ClinicalNeurophysiology3e2009 PDF

CLINICAL NEUROPHYSIOLOGY
Third Edition
SERIES EDITOR
Sid Gilman, MD, FRCP
William J. Herdman Distinguished University Professor of Neurology
University of Michigan
Contemporary Neurology Series
53 SLEEP MEDICINE 67 THE CLINICAL SCIENCE OF

Michael S. Aldrich, MD NEUROLOGIC REHABILITATION,
54 BRAIN TUMORS Second Edition
Harry S. Greenberg, MD, Bruce H. Dobkin, MD
William F. Chandler, MD, and 68 NEUROLOGY OF COGNITIVE AND
Howard M. Sandler, MD BEHAVIORAL DISORDERS
56 MYASTHENIA GRAVIS AND Orrin Devinsky, MD and
MYASTHENIC DISORDERS Mark D’Esposito, MD
Andrew G. Engel, MD, Editor 69 PALLIATIVE CARE IN NEUROLOGY
57 NEUROGENETICS Raymond Voltz, MD, James L. Bernat, MD,
Stefan-M. Pulst, MD, Dr. Med., Editor Gian Domenico Borasio, MD, DipPallMed,
58 DISEASES OF THE SPINE AND Ian Maddocks, MD, David Oliver, FRCGP,
SPINAL CORD and Russell K. Portenoy, MD
Thomas N. Byrne, MD, Edward C. Benzel, 70 THE NEUROLOGY OF EYE
MD, and Stephen G. Waxman, MD, PhD MOVEMENTS, Fourth Edition
59 DIAGNOSIS AND MANAGEMENT R. John Leigh, MD, FRCP and
OF PERIPHERAL NERVE DISORDERS David S. Zee, MD
Jerry R. Mendell, MD, John T. Kissel, MD, 71 PLUM AND POSNER’S DIAGNOSIS
and David R. Cornblath, MD OF STUPOR AND COMA,
60 THE NEUROLOGY OF VISION Fourth Edition
Jonathan D. Trobe, MD Jerome B. Posner, MD, Clifford B. Saper,
61 HIV NEUROLOGY MD, PhD, Nicholas D. Schiff, MD, and
Bruce James Brew, MBBS, MD, FRACP Fred Plum, MD
62 ISCHEMIC CEREBROVASCULAR 72 PRINCIPLES OF DRUG THERAPY
DISEASE IN NEUROLOGY,
Harold P. Adams, Jr., MD, Vladimir Second Edition
Hachinski, MD, and John W. Norris, MD Michael V. Johnston, MD and
63 CLINICAL NEUROPHYSIOLOGY OF Robert A. Gross, MD, PhD, Editors
THE VESTIBULAR SYSTEM, 73 NEUROLOGIC COMPLICATIONS OF
Third Edition CANCER, Second Edition
Robert W. Baloh, MD and Vicente Lisa M. DeAngelis, MD, and
Honrubia, MD Jerome B. Posner, MD
65 MIGRAINE: MANIFESTATIONS, 74 NEUROLOGIC COMPLICATIONS
PATHOGENESIS, AND OF CRITICAL ILLNESS,
MANAGEMENT, Second Edition Third Edition
Robert A. Davidoff, MD Eelco F.M. Wijdicks, MD, PhD, FACP
Third Edition
Edited by
Jasper R. Daube, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota;
Professor of Neurology, College of Medicine, Mayo Clinic
Devon I. Rubin, MD
Consultant, Department of Neurology, Mayo Clinic, Jacksonville, Florida;
Associate Professor of Neurology, College of Medicine, Mayo Clinic
1
2009
1
Oxford University Press, Inc., publishes works that further
Oxford University’s objective of excellence
in research, scholarship, and education.
Oxford New York

Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Copyright © 2009 by Mayo Foundation for Medical Education and Research

Published by Oxford University Press, Inc.
198 Madison Avenue, New York, New York 10016
www.oup.com
Oxford is a registered trademark of Oxford University Press
All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Mayo Foundation for Medical Education
and Research. Inquiries should be addressed to Scientific Publications,
Plummer 10, Mayo Clinic, 200 First Street SW, Rochester, MN 55905
Library of Congress Cataloging-in-Publication Data

Clinical neurophysiology / [edited by] Jasper R. Daube, Devon I. Rubin. — 3rd ed.
p. ; cm. — (Contemporary neurology series ; 75)
Includes bibliographical references and index.
ISBN-13: 978-0-19-538511-3
1. Electroencephalography. 2. Electromyography. 3. Nervous system—Diseases—Diagnosis.
4. Evoked potentials (Electrophysiology) 5. Neurophysiology.
I. Daube, Jasper R. II. Rubin, Devon I. III. Series.
[DNLM: 1. Nervous System Diseases—diagnosis. 2. Electroencephalography. 3. Electromyography.
4. Evoked Potentials. 5. Nervous System Diseases—therapy. 6. Neurophysiology.
W1 CO769N v.75 2009 / WL 141 C64155 2009]
RC386.6.E43C585 2009
616.8 047547—dc22 2008042251
Nothing in this publication implies that the Mayo Foundation endorses any of the products mentioned in this book. The
science of medicine is a rapidly changing field. As new research and clinical experience broaden our knowledge, changes
in treatment and drug therapy occur. The author and publisher of this work have checked with sources believed to be
reliable in their efforts to provide information that is accurate and complete, and in accordance with the standards
accepted at the time of publication. However, in light of the possibility of human error or changes in the practice of
medicine, neither the author, nor the publisher, nor any other party who has been involved in the preparation or
publication of this work warrants that the information contained herein is in every respect accurate or complete. The
opinion and suggestions herein are solely that of the author and not necessarily shared by the publisher. Readers are
encouraged to confirm the information contained herein with other reliable sources, and are strongly advised to check
the product information sheet provided by the pharmaceutical company for each drug they plan to administer.
9 8 7 6 5 4 3 2 1
Printed in the United States of America

on acid-free paper
Foreword
Clinical neurophysiology is a mature field. Many of its techniques are standard operating
procedures. Clinical neurophysiological approaches are logical extensions of the neurologic
examination and can add information that is helpful in making a diagnosis. Because of its use-
fulness, clinical neurophysiology is practiced by a large percentage of neurologists and physiatrists.
Even the neurologists and physiatrists who do not actively practice clinical neurophysiology
are expected to understand it. Therefore, it is important for practitioners to understand the
fundamental facts and principles of the field and to be current with key advances.
Clinical neurophysiology is also a large field. Like neurology, it encompasses a wide spectrum
of issues and illnesses, ranging from the peripheral nervous system to the central nervous system.
As in neurology, it is difficult to be an expert in all aspects of clinical neurophysiology, and most
practitioners have a focused interest in the field. However, as is true for neurology, clinical neu-
rophysiology has an essential unity. Problems are approached physiologically with methods that
measure the electric activity of the nervous system. This is another reason for practitioners to be
acquainted with the whole field even if they practice only a part of it.
Currently, there is considerable interest and activity in clinical neurophysiology. Numerous
societies in the United States and throughout the world are devoted to this field, and their mem-
bership is growing. The two principal societies in the United States are the American Association
of Neuromuscular and Electrodiagnostic Medicine, with its journal Muscle and Nerve, and the
American Clinical Neurophysiology Society, with its journal Journal of Clinical Neurophysiology.
The umbrella organization for the societies worldwide, the International Federation of Clinical
Neurophysiology has members in 58 countries and its journal Clinical Neurophysiology. There
are several examining bodies for competence in clinical neurophysiology. In the United States,
the American Board of Psychiatry and Neurology examines for competence in the broad field,
the American Board of Electrodiagnostic Medicine examines in the area commonly known as
electromyography, and the American Board of Clinical Neurophysiology examines in the area of
electroencephalography.
Where can a physician turn to learn the basics of clinical neurophysiology and be sure the infor-
mation is up-to-date? When Mayo Clinic neurologists speak about clinical neurophysiology, they
speak with special authority. The Mayo Clinic has been a central force in the United States in
many areas of the field. In the area of electromyography, Dr. Edward Lambert, a pioneer in the
field, made many basic observations that still guide current practice, and, of course, he identified
an illness that now bears his name. He has trained many leaders of modern electromyography in
the United States. In electroencephalography, Dr. Reginald Bickford was a pioneer and was active
in many areas, including evoked potentials and even early attempts at magnetic stimulation of the
brain. Many other leaders in electroencephalography have been at the Mayo Clinic, and four of
them, in addition to Dr. Bickford, have been presidents of the American Clinical Neurophysiology
Society. No one is better suited to orchestrate the writing of a textbook on Clinical Neurophysiol-
ogy than Dr. Jasper Daube, a leader in clinical neurophysiology at Mayo and former head of the
Neurology Department there. Dr. Daube is well recognized internationally as an expert in elec-
tromyography; he is very knowledgeable about all areas of the subject, basic and applied. He is an
outstanding leader with a gift for organization. He has been ably assisted by Dr. Devon I. Rubin,
another Mayo clinical neurophysiologist, who has worked with Dr. Daube on several projects in
addition to this book.
v
vi Foreword
For all these reasons, it is nice to see this third edition of Clinical Neurophysiology. Its many
chapters cover the field in a broad way. The first several chapters discuss the basic issues of neu-
ronal generators, biologic electricity, and measurement techniques central to all areas of clinical
neurophysiology. A new chapter in this section deals with fundamental membrane and synaptic
physiology. Next, the individual areas of the field are discussed: areas including classic electromyo-
graphy, electroencephalography, and evoked potentials and extending to autonomic nervous system
testing, sleep, surgical monitoring, motor control, vestibular testing, and magnetic stimulation. The
text is organized for physicians who want to know how to make an assessment of a particular symp-
tom, of a particular system, or for a particular disease. There is valuable information on the use
of clinical neurophysiologic testing in a practical setting. Each chapter has periodic summaries
of key points, which help understanding and learning. The book is profusely illustrated and has
an accompanying CD that includes instructions with pictures of standard nerve conduction stud-
ies, anatomical illustrations for performing needle EMG on standard muscles, protocols for the
approach to a wide range of clinical problems, and normal value tables.
Clinical neurophysiology, even though mature, like all other fields of medicine, is evolving.
Analysis and management of data are becoming more heavily computerized. New methods of
quantification are now possible and are being used clinically. New techniques are being devel-
oped. Perhaps most important, increasing emphasis is placed on how to improve patient care with
better integration of clinical neurophysiologic testing; the third section of the book is devoted to
these issues. This authoritative third edition should serve both students and practitioners, keeping
them up-to-date about important new advances.
Mark Hallett, MD
Human Motor Control Section
National Institute of Neurological Disorders and Stroke
Bethesda, Maryland
Past Editor-in-Chief, Clinical Neurophysiology
Preface
Clinical Neurophysiology is the result of more than 60 years of experience at the Mayo Clinic in
training clinicians in the neurophysiologic methods for assessing diseases of the central and periph-
eral nervous systems. The lectures and handouts that were developed initially by Doctors Reginald
Bickford and Edward Lambert in electroencephalography and electromyography, respectively,
were the seeds of what has grown into the far-reaching field of endeavor of clinical neurophysi-
ology at Mayo Clinic. The clinical neurophysiology teaching programs at Mayo Clinic Rochester,
Jacksonville, and Arizona have continued to evolve into a formal, unified, 2-month course in clini-
cal neurophysiology that provides trainees with the knowledge and experience needed to apply the
principles of neurophysiology clinically.
The development of clinical neurophysiology at Mayo has paralleled developments in the
field of medicine at large. The expansion during the past 25 years of neurophysiology of dis-
eases of the central and peripheral nervous system has been recognized by the American Board
of Psychiatry and Neurology, by the American Board of Medical Specialties with a Special
Qualifications Examination in Clinical Neurophysiology, and by the Accreditation Council for
Graduate Medical Education Residency Review Committee for postresidency fellowships in
Clinical Neurophysiology.1
The Mayo course in clinical neurophysiology serves as an introduction to clinical neurophysiol-
ogy for residents, fellows, and other trainees. The course includes lectures, small group seminars,
practical workshops, and clinical experience in each of the areas of clinical neurophysiology. The
faculty for the course consists entirely of Mayo Clinic staff members. These staff members are the
authors of the chapters of this textbook.
Over the years, the material for the clinical neurophysiology course was consolidated from indi-
vidual lecture handouts into manuals. Persons outside Mayo who had learned about these manuals
by word of mouth increasingly requested them. The success of these manuals prompted us to
publish the first edition of Clinical Neurophysiology in 1996 and a second edition in 2002. The
continued evolution and expansion of the field of clinical neurophysiology has resulted in this third
edition.
The organization of our textbook is unique: it is built around the concept of testing systems
within the nervous system, rather than separated by individual techniques. The book consists of
three major sections. The first section is a review of the basics of clinical neurophysiology, knowl-
edge that is common to each of the areas of clinical neurophysiology. The second section considers
the assessment of diseases by anatomical system. Thus, methods for assessing the motor system are
grouped together, followed by those for assessing the sensory system, higher cortical functions, and
the autonomic nervous system. The third section explains how clinical neurophysiologic techniques
are used in the clinical assessment of diseases of the nervous system.
This third edition includes new approaches, such as those described in the new chapters on EEG
coregistration with MRI imaging in epilepsy and motor unit number estimate studies in peripheral
neuromuscular diseases. The underlying physiologic and electronic principles in Clinical Neuro-
physiology have not changed but the approach to teaching them with bullet points and key points
has provided simplification and clarification. The clinical problems in which each of the clinical
neurophysiologic approaches can add to the diagnosis and management of neurologic disease have
been detailed, especially the assessment of clinical symptom complexes with electroencephalog-
raphy (EEG). The discussion of pediatric EEG disorders, ambulatory EEG, new equipment and
digital analyses, magneto-EEG, electromyographic (EMG) techniques, motor unit number esti-
mates, myoclonus on surface EMG, segmental sympathetic reflex, and postural hypotension has
vii
viii Preface
been expanded. Chapters on EMG quantification and single fiber EMG have been reorganized,
and major revisions have been made in the discussion of sensory potentials, somatosensory evoked
potentials, acoustic reflex testing, cardiovagal function, physiologic testing of sleep, and assessment
of sleep disorders. New approaches have been expanded in each of the four chapters on monitoring
neural function during surgery, particularly with motor evoked potentials.
For the first time, this edition also includes a CD with material immediately available during clin-
ical electromyography. Pictures are provided, depicting nerve conduction study and somatosensory
evoked potential techniques used in the Mayo Clinic EMG Laboratories including accompany-
ing Mayo normal values, images depicting muscle surface anatomy with superimposed illustrated
muscles for localization during needle EMG, and algorithms used for assessment of common
problems in the clinical EMG laboratory and during intraoperative monitoring. The CD also con-
tains the “EMG Sound Simulator and Synthesizer,” a unique, downloadable, interactive program
that teaches EMG waveform recognition, and motor unit potential assessment and interpretation.
Interactive CNP learning has been shown to be more effective than lectures.2
REFERENCES
1. Burns, R., J. Daube, and H. Royden Jones. 2000. Clinical neurophysiology training and certification in the United States:
2000: American Board of Psychiatry and Neurology, Neurology Residency Review Committee. Neurology 55:1773–8.
2. Schuh, L., D. E. Burdette, L. Schultz, and B. Silver. 2008. Learning clinical neurophysiology: Gaming is better than
lectures. Journal of Clinical Neurophysiology 25(3):167–9.
Acknowledgments
The authors of the third edition of Clinical Neurophysiology have made our work as editors
both educational and enjoyable. Each of the authors is active in clinical neurophysiology practice,
education, and research. They bring their experiences to bear in the chapters they have written.
Thus, our task was the remarkably easy one of organizing and coordinating the material. The editors
and authors appreciate the skill and professionalism of Roberta Schwartz of the Sections of Scien-
tific Publications; she has had an integral part in the development of this textbook. The work of the
medical illustrators, Paul Honerman, David Factor, and David Cheney, and of Raj Alphonse and
others in the Media Support Services at the Mayo Clinic, have been invaluable in the development
and preparation of the supplemental material including that in the accompanying CD.
Mayo Neurology leadership has continued to encourage and support the Division of Clinical
Neurophysiology in its combined efforts to provide trainees with the broad background of knowl-
edge they will need as they enter active practice. This support has provided strong encouragement
for this book. The staff in the Division of Neurophysiology—including staff at all three Mayo
Clinic sites in Jacksonville, Florida; Rochester, Minnesota; and Scottsdale, Arizona—have con-
tributed in a major way to the clinical neurophysiology course on which this textbook is based.
The laboratory directors have been particularly important: Drs. Eric Sorenson, Devon Rubin, and
Benn Smith, chairs of the Divisions of Clinical Neurophysiology and directors of the Electromyo-
graphy Laboratories at the three Mayo Clinics; Dr. Phillip Low, director of the Autonomic Reflex
Laboratory and the Nerve Physiology Laboratory; Dr. Elson So, director of the Electroencephalo-
graphic Laboratory; Dr. Michael Silber, director of the Sleep Disorders Center; and Dr. Robert
Fealey, director of the Thermoregulatory Sweat Laboratory.
The support of the Mayo Foundation has been critical in the development of new directions and
unique training programs in clinical neurophysiology. We acknowledge not only this support but
also the help given by many others: the trainees who have participated in our clinical neurophys-
iology program and the students in our courses in continuing medical education who have given
us feedback on our teaching material, the technicians who have been a major part of our teaching
program and who have provided a helpful critique of our activities, Jean M. Smith and the other
secretarial staff who have worked diligently to keep the project on track, and other physicians at our
institution who have found our help in clinical neurophysiology useful in the care of their patients.
Jasper R. Daube, MD and Devon I. Rubin, MD
ix
This page intentionally left blank
Contents
Contributors xxv
SECTION 1 BASIC NEUROPHYSIOLOGY AND AN

INTRODUCTION TO ANALYSIS OF ELECTROPHYSIOLOGIC
WAVEFORMS
1. ELECTRICITY AND ELECTRONICS FOR CLINICAL

NEUROPHYSIOLOGY 5
Terrence D. Lagerlund
BASIC PRINCIPLES AND DEFINITIONS IN ELECTRICITY 5
CIRCUIT ANALYSIS 7
RESISTIVE–CAPACITIVE AND RESISTIVE–INDUCTIVE CIRCUITS 9
CIRCUITS CONTAINING INDUCTORS AND CAPACITORS 10
FILTER CIRCUITS 14
TRANSISTORS AND AMPLIFIERS 16
2. ELECTRIC SAFETY IN THE LABORATORY AND HOSPITAL 21

INTRODUCTION 21
ELECTRIC POWER DISTRIBUTION SYSTEMS 21
ELECTRIC SHOCK 22
LEAKAGE CURRENT 25
ELECTRIC SAFETY PRINCIPLES AND IMPLEMENTATION 27
ELECTRIC STIMULATION SAFETY 30
3. VOLUME CONDUCTION 33
Terrence D. Lagerlund, Devon I. Rubin, and Jasper R. Daube
PRINCIPLES 33
SOURCES OF ELECTRICAL POTENTIALS 34
xi
xii Contents
CURRENT SOURCES: MONOPOLES, DIPOLES, AND

QUADRUPOLES 38
VOLUME CONDUCTION THEORY: ELECTRIC PROPERTIES IN
VOLUME CONDUCTORS 41
APPLICATIONS OF VOLUME CONDUCTION PRINCIPLES 43
4. DIGITAL SIGNAL PROCESSING 53

DIGITAL COMPUTERS IN CLINICAL NEUROPHYSIOLOGY 53
DIGITIZATION 56
COMMON USES OF DIGITAL PROCESSING 59
AVERAGING 60
DIGITAL FILTERING 62
TIME AND FREQUENCY DOMAIN ANALYSIS 63
5. BASICS OF NEUROPHYSIOLOGY 69
Jasper R. Daube and Squire M. Stead
INTRODUCTION 69
CELL MEMBRANE 69
RESTING POTENTIAL 76
LOCAL POTENTIALS 78
ACTION POTENTIALS 82
SYNAPTIC TRANSMISSION 88
CLINICAL CORRELATIONS 93
6. ELECTROPHYSIOLOGIC GENERATORS IN CLINICAL

NEUROPHYSIOLOGY 97
STRUCTURAL GENERATORS 97
7. WAVEFORMS AND ARTIFACTS 103

Jasper R. Daube
INTRODUCTION 103
CONTINUOUS WAVEFORMS 103
EVENT RECORDING 106
COMBINATIONS OF CONTINUOUS WAVEFORMS AND
EVENTS 108
WAVEFORM ALTERATIONS 108
Contents xiii
PHYSIOLOGIC ALTERATION OF WAVEFORMS 108

ARTIFACTUAL WAVEFORMS 111
SECTION 2 ELECTROPHYSIOLOGIC ASSESSMENT OF

NEURAL FUNCTION
Part A Assessment of Cortical Function
8. ELECTROENCEPHALOGRAPHY: ADULT, NORMAL, AND BENIGN

VARIANTS 119
Barbara F. Westmoreland
INTRODUCTION 119
RECORDING THE ELECTROENCEPHALOGRAM 119
DISPLAY OF EEG ACTIVITY 119
ACTIVATION PROCEDURES 120
ARTIFACTS 123
NORMAL EEG ACTIVITY OF ADULTS 124
BENIGN VARIANTS 130
9. EPILEPTIFORM ACTIVITY 137

Joseph F. Drazkowski
INTRODUCTION AND OVERVIEW OF EPILEPTIFORM
ACTIVITY 137
SPECIFIC FOCAL INTERICTAL DISCHARGES 138
GENERALIZED EPILEPTIFORM PATTERNS 141
ICTAL DISCHARGES 145
EPILEPTIFORM ACTIVITY WITH A POTENTIAL SEIZURE
ASSOCIATION 148
10. ADULT EEG: ABNORMAL NONEPILEPTIFORM ACTIVITY 151

INTRODUCTION 151
TYPES OF EEG ABNORMALITIES 152
FOCAL INTRACRANIAL PROCESSES CAUSING EEG
ABNORMALITIES 154
ELECTROENCEPHALOGRAPHIC MANIFESTATIONS OF DIFFUSE
DISORDERS 159
EVALUATION FOR SUSPECTED BRAIN DEATH 164
xiv Contents
11. ELECTROENCEPHALOGRAPHY: ELECTROENCEPHALOGRAMS OF

INFANTS AND CHILDREN 167
INTRODUCTION 167
NEONATAL EEG PATTERNS 168
DEVELOPMENTAL CHANGES DURING INFANCY, CHILDHOOD, AND
ADOLESCENCE 173
BENIGN VARIANTS IN CHILDREN 175
ABNORMALITIES 175
12. AMBULATORY ELECTROENCEPHALOGRAPHY 187

Jeffrey R. Buchhalter
INTRODUCTION 187
INDICATIONS 187
TECHNOLOGY 188
CLINICAL APPLICATIONS 189
13. PROLONGED VIDEO ELECTROENCEPHALOGRAPHY 193

Cheolsu Shin
INTRODUCTION 193
EQUIPMENT 194
CLINICAL APPLICATION 195
14. ELECTROENCEPHALOGRAPHIC SPECIAL STUDIES 203

Gregory A. Worrell and Terrence D. Lagerlund
INTRODUCTION 203
QUANTITATIVE METHODS OF ELECTROENCEPHALOGRAPHIC
ANALYSIS 203
MAGNETOENCEPHALOGRAPHY 211
15. ELECTROENCEPHALOGRAPHY IN THE SURGICAL EVALUATION

OF EPILEPSY 215
BACKGROUND 215
PRESURGICAL SELECTION AND EVALUATION 216
ROUTINE EEG IN THE SURGICAL EVALUATION OF PATIENTS WITH
SEIZURES 218
PREOPERATIVE VIDEO-EEG MONITORING 218
Contents xv
PRESURGICAL EVALUATION WITH CONTINUOUS OR CHRONIC

INTRACRANIAL MONITORING 219
INTRAOPERATIVE ELECTROCORTICOGRAPHY 223
16. MOVEMENT-RELATED CORTICAL POTENTIALS AND

EVENT-RELATED POTENTIALS 229
Virgilio Gerald H. Evidente and John N. Caviness
MOVEMENT-RELATED CORTICAL POTENTIALS 229
EVENT-RELATED POTENTIALS 232
Part B Sensory Pathways
17. SENSORY NERVE ACTION POTENTIALS 239

Eric J. Sorenson
INTRODUCTION 239
PATHOPHYSIOLOGY OF SNAPs 240
METHODS OF STUDY OF SNAPs 243
MEASUREMENTS 247
TECHNICAL FACTORS 248
PLANNING THE STUDY AND FINDINGS IN DISEASES 251
18. SOMATOSENSORY EVOKED POTENTIALS 257

Jonathan L. Carter and J. Clarke Stevens
INTRODUCTION 257
GENERATORS AND ORIGIN OF SEPs 258
METHODS 258
RECORDING 260
LOCALIZATION 268
19. BRAIN STEM AUDITORY EVOKED POTENTIALS IN CENTRAL

DISORDERS 281
Jonathan L. Carter
INTRODUCTION 281
AUDITORY ANATOMY AND PHYSIOLOGY 283
GENERATORS OF THE BRAIN STEM AUDITORY EVOKED
POTENTIALS 283
BRAIN STEM AUDITORY EVOKED POTENTIALS: METHODS 284
FACTORS AFFECTING THE BAEP RESPONSE 286
xvi Contents
INTERPRETATION OF BAEPs 287

20. AUDIOGRAM, ACOUSTIC REFLEXES, AND EVOKED OTOACOUSTIC

EMISSIONS 295
Christopher D. Bauch and Wayne O. Olsen
INTRODUCTION 295
AUDIOGRAM 296
ACOUSTIC REFLEX 298
EVOKED OTOACOUSTIC EMISSIONS 301
APPLICATIONS 303
21. BRAIN STEM AUDITORY EVOKED POTENTIALS IN PERIPHERAL

ACOUSTIC DISORDERS 305
Christopher D. Bauch
INTRODUCTION 305
STIMULI 306
ELECTRODES 306
INTERPRETATION 306
APPLICATIONS 308
22. VISUAL EVOKED POTENTIALS 311

Jonathan L. Carter
INTRODUCTION 311
VISUAL SYSTEM ANATOMY AND PHYSIOLOGY 312
VISUAL EVOKED POTENTIALS: METHODS 312
FACTORS AFFECTING THE VEP RESPONSE 314
INTERPRETATION OF VEPs 317
LOCALIZATION OF VISUAL SYSTEM LESIONS 318
Part C Motor Pathways
23. COMPOUND MUSCLE ACTION POTENTIALS 327

James C. Watson and Jasper R. Daube
INTRODUCTION 327
GENERAL CLINICAL APPLICATIONS 328
RECORDING CMAPs 329
STIMULATION 332
Contents xvii
CMAP MEASUREMENTS 334

F WAVES 338
AXON REFLEXES (A WAVES) 342
PHYSIOLOGIC VARIABLES AFFECTING THE CMAP 343
CMAP CHANGES IN DISEASE 344
FINDINGS IN PERIPHERAL NERVE DISORDERS 346
24. ASSESSING THE NEUROMUSCULAR JUNCTION WITH REPETITIVE

STIMULATION STUDIES 369
Andrea J. Boon
INTRODUCTION 369
ANATOMY AND PHYSIOLOGY OF THE NEUROMUSCULAR
JUNCTION 370
TECHNIQUE 372
CRITERIA OF ABNORMALITY 377
RAPID RATES OF STIMULATION 378
SELECTION OF NERVE–MUSCLE COMBINATIONS 379
CLINICAL CORRELATIONS 380
25. MOTOR EVOKED POTENTIALS 385

Jeffrey A. Strommen
INTRODUCTION 385
TECHNIQUE 386
MEP PHARMACOLOGY 392
APPLICATIONS 393
CONTRAINDICATIONS AND RISKS 395
Part D Assessing the Motor Unit
26. ASSESSING THE MOTOR UNIT WITH NEEDLE

ELECTROMYOGRAPHY 403
Devon I. Rubin
INTRODUCTION 404
KNOWLEDGE BASE OF NEEDLE EMG 404
TECHNIQUE OF NEEDLE EXAMINATION 405
CONDUCTING THE NEEDLE EXAMINATION 405
xviii Contents
POTENTIAL COMPLICATIONS DURING NEEDLE

EXAMINATION 409
EMG SIGNAL ANALYSIS 412
NEEDLE ELECTRODE CHARACTERISTICS 412
SKILLS OF EMG WAVEFORM RECOGNITION 413
ORIGIN OF EMG POTENTIALS 415
NORMAL EMG ACTIVITY 417
ABNORMAL SPONTANEOUS ELECTRIC ACTIVITY 424
ABNORMAL ELECTRICAL ACTIVITY—VOLUNTARY MUPs 437
ABNORMAL ELECTRICAL ACTIVITY—DISORDERS OF CENTRAL
CONTROL 445
PATTERNS OF ABNORMALITIES 445
27. QUANTITATIVE ELECTROMYOGRAPHY 451

Benn E. Smith
INTRODUCTION 451
CHARACTERISTICS OF THE MOTOR UNIT POTENTIAL 453
CHARACTERISTICS OF THE RECORDING EQUIPMENT 454
PROPERTIES OF MUPs EVALUATED USING STANDARD
ELECTRODES 456
PROPERTIES OF MUPs MEASURABLE ONLY WITH SPECIAL
ELECTRODES 460
QUANTITATIVE ANALYSIS OF SINGLE MUPs 462
PROPERTIES OF INTERFERENCE PATTERN AND METHODS OF
INTERFERENCE PATTERN ANALYSIS 467
TURNS AND AMPLITUDE ANALYSIS OF THE INTERFERENCE
PATTERN 469
POWER-SPECTRUM ANALYSIS 471
AUTOMATED METHODS OF ANALYSIS OF SPONTANEOUS
ACTIVITY 472
28. SINGLE FIBER ELECTROMYOGRAPHY 475

C. Michel Harper, Jr.
INTRODUCTION 475
TECHNIQUE 477
PITFALLS OF SFEMG 483
CLINICAL APPLICATIONS OF SFEMG 486
Contents xix
29. QUANTITATIVE MOTOR UNIT NUMBER ESTIMATES 493

Jasper R. Daube
INTRODUCTION 493
MUNE BY STANDARD EMG 495
MUNE BY STANDARD MOTOR NCS 496
QUANTITATIVE MUNE 497
Part E Reflexes and Central Motor Control
30. H REFLEXES 519

Ruple S. Laughlin
INTRODUCTION 519
PHYSIOLOGIC BASIS 519
TECHNIQUE 521
PEDIATRIC H REFLEXES 524
31. CRANIAL REFLEXES AND RELATED TECHNIQUES 529

Benn E. Smith
INTRODUCTION 529
BLINK REFLEX 530
LATERAL SPREAD OF THE FACIAL NERVE RESPONSE:
ASSESSMENT OF FACIAL SYNKINESIS AND HEMIFACIAL
SPASM 535
JAW JERK (MASSETER REFLEX) 537
MASSETER INHIBITORY REFLEX 538
GREAT AURICULAR SENSORY NERVE CONDUCTION
STUDIES 540
TRIGEMINAL CONTACT HEAT EVOKED POTENTIAL STIMULATOR
STUDIES 540
32. LONG LATENCY REFLEXES AND THE SILENT PERIOD 543

John N. Caviness
INTRODUCTION 543
LONG LATENCY REFLEXES 543
THE SILENT PERIOD 546
xx Contents
33. MOVEMENT DISORDERS 551

John N. Caviness
INTRODUCTION 551
TECHNIQUES 552
EEG 553
SURFACE EMG: NORMAL PATTERNS 554
TREMOR 554
MYOCLONUS 559
PSYCHOGENIC JERKS 568
STARTLE DISORDERS 568
PERIODIC LIMB MOVEMENTS OF SLEEP 568
DYSTONIA 569
TICS, CHOREA, AND ATHETOSIS 571
VOLUNTARY MOVEMENT ABNORMALITIES 571
34. VERTIGO AND BALANCE 575

David A. Zapala and Robert H. Brey
INTRODUCTION 575
LABORATORY EXAMINATION: VOR-BASED MEASURES 582
LABORATORY EXAMINATION: VSR-BASED MEASURES 599
CLINICAL APPLICATIONS OF VESTIBULAR TESTING:
ASSESSING SENSORINEURAL SYNDROMES OF THE
LABYRINTH 607
VESTIBULAR REHABILITATION 610
Part F Autonomic Function
35. AUTONOMIC PHYSIOLOGY 617

William P. Cheshire, Jr.
INTRODUCTION 617
SYMPTOMS AND DISEASES 617
GENERAL ORGANIZATION OF THE AUTONOMIC
SYSTEM 619
SYMPATHETIC FUNCTION 621
SYMPATHETIC INNERVATION OF THE SKIN 622
MUSCLE SYMPATHETIC ACTIVITY 623
Contents xxi
AUTONOMIC CONTROL OF HEART RATE 624

CARDIOVASCULAR REFLEXES 624
MAINTENANCE OF POSTURAL NORMOTENSION 626
36. QUANTITATIVE SUDOMOTOR AXON REFLEX AND

RELATED TESTS 629
Phillip A. Low
INTRODUCTION 629
LABORATORY EVALUATION OF AUTONOMIC FUNCTION 629
QUANTITATIVE SUDOMOTOR AXON REFLEX TEST 630
IMPRINT METHODS OF SWEAT MEASUREMENT 634
37. EVALUATION OF ADRENERGIC FUNCTION 637

Phillip A. Low
INTRODUCTION 637
SKIN VASOMOTOR REFLEXES 637
38. THERMOREGULATORY SWEAT TEST 645

Robert D. Fealey
INTRODUCTION 645
ROLE OF THERMOREGULATORY SWEAT TESTING:
CLINICAL SYNDROMES AND PROBLEMS
EVALUATED 646
METHOD 650
THERMOREGULATORY SWEAT DISTRIBUTION 653
REPORTING RESULTS 656
DIFFICULTIES AND PITFALLS IN INTERPRETATION 657
39. CARDIOVAGAL REFLEXES 661

INTRODUCTION 661
HEART RATE RESPONSE TO DEEP BREATHING 662
THE VALSALVA MANEUVER 665
CARDIOVAGAL SCORING 670
POWER SPECTRUM ANALYSIS 670
HEART RATE RESPONSE TO STANDING 671
OTHER TESTS OF CARDIOVAGAL FUNCTION 672
xxii Contents
40. ELECTROPHYSIOLOGY OF PAIN 677

Rose M. Dotson and Paola Sandroni
INTRODUCTION 677
QUANTITATIVE SENSORY TEST 678
AUTONOMIC TESTS 681
MICRONEUROGRAPHY 683
LASER EVOKED POTENTIALS 684
CONTACT HEAT EVOKED POTENTIALS 688
Part G Sleep and Consciousness
41. ASSESSMENT OF SLEEP AND SLEEP DISORDERS 697

Michael H. Silber, Cameron D. Harris, and Peter J. Hauri
INTRODUCTION 697
TECHNIQUES USED IN STUDYING SLEEP 698
STAGING OF SLEEP 706
ASSESSING RESPIRATION DURING SLEEP 712
ASSESSING MOVEMENTS IN SLEEP 715
ASSESSING OTHER PHYSIOLOGIC VARIABLES 718
PERFORMANCE OF A SLEEP STUDY 718
ASSESSING SLEEP DISORDERS 719
Part H Intraoperative Monitoring
42. CEREBRAL FUNCTION MONITORING 727

Elson L. So and Frank W. Sharbrough
INTRODUCTION 727
TECHNICAL FACTORS IN INTRAOPERATIVE EEG MONITORING 727
EFFECTS OF ANESTHESIA ON ELECTROENCEPHALOGRAPHY
SYMMETRICAL EEG PATTERNS DURING ANESTHESIA 731
PREOPERATIVE FOCAL ABNORMALITIES SEEN WITH
ANESTHESIA 732
EEG CHANGES DURING CAROTID ENDARTERECTOMY 733
SEP RECORDING DURING CAROTID ENDARTERECTOMY 735
OTHER MONITORING TECHNIQUES DURING CAROTID
ENDARTERECTOMY 735
Contents xxiii
EEG MONITORING DURING CARDIAC SURGERY 736

BISPECTRAL ANALYSIS OF EEG FOR MONITORING DEPTH OF
ANESTHESIA 736
EEG MONITORING FOR EPILEPSY SURGERY 736
43. BRAIN STEM AND CRANIAL NERVE MONITORING 739

Brian A. Crum
INTRODUCTION 739
METHODS 740
APPLICATIONS 743
44. SPINAL CORD MONITORING 751

Jeffrey A. Strommen
INTRODUCTION 751
GENERAL PRINCIPLES OF INTRAOPERATIVE
MONITORING 752
EQUIPMENT AND ELECTRICAL SAFETY 753
MONITORING METHODS—SOMATOSENSORY EVOKED
POTENTIALS 754
MOTOR EVOKED POTENTIALS 760
ELECTROMYOGRAPHY AND NERVE CONDUCTION
STUDIES 765
TYPES OF SPINAL SURGERIES 768
45. PERIPHERAL NERVOUS SYSTEM MONITORING 777

INTRODUCTION 777
METHODS 778
APPLICATIONS 780
SECTION 3 APPLICATIONS OF CLINICAL

NEUROPHYSIOLOGY: ASSESSING SYMPTOM COMPLEXES
AND DISEASE ENTITIES
46. ASSESSING CENTRAL NERVOUS SYSTEM SYMPTOMS 791

Elson L. So
INTRODUCTION 791
ASSESSMENT OF MOTOR SYMPTOMS OF CENTRAL ORIGIN 791
ASSESSMENT OF SENSORY SYMPTOMS OF CENTRAL ORIGIN 792
xxiv Contents
ASSESSING IMPAIRMENT OF CONSCIOUSNESS AND COGNITION 792

ASSESSING IMPAIRMENT OF VISCERAL FUNCTION AND SLEEP 792
IDENTIFYING DISEASE TYPES 793
PROGNOSIS 793
ASSESSING CLINICAL DISORDERS WITH EEG 793
47. APPLICATION OF CLINICAL NEUROPHYSIOLOGY: ASSESSING

PERIPHERAL NEUROMUSCULAR SYMPTOM COMPLEXES 801
Devon I. Rubin and Jasper R. Daube
CLINICAL NEUROPHYSIOLOGY IN THE ASSESSMENT OF
PERIPHERAL NERVOUS SYSTEM DISORDERS 802
ASSESSING CLINICAL DISORDERS: ASSESSMENT WITH EMG
AND NCS 806
RADICULOPATHIES 806
COMMON FOCAL MONONEUROPATHIES 808
PERIPHERAL NEUROPATHY 810
BRACHIAL PLEXOPATHY 817
GENERALIZED WEAKNESS 822
MYOPATHY 823
MYALGIAS, MUSCLE STIFFNESS, AND EPISODIC MUSCLE
WEAKNESS 827
NMJ DISORDERS 827
POLYRADICULOPATHY 829
MOTOR NEURON DISEASE 830
FACIAL WEAKNESS 831
ANOMALOUS INNERVATION 831
UNEXPECTED FINDINGS ON NERVE CONDUCTION STUDIES: CAUSE
AND ACTION 835
Glossary of Electrophysiologic Terms 839
Index 869
Contributors
Christopher D. Bauch, PhD

Chair, Division of Audiology, Mayo Clinic, Rochester, Minnesota; Associate Professor of
Audiology, College of Medicine, Mayo Clinic
Andrea J. Boon, MD
Consultant, Department of Physical Medicine and Rehabilitation, Mayo Clinic, Rochester,
Minnesota; Assistant Professor of Neurology and of Physical Medicine and Rehabilitation,
College of Medicine, Mayo Clinic
Robert H. Brey, PhD

Consultant, Division of Audiology, Mayo Clinic, Rochester, Minnesota; Professor of Audiology,
College of Medicine, Mayo Clinic
Jeffrey R. Buchhalter, MD, PhD

Supplemental Consultant, Department of Neurology, Mayo Clinic, Scottsdale, Arizona; Associate
Professor of Neurology, College of Medicine, Mayo Clinic; Chief, Division of Pediatric
Neurology, Phoenix Children’s Hospital, Phoenix, Arizona
Jonathan L. Carter, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Associate Professor
of Neurology, College of Medicine, Mayo Clinic
John N. Caviness, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Professor of
Neurology, College of Medicine, Mayo Clinic
William P. Cheshire, Jr., MD

Consultant, Department of Neurology, Mayo Clinic, Jacksonville, Florida; Professor of
Brian A. Crum, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Assistant Professor
Jasper R. Daube, MD
Rose M. Dotson, MD
Formerly, Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Formerly,
Assistant Professor of Neurology, College of Medicine, Mayo Clinic; Presently, University
Hospitals Medical Pediatrics, Chardon, Ohio
Joseph F. Drazkowski, MD
Consultant, Department of Neurology, Mayo Clinic, Scottsdale, Arizona; Associate Professor of
xxv
xxvi Contributors
Virgilio G. H. Evidente, MD
Robert D. Fealey, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Assistant Professor
C. Michel Harper, Jr., MD

Cameron D. Harris
Associate, Sleep Disorders Center, Mayo Clinic, Rochester, Minnesota; Assistant Professor of
Medicine, College of Medicine, Mayo Clinic
Peter J. Hauri, PhD

Supplemental Consultant, Department of Psychiatry and Psychology, Mayo Clinic, Rochester,
Minnesota; Emeritus Professor of Psychology, College of Medicine, Mayo Clinic
Terrence D. Lagerlund, MD
Ruple S. Laughlin, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Instructor in
Phillip A. Low, MD
Wayne O. Olsen, MD
Emeritus Consultant, Department of Otorhinolaryngology, Mayo Clinic, Rochester, Minnesota;
Emeritus Professor of Otolaryngology, College of Medicine, Mayo Clinic
Devon I. Rubin, MD
Consultant, Department of Neurology, Mayo Clinic, Jacksonville, Florida; Associate Professor of
Paola Sandroni, MD, PhD

Frank W. Sharbrough, MD
Cheolsu Shin, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Associate Professor of
Neurology and Assistant Professor of Pharmacology, College of Medicine, Mayo Clinic
Michael H. Silber, MB ChB

Contributors xxvii
Benn E. Smith, MD
Elson L. So, MD
Eric J. Sorenson, MD
Squire M. Stead, MD, PhD

Senior Associate Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota;
Assistant Professor of Neurology, College of Medicine, Mayo Clinic
J. Clarke Stevens, MD
Jeffrey A. Strommen, MD
Consultant, Department of Physical Medicine and Rehabilitation, Mayo Clinic, Rochester,
Minnesota; Assistant Professor of Physical Medicine and Rehabilitation, College of Medicine,
Mayo Clinic
James C. Watson, MD
Consultant, Department of Neurology, Mayo Clinic, Rochester, Minnesota; Assistant Professor of
Barbara F. Westmoreland, MD
Gregory A. Worrell, MD, PhD

David A. Zapala, PhD

Consultant, Section of Audiology, Mayo Clinic, Jacksonville, Florida; Assistant Professor of
Audiology, College of Medicine, Mayo Clinic
Third Edition
SECTION 1
BASIC NEUROPHYSIOLOGY
AND AN INTRODUCTION
TO ANALYSIS OF
ELECTROPHYSIOLOGIC
WAVEFORMS
Clinical neurophysiology is an area of medical The first section of this introductory textbook
practice focused primarily on measuring func- reviews the basic concepts of neurophysiol-
tion in the central and peripheral nervous ogy, including the generation, recording, and
systems, the autonomic nervous system, and analysis of the waveforms studied in the prac-
muscles. The specialty identifies and charac- tice of clinical neurophysiology. The principles
terizes diseases of these areas, understands of electricity and electronics needed to make
their pathophysiology, and, to a limited extent, the recordings are reviewed in Chapter 1.
treats them. Clinical neurophysiology relies To make the appropriate measurements, clin-
entirely on the measurement of ongoing ical neurophysiologists rely on equipment
function—either spontaneous or in response with technical specifications; this requires that
to a defined stimulus—in a patient. Each of they understand the basic principles reviewed
the clinical neurophysiology methods measures in Chapter 1. All electrical stimulation and
function by recording alterations in physiol- recording methods require applying electrical
ogy as manifested by changes in electrical connections that pass small amounts of cur-
waveforms, electromagnetic fields, force, or rent through human tissue. Although the risks
secretory activities. Each of these variables is of harm from this current flow are small, they
measured as a waveform that changes over must be understood. The principles of elec-
time. Electrical measurements in which the trical safety necessary to minimize, reduce, or
voltage or current flow associated with activ- eliminate any risk are discussed in Chapter 2.
ity is plotted on a temporal basis are the These unchanging principles are of critical
most common measurements used in clini- importance to the practice of clinical neuro-
cal neurophysiology, but other measures such physiology.
as blood pressure, pulse, sweat production, The circuits reviewed in Chapter 1 describe
and respiration are also sometimes measured. the familiar forms of electricity found in the
Knowledge of the basics of neurophysiology, home and in business. Electrical recordings
including the origin and generation of elec- made from human tissue are distinctly differ-
trical activity and the recording, measure- ent because the electric currents are carried
ment, and analysis of the waveforms generated by charged ions that are present through-
by the electrical signals, is critical in learn- out the tissue, rather than by electrons as in
ing how to perform and apply the methods wires. Electric currents flowing throughout the
of clinical neurophysiology in the study of human body are limited by the resistance and
disease. capacitance of the tissues. While the resistance
3
4 Clinical Neurophysiology
and capacitance vary among different tissues, for selecting, displaying, and storing the
it does not stop it, resulting in widespread flow waveforms are described in Chapter 4.
of electricity throughout the body, referred to Virtually all tissues (and the cells compos-
as volume conduction. Volume conduction pro- ing them) in the human body have elec-
duces the unique aspects of the generation and tric potentials associated with their activities.
recording of physiologic waveforms recorded These potentials are much larger for nerve
from human tissue. The immutable principles and muscle tissue than for other tissues and
of volume conduction described in Chapter 3 can easily be recorded for analyzing function
are applicable to the many forms of electri- and its alteration with disease, just as they can
cal recording used in clinical neurophysiology, for the heart. Chapters 5 and 6 review the
whether the waveforms are recorded from the basic concepts of neurophysiology, including
head (electroencephalography), nerves (nerve the generators and processes on a cellular level
conduction studies), muscles (electromyogra- that give rise to the signals recorded in clini-
phy), or skin (autonomic function testing). cal neurophysiology. Chapter 6 describes the
Measurement of current flow, or poten- underlying physiology of all electrical wave-
tial differences, between areas of the body forms, whatever be their sources. The range
was first made using analog electronic devices of alterations that occur in these waveforms in
that have been replaced entirely by digital disease and the electric artifacts that occur in
recordings throughout clinical neurophysiol- association with the physiologic waveforms are
ogy. The basic principles of digital techniques reviewed in Chapter 7.
Chapter 1
Electricity and Electronics for

Clinical Neurophysiology
BASIC PRINCIPLES AND CIRCUITS CONTAINING

DEFINITIONS IN ELECTRICITY INDUCTORS AND CAPACITORS
Electric Charges and Force Inductive–Capacitive Circuits
Electric Potential Inductive–Resistive–Capacitive
Electric Current and Resistors Circuits
Capacitors Root-Mean-Square Potentials or
Coils (Inductors) Currents
Calculation of Reactance
CIRCUIT ANALYSIS Calculation of Impedance and the
Kirchhoff’s First Law Phenomenon of Resonance
Rules for Seats of Electromotive
Force, Resistors, Capacitors, and FILTER CIRCUITS
Inductors High-Pass Filters
Kirchhoff’s Second Law Low-Pass Filters
RESISTIVE–CAPACITIVE AND TRANSISTORS AND AMPLIFIERS
RESISTIVE–INDUCTIVE CIRCUITS Semiconductors and Doping
Resistive–Capacitive Circuits and Diodes and Rectification
Time Constant Transistors and Amplification
Resistive–Inductive Circuits and Differential Amplifiers
Time Constant
SUMMARY
BASIC PRINCIPLES AND opposite charges attract. An electric charge can

DEFINITIONS IN ELECTRICITY be thought of as generating an electric field
in the surrounding space. The field around
a positive charge points radially outward in
Electric Charges and Force all directions from that charge, whereas the
field around a negative charge points radially
Electric charges are of two types, designated inward. Numerically, the electric force F on
positive and negative. Charges exert electric charge q in a region of space in which there is
forces on each other; like charges repel, an electric field E is given by F = qE. Thus, the
5
electric field can be thought of as the electric is supplied is called power (P = dU/dt); it is
force per unit charge (E = F/q). measured in watts. One watt is one joule per
Ordinary matter consists of atoms contain- second.
ing a nucleus composed of positively charged
protons and uncharged neutrons. Negatively Key Points
charged electrons occupy the space around the • Energy required to move a charge in an
nucleus, to which they normally are bound by
electric field is proportional to charge.
the attractive electric force between them and • Electric potential is energy per unit
the nucleus. In unionized atoms, the net charge
charge (measured in volts).
of the electrons is equal and opposite to the • Electrochemical cells and batteries con-
charge of the nucleus, so that the atom as a
vert chemical energy to electric energy.
whole is electrically neutral. The charge car- • Electric generators convert mechanical
ried by 6.24 × 1018 protons is one coulomb
energy to electric energy.
(the SI [Système International] unit of electric • Power is the rate at which energy is sup-
charge).
plied (measured in watts).
Key Points
• There are positive and negative electric
charges.
Electric Current and Resistors
• Electric force between two charges dep-
The movement of electric charges is called
ends on the charges and the distance
electric current. The current i is numerically
between them.
• Electric field is force per unit charge. equal to the rate of flow of charge q (i = dq/dt)
and is measured in amperes. One ampere is
one coulomb per second. Current in a circuit
is somewhat analogous to flow in fluid dynam-
Electric Potential ics. A conductor is a substance that has free
charges that can be induced to move when an
Energy (work) is required to move a charge in electric field is applied. For example, a salt
an electric field because of the electric force solution contains sodium and chloride ions.
acting on that charge. The energy required, U, When such a solution is immersed in an elec-
is proportional to the charge, q; thus, it makes tric field, the sodium ions move in the direction
sense to talk about the energy per unit charge. of the field, while the chloride ions move in
This quantity is called the electric potential the opposite direction. The direction of flow of
(V = U/q) and is measured in volts; one volt is current is determined by the movement of pos-
one joule of energy per coulomb of charge. The itive charges and, hence, is in the direction of
energy required to move a charge in a uniform the applied electric field. A metal contains free
electric field is also proportional to the distance electrons. When an electric field is applied, the
moved. It can be shown that the difference in electrons (being negatively charged) move in
electric potential between two points at a dis- the direction opposite to the electric field, but
tance l apart in a region of space containing the current by convention is still taken to be
an electric field E is given by V = El. Electric in the direction of the field (i.e., opposite to
potential in a circuit is somewhat analogous to the direction of charge movement). The cur-
pressure in fluid dynamics. rent flowing in a conductor, for example a wire,
To have continuous movement of charges, as divided by the cross-sectional area A of that
in an electric circuit, energy must be supplied conductor is called the current density (J), that
continuously by a device such as an electro- is, J = i/A.
chemical cell, or a battery of such cells (which Movement of charges in an ordinary con-
convert chemical energy to electric energy), or ductor is not completely free; there is friction,
an electric generator (which converts mechan- which is called resistance. Many conductors are
ical energy to electric energy). Such a device linear, that is, the electric field that causes cur-
is called a seat of electromotive force (EMF). rent flow is proportional to the current density
The EMF of a battery or generator is equal to in the conductor. The resistivity (ρ) of a sub-
the energy supplied per unit of charge and is stance determines how much current it will
measured in volts. The rate at which energy U conduct for a given applied electric field and
Electricity and Electronics for Clinical Neurophysiology 7
is numerically equal to the ratio of the elec- • Charge stored in a capacitor is propor-
tric field to the current density (ρ = E/J). The tional to potential difference.
resistivity is a constant for any given substance. • Capacitance is charge stored per unit
In contrast, the resistance R of an individ- potential (measured in Farads).
ual conductor (also called a resistor in this
context), which is equal to the ratio of the
potential difference (V) across the resistor to
the current flow i in the resistor (R = V/i,
Coils (Inductors)
called Ohm’s law), depends on the geome-
A coil (also known as an inductor or electro-
try of the resistor as well as on the material
magnet) is a device that generates a magnetic
of which it is made. Resistance is measured
field when a current flows in it. Physically, it
in volts per ampere, or ohms. The resistance
consists of a coil of wire that may be wrapped
R of a long cylindrical conductor of length l,
around a magnetic core; for example, a fer-
cross-sectional area A, and resistivity ρ is given
romagnetic substance such as iron, cobalt, or
by R = ρl/A.
nickel. A coil has a property, called inductance,
Sometimes it is more convenient to discuss
that is analogous to the mechanical property
the conductivity of a substance, σ = 1/ρ. This
of inertia; it resists any change in current flow
is the ratio of the current density to the elec-
(either increase or decrease). More precisely,
tric field (σ = J/E). Similarly, the conductance
a coil is capable of generating an EMF in
G of a resistor is the reciprocal of the resistance
response to any change in the current flowing
(G = 1/R). Ohm’s law in terms of conductance
in it; the direction of the EMF is always such
may be written as G = i/V.
as to oppose the current change, and numeri-
Key Points cally the potential difference V across a coil is
proportional to the rate of change of the cur-
• Current is the rate of flow of charge (mea- rent i in the coil: V = − L (di/dt), where L is
sured in amperes). the inductance of the coil and the minus sign
• A conductor has free charges that move indicates that the potential is in the direction
when an electric field is applied. opposite to the current change. Inductance
• The conductors in metal are negatively is measured in volt-seconds per ampere, or
charged electrons. Henrys (H).
• Current density is charge per unit cross-
sectional area. Key Points
• Linear conductors obey Ohm’s law (cur-
rent flow is proportional to potential • A coil (inductor) generates a magnetic
difference). field when current flows in it.
• Resistance is the ratio of potential differ- • A coil generates an EMF in response to a
ence to current flow (measured in ohms). change in current flow.
• The potential difference across a coil is
proportional to the rate of change of
Capacitors current.
• Inductance is the potential difference
A capacitor is a device for storing electric divided by rate of change of current (mea-
charge; it generally consists of two charged sured in Henrys).
conductors separated by a dielectric (insula-
tor). A capacitor has the property that the
charge q stored is proportional to the poten-
tial difference V across the capacitor: q = CV,
where C, the charge per unit potential, is called
CIRCUIT ANALYSIS
the capacitance. Capacitance is measured in
A circuit is a closed loop or series of
coulombs per volt, or Farads (F).
loops composed of circuit elements con-
Key Points nected by conducting wires. Circuit ele-
ments include a source of EMF (power
• A capacitor is two charged conductors sep- supply), resistors, capacitors, inductors, and
arated by an insulator. transistors.
Kirchhoff’s First Law i

a
For any loop of a circuit, the energy per unit
charge (potential) imparted to the loop must
equal the energy per unit charge dissipated +
(principle of conservation of energy). Stated ε B R i
–
another way, if electric potential is to have
any meaning, a given point can have only one i
value of potential at any given time. If we
start at any point in a circuit and, in imagina- Figure 1–1. A simple electric circuit containing an elec-
tion, go around the circuit in either direction, tromotive force, ε, current, i, and a resistor, R. The applica-
adding up algebraically the changes in poten- tion of Kirchhoff’s first law to the circuit starts at point “a.”
tial that we encounter, we must arrive at the B is a battery (seat of electromotive force) (From Halliday,
D., and R. Resnick. 1962. Physics. Part II, Rev. 2nd ed.,
same potential when we return to the starting 790. New York: John Wiley & Sons. By permission of the
point. In other words, the algebraic sum of the publisher.)
changes in potential encountered in a complete
traversal of a circuit loop must be zero. This is
Kirchhoff’s first law.1
Figure 1–1 shows a simple circuit containing a
Key Points source of EMF and a resistor, to which Kirch-
hoff’s first law may be applied to determine the
• Energy is conserved in an electric circuit. current flow, as follows:
• Therefore, the sum of all changes in
potential around an entire circuit is −iR + ε = 0
zero. ε = iR
i = ε/R
Rules for Seats of Electromotive Thus, the current is given by the EMF divided
Force, Resistors, Capacitors, and by the resistance. A similar analysis can be
Inductors made when a circuit contains multiple sources
of EMF, or multiple resistors, connected in
To apply Kirchhoff’s first law to a circuit, the series. The effective net EMF is the algebraic
following rules must be used to determine the sum of the individual EMFs. The effective
algebraic signs of the potentials across circuit net resistance is the sum of the individual
components: resistances.
1. If a seat of EMF ε is traversed in the

direction of the EMF (i.e., from the neg- Kirchhoff’s Second Law
ative to the positive terminal), the change
in potential is +ε; in the opposite direc- A node is a junction point of two or more
tion, it is −ε. conductors in a circuit. The node cannot act
2. If a resistor is traversed in the direction as a repository of electric charge; therefore,
of the current, the change in potential is the sum of all the current flowing into a node
−iR; in the opposite direction, it is +iR. must equal the sum of all the current flow-
3. If a capacitor is traversed in the direction ing out of the node (otherwise, charge would
of its positively charged plate to its nega- be constantly accumulating at the node). This
tively charged plate, the change in poten- is Kirchhoff’s second law. By using this law
tial is −q/C; in the opposite direction, it for each node in the circuit, together with the
is +q/C. first law for each loop in the circuit, any arbi-
4. If an inductor is traversed in the direction trary circuit problem, no matter how complex,
of the current flow, the change in poten- can be expressed as mathematical equations.
tial is −L(di/dt); in the opposite direction, Whether the equations can be solved and how
it is +L(di/dt). difficult it is to obtain a solution depend on the
i Key Points
a
• A node is a junction point in an electric
circuit.
• Charge is conserved in a circuit.
ε R1 i1 R2 i2 R3 i3 • Therefore, the sum of current flowing into
a node equals the sum of current flowing
out of a node.
b
i
Figure 1–2. A circuit containing an EMF, ε, and three
RESISTIVE–CAPACITIVE AND
resistors (R1 , R2 , R3 ) in parallel. a, Junction point (node); RESISTIVE–INDUCTIVE
b, junction point (node); i, current. (From Halliday, D., and CIRCUITS
R. Resnick. 1962. Physics. Part II. Rev., 2nd ed., 800. New
York: John Wiley & Sons. By permission of the publisher.)
Resistive–Capacitive Circuits and
nature of the circuit and the elements it con- Time Constant
tains and on the mathematical skills available
for the task. Figure 1–3 shows a resistive–capacitive (RC)
Figure 1–2 shows a circuit containing a sin- circuit containing a single source of EMF con-
gle source of EMF connected to three resistors nected to a resistor and capacitor in series.
in parallel. Kirchhoff’s second law is applied to When the switch is placed in position a, the
the junction of the three resistors, and Kirch- current flows in such a way as to charge the
hoff’s first law is applied to each branch of the capacitor. Because of the presence of the resis-
circuit, to give four independent equations in tor, the capacitor is not charged all at once
the four current variables i, i1 , i2 , and i3 , as but gradually over time. When Kirchhoff’s
follows: first law is applied to this circuit and use is
made of the fact that the current is charg-
i = i1 + i2 + i3
ing the capacitor at the rate i = dq/dt,
− i1 R1 + ε = 0 a differential equation results that can be
− i2 R2 + ε = 0 solved for the charge or current as a function
of time.
− i3 R3 + ε = 0 Figure 1–4A shows the exponential rise in
the charge q on the capacitor, and Figure 1–4B
These equations are solved for the four cur-
shows that the current i (which is the slope
rents as follows:
of the curve representing q against time) falls
ε = i1 R1 = i2 R2 = i3 R3 exponentially to zero as the capacitor becomes
i1 = ε/R1 fully charged. The time constant of the RC cir-
cuit is the time required for the current to fall
i2 = ε/R2
i3 = ε/R3
i
i = i1 + i2 + i3 a
S
= ε(1/R1 + 1/R2 + 1/R3 ) R
b
+
Equivalent resistance: ε C
–
1
R=
(1/R1 + 1/R2 + 1/R3 )
The net current i in the circuit can be calcu- Figure 1–3. A circuit containing an EMF, ε, a resistor, R,
lated from Ohm’s law, using an effective net and a capacitor, C. With the switch, S, in position a, the
capacitor is charged. In position b, it is discharged. (From
resistance given as the reciprocal of the sum Halliday, D., and R. Resnick. 1962. Physics. Part II. Rev.,
of the reciprocals of the three resistances in 2nd ed., 802. New York: John Wiley & Sons. By permission
parallel. of the publisher.)
A 12 • The time constant is RC.

Cε • The current decays exponentially to zero
with the same time constant.
8 • During discharge of the capacitor, the
q, microcoulombs
charge and current both decay exponen-

tially.
4
RC
0 Resistive–Inductive Circuits and
0 2 4 6 8 10
t, milliseconds
Time Constant
B 6
ε/R A circuit containing a resistor R and an induc-

tor L (an RL circuit) as well as a source of EMF
i, milliamperes
4
can be studied by similar methods. Because
of the inductance, the current flow in this cir-
cuit does not rise immediately to its eventual
2
value when a switch is closed but rather rises
exponentially with a time constant (time to
RC
0
reach 63% of final value) given by L/R. This
0 2 4 6 8 10 is caused by the “inertial” effect of the induc-
t, milliseconds tor. The form of the exponential rise in current
Figure 1–4. A, The variation of charge, q, with time, t, is similar to the shape of the exponential rise
during the charging process. B, The variation of current, in charge shown for the previous RC circuit in
i, with time, t. For R = 2000 , C = 1.0 μf, and ε = 10 V. Figure 1–4A. Similarly, when a switch bypasses
(From Halliday, D., and R. Resnick. 1962. Physics. Part II.
Rev., 2nd ed., 804. New York: John Wiley & Sons. By the source of EMF in this circuit, the current
permission of the publisher.) will not drop to zero immediately but will fall
off exponentially in time with time constant
to 1/e (37%) of its initial value; it is also the time L/R because of the effect of the inductor.
needed for the charge to rise to 1 − 1/e (63%)
of its final value; e is the base of the natural Key Points
logarithm (e = 2.718282). The time constant is
• In a resistive–inductive (RL) circuit, cur-
equal to RC.
When the switch shown in Figure 1–3 is rent flow gradually increases over time.
• The current approaches its maximum
placed in position b, the direction of cur-
rent flow is reversed and the capacitor is dis- value with an exponential time constant.
• The time constant is L/R.
charged. Application of Kirchhoff’s first law to
• When EMF is removed, the current in a
this circuit yields a differential equation that
can be solved for the charge or current as a RL circuit decays exponentially to zero.
function of time.
In this situation, both the charge on the
capacitor and the current (slope of the curve CIRCUITS CONTAINING
representing q against time) fall exponentially INDUCTORS AND CAPACITORS
with time with the same time constant, RC.
The current is negative in this case (i.e., it flows
counterclockwise in the circuit as the capacitor
Inductive–Capacitive Circuits
is discharged).
Figure 1–5 shows an ideal circuit containing
Key Points a capacitor and an inductor, an inductive–
capacitive, or LC, circuit. The circuit is ideal
• In a resistive–capacitive (RC) circuit, the because it contains no resistance. In stage
capacitor is charged gradually over time. (a), the capacitor is fully charged and there
• The charge on the capacitor approaches is no current flow. All the energy present in
its maximum value with an exponential the circuit is stored as electric energy (UE )
time constant. in the capacitor. By stage (b), a current flow
(c)
L L
UB UE
(b) (d )
UB UE UB UE
L L
(a) (e)
UB UE UB UE
L L
(h) (f )
UB UE UB UE
(g)
UB UE
Figure 1–5. A simple LC circuit showing eight stages (a–h) in one cycle of oscillation. The bar graphs under each stage
show the relative amounts of magnetic (UB ) and electric (UE ) energy stored in the circuit at any time. (From Halliday, D.,
and R. Resnick. 1962. Physics. Part II. Rev., 2nd ed., 944. New York: John Wiley & Sons. By permission of the publisher.)
has partially discharged the capacitor but at constant τ of this circuit is the square root of
the same time has created a magnetic field in LC, and the frequency of oscillations is 1/2π τ .
the inductor. The total energy is split between
electric energy in the capacitor and magnetic
energy in the inductor (UB ). At stage (c), the Key Points
current flow has reached its maximum and the • An ideal inductive–capacitive (LC) circuit
capacitor is fully discharged. The inductance has no resistance and demonstrates con-
effect, however, causes the current to continue tinuous oscillations.
to flow, now charging the capacitor in the oppo- • The frequency of oscillation depends
site direction, as shown in stages (d) and (e). In inversely on the square root of LC.
stages (f) through (h), the scenario is repeated,
with the capacitor discharging and the cur-
rent flowing in the opposite direction. Finally,
stage (a) is reached again, and the entire cycle Inductive–Resistive–Capacitive
repeats. Thus, this LC circuit is an oscillator, Circuits
and the charge on the capacitor is a cosine
function of time; similarly, the current in the A more realistic circuit is shown in Figure 1–6;
circuit is a sine function of time. The time it contains a resistor, capacitor, and inductor.
R • Oscillations in an LRC circuit are damped

unless power is continuously supplied by
an alternating current EMF.
• When a generator supplies power to an
L C
LRC circuit, the current flow is propor-
tional to the EMF of the generator.
• Impedance is the ratio of the amplitude of
ε(t) the applied EMF to the amplitude of the
current flow (measured in ohms).
Figure 1–6. An AC circuit containing a generator, • Impedance depends on L, R, and C and
ε(t), inductor, L, resistor, R, and capacitor, C. (From
Halliday, D. and R. Resnick. 1962. Physics. Part II. Rev., also is dependent on the frequency of the
2nd ed., 952. New York: John Wiley & Sons. By permission applied EMF of the generator.
of the publisher.)
It also contains a source of alternating cur- Root-Mean-Square Potentials or

rent EMF (an AC generator). This is a device
that generates an EMF ε(t) that varies sinu-
Currents
soidally over time (which is exactly what the
Often, when measuring potentials and currents
electric power company generators do) with a
in AC circuits that vary sinusoidally with time,
frequency of f (typically 60 Hz for line power in
it is more convenient to deal with an average
the United States).
potential or current than the amplitude (the
If the AC generator is omitted, the circuit
peak positive or negative value) over a cycle. A
behaves as a damped oscillator; if the capacitor
simple average is not useful, however, because
is initially charged up and then switched into
it is always zero; that is, the values are pos-
the circuit, the current in the circuit will vary
itive for half the cycle and negative for the
sinusoidally over time, but its amplitude decays
other half. The most useful “average” quan-
exponentially to zero because of power dissipa-
tity that can be used is the root-mean-square
tion in the resistor.
(rms) potential or current. This is defined as
If, on the other hand, the AC generator sup-
the square root of the average of the squares.
plies energy continuously at a fixed frequency f,
Because the square of a quantity, whether neg-
then the current flow in the circuit also varies
ative or positive, is always positive, the rms
sinusoidally with the frequency f of the driv-
value over a cycle is nonzero. It can be shown
ing EMF, although there generally is a phase
that the rms value of a sinusoidally varying
shift (effectively, a time delay) between the cur-
quantity is equal to the amplitude divided by
rent and the driving EMF. For such a circuit, it
the square root of two (approximately 0.707
can be shown that the amplitude of the current
times the amplitude). When it is said that the
flow i0 is directly proportional to the ampli-
line voltage for electric service in the United
tude of the driving EMF ε0 , that is, i0 = ε0 /Z.
States is 120 V, an rms value is implied; the
This is analogous to the situation in a direct
amplitude of the voltage variation is actually
current (DC) circuit containing a battery and
±170 V. Similarly, the rating of a fuse or circuit
a resistor, in which current is proportional to
breaker is an rms current rather than current
the EMF of the battery (i = ε/R, Ohm’s law).
amplitude.
The quantity in the AC circuit that is analo-
gous to resistance in a DC circuit is impedance
(Z); like resistance, it is measured in ohms. Key Points
Impedance is determined by all three circuit
• Root-mean-square values of potential and
elements (L, C, and R); also, unlike DC resis-
tance, impedance is a function of the frequency current are more commonly used than
f of the AC generator. amplitudes.
• Root-mean-square values are the square
root of the average of the squares over an
Key Points entire cycle.
• A realistic inductive–capacitive circuit has • An rms value is approximately 0.707 times
resistance. the amplitude.
Calculation of Reactance
Z
In general, impedance (Z) is made up of XL – Xc
three parts: the resistance (R), the reactance
of the capacitor (XC ), and the reactance of the
inductor (XL ). Reactance, which is measured
in ohms, is the opposition that a capacitor or
R
inductor offers to the flow of AC current; it is a
function of frequency. Figure 1–7. A right triangle symbolizing the relationship
The reactance of a capacitor is inversely pro- among resistance (R), inductive and capacitive reactance
(XL − XC ), and impedance (Z).
portional to the frequency and the capacitance
(XC = 1/2πfC). Thus, it is least at high frequen-
cies, becomes progressively greater at lower impedance, resistance, and reactance. Note
frequencies, and is infinite at zero frequency that the capacitive reactance XC is subtracted
(DC), because an ideal capacitor uses a perfect from the inductive reactance XL in calculating
insulator between the plates that is not capable impedance. This, together with the frequency
of carrying any direct current. (The only rea- dependence of the reactances described above,
son a capacitor appears to conduct AC current leads to an important phenomenon in AC cir-
is that AC current is constantly reversing direc- cuits: there is always one frequency at which
tion; the capacitor in this case is merely being the impedance of an LRC circuit is a min-
charged, discharged, and charged again in the imum. This frequency may be calculated by
opposite polarity [Fig. 1–5].) The reactance of setting XC = XL , because when this is true,
an inductor is directly proportional to the fre- impedance Z equals resistance R (the small-
quency and the inductance (XL = 2πfL). Thus, est possible value). It may be shown by this
it is zero at zero frequency (DC) and increases method that the frequency at which impedance
progressively with increasing frequency. This is a minimum is exactly equal to the frequency
happens because the effect of an inductor is of oscillations of the LC or LRC circuit without
to oppose changes in current, and the more an AC generator (the frequency of the circuit
rapidly the current changes, the greater the shown in Fig. 1–5). This can be restated as
induced EMF opposing that change will be. follows: when an LRC circuit is driven by an
AC source of EMF, the largest current flow
Key Points occurs when the frequency of the driving EMF
is exactly equal to the natural, or resonant,
• Reactance is the opposition that a capac-
frequency of the circuit. The current flow at
itor or inductor offers to the flow of AC driving frequencies above or below the reso-
current (measured in ohms). nant frequency is less; that is, the impedance
• Reactance of a capacitor is inversely pro-
of the circuit is greater. This phenomenon is
portional to capacitance and frequency. known as resonance; it is exploited in tuner
• Reactance of an inductor is directly pro-
circuits to select a signal of one particular
portional to inductance and frequency. frequency (i.e., one broadcast station) and to
reject signals of all other frequencies. Simi-
lar circuits can be used as narrow band-pass
Calculation of Impedance and the filters to eliminate all but a narrow range of
Phenomenon of Resonance frequencies from a signal or as notch filters to
eliminate a narrow range of frequencies from a
After the reactances and resistances have been signal.
calculated, the impedance is calculated as fol-
lows:
Key Points

Z = R2 + (XL − XC )2 • Impedance in an LRC circuit depends
on resistance, capacitive reactance, and
The right triangle shown in Figure 1–7 (with inductive reactance.
impedance Z being the hypotenuse) symbol- • A right triangle symbolizes in geomet-
izes in geometric form the relationship among ric form the calculation of impedance;
the difference in inductive and capacitive the place of the battery EMF. The output
reactance is on one side and the resistance potential Vout is proportional to the current
is on the other side of the triangle. i in the circuit. This decreases exponentially
• The impedance is the hypotenuse of the to zero with time constant (TC) equal to RC
triangle. when the input potential is “turned on” and the
• There is a particular frequency at which capacitor charges, and it becomes negative and
the impedance of an LRC circuit is a decreases exponentially when the input poten-
minimum. tial is “turned off” and the capacitor discharges.
This accounts for the shape of the output in
response to a square-wave calibration pulse.
Alternatively, one can imagine applying a
FILTER CIRCUITS sinusoidal AC potential to the input of this fil-
ter circuit. The current i is then equal to the
High-Pass Filters input potential divided by the impedance Z of
the circuit, which can be calculated from the
Figure 1–8A shows a simple high-pass (low- resistance and capacitive reactance. When the
frequency) filter circuit, which consists of a formula for impedance given earlier is used,
capacitor C in series with and a resistor R the ratio of the output to the input potential
in parallel with the output circuit. The input Vout /Vin can be calculated as a function of fre-
potential Vin is applied between the input ter- quency f. The calculation shows that the output
minal and ground, and the output potential Vout is strongly attenuated at low frequencies (when
is developed across the resistor R. This circuit f is near zero) but is essentially equal to the
may be analyzed in two ways. First, it may be input at high frequencies (when f is large). The
treated as an RC circuit similar to that shown in cut-off frequency f of the high-pass filter is
Figure 1–3, with the input potential Vin taking usually specified as the frequency at which the
Test signals
A
Component circuit
High pass = Low cut
Gain
R in 1.0
Celectrode 0.7
0.1 1
f=
2π RC
Effective circuit
0.01
C
1 10 100 1000 10,000
R Hz
B
Component circuit Low pass = High cut
1.0
0.7
Cstray
0.1 1
f=
2π RC
Effective circuit
R 0.01
1 10 100 1000 10,000
C Hz
Figure 1–8. A, A high-pass (low-frequency) filter circuit. B, A low-pass (high-frequency) filter circuit. C, Capacitor; R,
resistor; TC, time constant. The effects on square waves and on EMG signals are shown to the right.
A B
DC
DC
.1
.1
.3
.3
1
1
5 5
10 seconds 1 seconds
0.1-HZ input 1-HZ input
Filter step Output (....)

Input (—) × response (– –) = attenuated and
shifted to left
Figure 1–9. A high-pass filter shifts the latency of lower frequency waveforms more than higher frequency. Note the
greater effect on the 0.1-Hz waveform than on the 1.0-Hz waveform.
attenuation factor Vout /Vin is one divided by the of a resistor R in series with and a capacitor
square root of two, or 0.707; this occurs when C in parallel with the output circuit. This is
f = 1/2πRC. Equivalently, the time constant of also an RC circuit, but the output potential
the filter is given by 1/2πf, where f is the filter Vout in this case is developed across the capac-
cut-off frequency. itor and is proportional to the charge on the
High-pass filters can alter waveforms by capacitor. Comparison with Figure 1–4 shows
shifting the phase. The phase shift or time that the output potential increases exponen-
delay is more prominent with lower frequency tially to a maximum with TC equal to RC
inputs. (Fig. 1–9). when the input potential is “turned on” and the
capacitor charges, and it decreases exponen-
Key Points tially when the input potential is “turned off”
• A high-pass (low-frequency) filter circuit and the capacitor discharges. This accounts for
the shape of the output in response to a square-
is an RC circuit with the capacitor in series
wave calibration pulse. In practice, the time
and the resistor in parallel.
• The shape of a square-wave calibration constant of a high-frequency filter is discussed
less often than that of a low-frequency filter,
pulse after high-pass filtering is deter-
because it is much smaller, for example only
mined by the behavior of the RC circuit.
• By calculating the impedance of the filter 2 ms for a 70-Hz filter, and cannot be mea-
sured on electroencephalographic (EEG) trac-
circuit, the filter output in response to a
ings made at standard paper speeds by visual
sinusoidal input can be found.
• The output is attenuated at low frequen- inspection of the calibration pulse.
This filter circuit can also be analyzed in
cies but is nearly equal to the input at high
the context of a sinusoidal AC input potential.
frequencies.
• The cut-off frequency of the filter (at The output potential Vout is equal to the cur-
rent flow times the capacitive reactance XC (the
which the attenuation factor is 0.707) is
equivalent of Ohm’s law for a capacitor, with
1/2πRC.
reactance substituting for resistance in an AC
circuit), and the current i is equal to the input
Low-Pass Filters potential Vin divided by the total impedance Z,
which is the same as before. When the formula
Figure 1–8B shows a simple low-pass for impedance given earlier is used, the ratio
(high-frequency) filter circuit, which consists of the output to the input potential Vout /Vin can
be calculated as a function of frequency f. The such as arsenic provides extra “free” electrons
calculation shows that the output is attenuated that can conduct an electric current. Because
at high frequencies but becomes nearly equal these electrons carry negative charge, the semi-
to the input at low frequencies. The cut-off conductor that results is referred to as an
frequency f of the low-pass filter is usually n-type semiconductor. Alternatively, the base
specified as the frequency at which the attenu- material can be doped with a trivalent element
ation factor Vout /Vin is one divided by the square such as gallium. An absence of sufficient elec-
root of two, or 0.707; this occurs when f = trons to fill all of the orbitals is the result; the
1/2πRC, as for the high-pass filter. unfilled, or electron-deficient, areas are called
Note that the only essential difference holes, and they behave as positive charges that
between a high-pass filter and a low-pass fil- are free to move and, thus, conduct a current.
ter is the source of the output potential (to be The resulting semiconductor is referred to as a
fed to the amplifier); the high-pass filter devel- p-type semiconductor. (What actually happens
ops its output potential across the resistor R, is that electrons from a neighboring atom move
whereas the low-pass filter develops its output to fill in the hole, resulting in a hole moving
potential across the capacitor C. to a new position.) Thus, n-type semiconduc-
tors have electrons available for conducting
Key Points current, whereas p-type semiconductors have
• A low-pass (high-frequency) filter circuit holes (a potential space for electrons) available
is an RC circuit with the resistor in series for conducting current.
and the capacitor in parallel.
• The shape of a square-wave calibration Key Points
pulse after low-pass filtering is determined • Transistors are made of semiconductors
by the behavior of the RC circuit. such as silicon and germanium doped with
• By calculating the impedance of the filter trace quantities of other elements.
circuit, the filter output in response to a • N-type semiconductors have extra “free”
sinusoidal input can be found. electrons that act as negative charges to
• The output is attenuated at high frequen- conduct current.
cies but is nearly equal to the input at low • P-type semiconductors have an absence
frequencies. of sufficient electrons to fill all orbitals
• The cut-off frequency of the filter (at (“holes”) that act as positive charges.
which the attenuation factor is 0.707) is
1/2πRC.
Diodes and Rectification
TRANSISTORS AND AMPLIFIERS A useful electronic device can be made when
two or more dissimilar semiconductors are
Semiconductors and Doping adjacent. When an n-type semiconductor slab
is fused along one face with a p-type semicon-
Transistors are constructed of materials called ductor, electrons diffuse from the n region to
semiconductors, which have resistivities inter- the p region, filling some of the empty holes
mediate between those of good conductors of the p region, up to the point at which the
(such as metals) and insulators (most non- relative attraction of the holes for electrons is
metals). Silicon and germanium are the most exactly counterbalanced by the effect of the
frequently used substances. They are very poor electric field set up between the regions by the
conductors when in pure form, but when migration of electrons. This leaves the p region
doped with trace quantities of elements capa- with a net negative charge and the n region
ble of acting as electron donors or acceptors, with a net positive charge. If such a device is
they become semiconductors. The resistivity of connected in a circuit to a source of EMF with
the semiconductor can be altered by control- the positive potential applied at the p region,
ling the doping process. a process called forward biasing, the electric
Doping the tetravalent base material, sili- field across the junction is reduced and fur-
con or germanium, with a pentavalent element ther migration of electrons from n to p occurs,
A B names npn and pnp refer to the types of semi-

conductors and their order in the sandwich;
p n npn transistors, for example, have a positively
+ + doped material in the middle layer and nega-
V V tively doped materials in the outer layers. In
– –
n p an npn transistor, the primary movement of
electrons is from one of the outer layers (the
emitter), which is connected to the negative
Figure 1–10. A, A forward-biased pn semiconductor terminal of an external battery or power sup-
junction permits current flow. B, A reverse-biased junction
prevents current flow. V, Potential across the semicon- ply, to the other outer layer (the collector),
ductor junction. (From Misulis, K. E. 1989. Basic elec- which is connected to the positive terminal,
tronics for clinical neurophysiology. Journal of Clinical through the middle layer (the base). Under
Neurophysiology 6:41–74. By permission of the American resting conditions, little current flow occurs,
Clinical Neurophysiology Society.)
because although the emitter–base np junc-
tion is forward biased and can conduct current,
which constitutes a current flowing (by conven- the base–collector pn junction is reverse biased
tion) from the p to the n terminal (Fig. 1–10). and blocks current (Fig. 1–11). However, if a
If, however, the positive potential is applied at small positive potential is applied to the base
the n region, a process called reverse biasing, through an external connection, the junctional
the electric field across the junction is actually electric field at the base-collector pn junction
increased and current flow is blocked, because is reduced, because some electrons entering
in the region of the junction, the electrons the p-type base are allowed to leave through
that would carry current in the n region have the external connection, preventing the filling
been depleted and the “holes” that would carry of holes in the base. This allows a continu-
current in the p region have also been filled. ous movement of electrons from the emitter
This type of device is called a diode; it allows through the base into the collector. The mid-
current flow only in one direction. If a sinu- dle, or base, layer is very thin, so that even
soidal AC potential is applied to such a device, a small positive potential applied to the base
rectification results—the current flow is pul- is sufficient to facilitate a large conductance
satile but constrained to a single direction. This between collector and emitter. Thus, a small
is the first step that occurs in power supply cir- controlling voltage across base and emitter gov-
cuits that convert AC line voltage to DC volt- erns the flow of a much larger current through
ages suitable for use in most electronic circuits. the collector-to-emitter circuit. This, in effect,
is the basis of an electric amplifier.
Key Points
• Joining an n-type to a p-type semiconduc- (Collector)
tor creates a diode.
• A diode allows current flow only in one n
C
direction. B
• A diode “rectifies” an AC potential (the +
p
(Base) +
first step in converting AC to DC for use + E
– + –
in electronic circuits). –
n –
(Emitter)
Transistors and Amplification Figure 1–11. An npn junction bipolar transistor showing
the potential applied between the emitter, E, and base, B,
A transistor is a device that controls the that controls the flow of current between the emitter and
transfer of electric charge across a resistor. the collector, C. The figure on the right shows the circuit,
Junction bipolar transistors (the most common using the conventional symbol for the transistor (From
Misulis, K. E. 1989. Basic electronics for clinical neuro-
type) can be made in two forms called npn and physiology. Journal of Clinical Neurophysiology 6:41–74.
pnp. Both are composed of a three-layer sand- By permission of the American Clinical Neurophysiology
wich of semiconductors of different types. The Society.)
Because the gain, or amplification, of a A differential amplifier is actually composed

single transistor is limited, multiple stages of two amplifier circuits (two transistors), one
of amplification are used in an EEG or for the G1 input and one for the G2 input (the
electromyographic (EMG) machine. For exam- nomenclature G1 and G2 is still in common
ple, six stages of amplification, each with a gain use, even though it originated in the early days
of 10, give an overall amplification factor of of EEG and EMG when vacuum tube ampli-
1,000,000. A complete EEG or EMG amplifier fiers were used; G was used to indicate a grid
contains preamplification stages, which typi- in the vacuum tubes). The G2 transistor is
cally boost the signal from the patient by a fac- connected to a negative power supply voltage,
tor of 10 to 1000, followed by low-frequency, and the G1 transistor is connected to a posi-
high-frequency, and 60-Hz notch filter cir- tive power supply voltage. The outputs of both
cuits, followed by driver amplification stages transistors are connected together (Fig. 1–12).
that provide the remaining amplification and In this fashion, increased current flow in the
produce an output signal capable of driving emitter-to-collector circuit of the G1 transis-
the oscilloscope display or the pen motors of tor produces a positive change in the output
the paper display unit or the analog-to-digital potential, whereas increased current flow in
converter in a digital instrument. the emitter-to-collector circuit of the G2 tran-
sistor produces a negative change in the output
potential.
Key Points Although an ideal differential amplifier
• A transistor is a solid-state device that con- would be sensitive only to the difference in
trols the transfer of electric charge across potential between the two inputs, in prac-
a resistor. tice, a large enough signal applied to both
• Junction bipolar transistors are a three- inputs simultaneously, called common mode,
layer sandwich of semiconductors of dif- produces a small output signal (Fig. 1–13).
ferent types (npn or pnp). This occurs because the input impedance and
• The outer layers are the emitter and the the gain and frequency response of the two
collector, while the thin middle layer is the transistors in the differential amplifier are not
base.
• A small controlling voltage applied to the
base governs the flow of a much larger V1
current from collector to emitter. 10x
• This is the basis of an electronic amplifier.
• Multiple stages of amplification are used Output
to achieve an overall gain of many orders [(V 1–V 2)×10]
V2 –10x
of magnitude.
Ref Ref
Differential Amplifiers V1 V2 (V1 – V2)×10

The type of amplifier used predominantly
in clinical neurophysiology is the differential
amplifier. This type of amplifier is constructed
to amplify only the difference in the poten-
tial between its two inputs. This is one way
of reducing contamination of the physiologic Figure 1–12. A differential amplifier constructed from
two single-ended amplifiers, with input potentials V1 and
signal by electrical noise—for example, 60-Hz V2 with respect to the reference (Ref), the second of which
noise from line voltage devices—because this produces an inverted output; the net output is given by
noise tends to be the same at all electrode the product of the amplifier gain (10 in this case) and
positions and cancels out when a difference in the difference in potential between the two inputs (From
Misulis, K. E. 1989. Basic electronics for clinical neuro-
potential is formed. Physiologic signals, on the physiology. Journal of Clinical Neurophysiology 6:41–74.
other hand, are usually different at different By permission of the American Clinical Neurophysiology
electrode positions. Society.)
+1
V + +1
–1 V
x10
+1
– –1
V
–1
Enoise = ±1 V Enoise = ±0.1 V
Esignal = ±0.1 V Esignal = ±1.0 V
SNR = 0.1 SNR = 10
CMRR = 100 = 40 dB
Figure 1–13. A differential amplifier with a small signal on one input and a large artifactual signal on both inputs. The
small signal is amplified while the large signal is reduced, but the common mode rejection is not sufficient to totally
eliminate the large signal.
quite identical. The common mode rejection input effectively becomes the ground electrode
ratio (CMRR) of a differential amplifier can potential. In addition to introducing more
be calculated as the applied common input 60 Hz and other noise into the recorded signal,
potential divided by the output potential.2 For artifacts and mislocalization of cerebral electric
modern amplifiers, this ratio is approximately activity can result by the unexpected introduc-
10,000. However, if the electrode impedances tion of a signal coming from an EEG ground
are high or differ significantly between the G1 electrode into one or more channels.
and the G2 inputs, the effective signal per-
ceived by the two transistors in the differen- Key Points
tial amplifier can differ significantly and the
CMRR can be drastically reduced, thus allow- • Predominantly differential amplifiers are
ing significant amounts of noise to contaminate used in clinical neurophysiology.
the signals being recorded. • They amplify only the difference in poten-
Even though the differential amplifier out- tial between two inputs, reducing contam-
put depends on the difference in the potentials ination by electrical noise.
at its two inputs, each input potential as per- • A differential amplifier has two transistor
ceived by the amplifier is relative to a common amplifier circuits connected to opposite-
reference, or ground, potential. This ground polarity power supply voltages.
potential is in practice equal to the poten- • Slight differences in the two transistors
tial at a single ground electrode that must cause some output when a large signal
be attached to the patient. For example, the reaches both inputs simultaneously.
ground electrode for EEG recording is typi- • CMRR is the ratio of the common input
cally placed on the mastoid process; occasion- potential to the output potential.
ally other locations, such as the frontal area, are • If one input of a differential amplifier is
used. As long as the electrode-to-patient con- of very high impedance, the ground elec-
nections are adequate (i.e., have low enough trode becomes the input.
impedance), the location of the ground elec-
trode does not matter. Because each input
of the differential amplifier receives a poten- SUMMARY
tial that is relative to the same ground and
these potentials are subtracted in the output, This chapter reviews the basic principles of
the potential of the ground electrode cancels electric and electronic circuits that are impor-
out. However, if there is a very poor (i.e., tant to clinical neurophysiology. Knowledge
one with high impedance) electrode connec- of these basic principles and how to solve
tion or, in the extreme case, if an electrode simple circuit problems is necessary for a com-
is left unconnected, the differential amplifier plete understanding of the proper operation
of equipment used in clinical neurophysiology Filter characteristics

and of the terminology and specifications given • High-pass filter—attenuates signals at low
in equipment manuals.
frequency while leaving high-frequency
signals unchanged.
• Low-pass filter—attenuates signals at high
Circuit definitions frequency while leaving low-frequency
• Resistance—opposition to direct current signals unchanged.
• Cut-off filter frequency—frequency at
flow; ratio of potential (voltage) to current.
• Conductance—ability to conduct direct which there is a 30% attenuation in signal
current; ratio of current to potential amplitude.
• Filter phase shift—Time delay (latency
(voltage).
• Capacitance—ability to store charge; ratio change) caused by a filter, especially
of charge to potential (voltage). affecting low-frequency components of a
• Inductance—ability due to magnetic effects signal.
• Filter time constant—resistance times
to produce a potential when current
changes; ratio of potential (voltage) to rate capacitance, inversely related to the cut-
of current change. off filter frequency.
• Reactance—opposition to alternating cur-
rent flow by capacitance or inductance;
ratio of potential (voltage) to AC current REFERENCES
in a capacitor or inductor.
• Impedance—net opposition to current 1. Halliday, D., and R. Resnick. 1962. Physics. Part II.
Rev., 2nd ed. New York: John Wiley & Sons.
flow by both resistance and reactance; 2. Misulis, K. E. 1989. Basic electronics for clinical
ratio of potential (voltage) to current in an neurophysiology. Journal of Clinical Neurophysiology
AC circuit. 6:41–74.
Chapter 2
Electric Safety in the Laboratory

and Hospital
INTRODUCTION ELECTRIC SAFETY PRINCIPLES

ELECTRIC POWER AND IMPLEMENTATION
DISTRIBUTION SYSTEMS Equipment Grounding
ELECTRIC SHOCK Tests for Equipment Grounding and
Requirements for Electric Current Leakage Current
to Flow Through the Body Rules for Electric Safety
Physiologic Effect of Electric Electric Safety Procedures for
Current Technicians
Factors Reducing Risk of Electric
ELECTRIC STIMULATION
Shock
SAFETY
Factors Increasing Risk of Electric
Stimulating Near Pacemakers and
Shock in Hospitals
Other Implanted Electrical Devices
Transcranial Electrical
LEAKAGE CURRENT and Magnetic
Origin Stimulation
Methods by Which Leakage Therapeutic Cortical and Deep
Current Reaches Patients Brain Stimulation
Methods to Reduce Leakage
Current Reaching Patients SUMMARY
INTRODUCTION tissue through which current passes. This chap-

ter discusses the concepts of electrical safety in
The principles of electrical safety are of great clinical neurophysiology.
importance in clinical neurophysiology. All of
the electrophysiologic studies that are per-
formed require the application of electrical ELECTRIC POWER
connections to equipments that, through con- DISTRIBUTION SYSTEMS
nections with the patient, pass small amounts
of electrical current to the patient. Although The electric systems of buildings are designed
small, there is always an inherent risk to the to distribute electric energy from one central
21
point of entrance to all the electric applian- • A “neutral” wire acts as the return path
ces and receptacles. Power companies provide and is connected to earth ground.
electric energy at high voltage (typically 4800 • A “hot” wire carries 120 or 240 V electric-
or 4160 V for a hospital or medical clinic) to ity to lights and appliances.
minimize transmission losses. A step-down tra- • An equipment ground wire provides a sep-
nsformer converts the high-voltage energy to arate pathway to ground from the “neu-
safer, usable voltages (usually 120 and 240 V). tral” wire.
Figure 2–1 shows the wiring of a typical 120 V
circuit. The secondary coil of the transformer
has a center tap that acts as the return path ELECTRIC SHOCK
(“neutral”) for the circuit; it is connected to
earth ground through a grounding stake at the Electric shock is the consequence of the flow of
electrical junction site. Each of the two outer current through the body. The effect of electric
ends of the 240 V secondary coil can be used to shock depends both on the magnitude of the
drive one 120 V circuit; this provides a hot line current flow and the path taken by the current
whose potential is 120 V from ground. This hot in the body, which is determined by the points
line incorporates a circuit breaker that limits of entry and exit.
current flow to a level (e.g., 20 A) that will not
cause excessive heating in wiring in the build-
ing. For reasons explained later, each recepta-
cle also includes a ground contact connected
Requirements for Electric Current
to earth ground through a conductor separate to Flow Through the Body
from the “neutral” conductor (see Fig. 2–1).
When an externally generated current flows
Key Points through the body, it has a point of entrance
and a point of exit. The current may be thought
• Electric power for a building is distributed of as originating from some electric apparatus,
from a step-down transformer to wall out- or source, flowing through a conducting mate-
lets and lighting. rial from the apparatus to the body, flowing
Figure 2–1. Scheme of a building’s electric power distribution system showing a step-down transformer, circuit breaker,
grounding stake, and a third wire for equipment grounding in the conduit. (From Cromwell, L., F. J. Weibell, and
E. A. Pfeiffer.1980. Biomedical Instrumentation and Measurements, 2nd ed., 437. Englewood Cliffs, NJ: Prentice-Hall.
By permission of the publisher.)
Electric Safety in the Laboratory and Hospital 23
through the body, and finally flowing through through direct depolarization of muscle fibers,
a second conducting material from the body to a victim of electric shock may not be able to
ground. Thus, to have an electric shock, there let go of the source of the current when it
must be at least two connections to the body: exceeds about 10–20 mA. The threshold for
one to the current source and the other to induction of ventricular fibrillation is approxi-
ground. An apparatus can act as a source of cur- mately 100 mA. The externally applied current
rent either (1) because a point of connection spreads out as it passes through the body, so
between it and the body, such as an exposed that the fraction passing through the heart
metal part of the chassis or other metal contacts is small, less than 0.1% in most situations
or terminals, is in direct continuity with the hot depending on entry and exit points.
line through a very low resistance path caused In hospitals, one of the two required con-
by some fault such as a mechanical break in tacts between an external source or ground and
insulation or fluid spilled into the circuit or the body may be an intracardiac catheter. If
inadvertent direct connection of electrode lead current enters or leaves through this device,
wires to energized, detachable power-line-cord essentially the entire current flows through
plugs1 or, more commonly, (2) because of a the myocardium. In this case, the threshold
low-level leakage of current through a moder- for inducing ventricular fibrillation is far less
ate resistance path, which may be inherent in than for externally applied current. In humans,
the design of the apparatus. A further require- this threshold is estimated to be approxi-
ment for significant electric shock is that the mately 50 μA, but experiments in dogs have
entire pathway to, through, and out of the shown that as little as 20 μA is sufficient.2
body must have a sufficiently low resistance for Furthermore, the results of a recent study
normal line voltage to be hazardous. have suggested that the threshold for induc-
An additional requirement for a lethal elec- tion of cardiovascular collapse (which is less
tric shock is that the current must take a path than the threshold for inducing ventricular fib-
through the body that includes the heart (e.g., rillation) is the more relevant quantity, and this
when current enters through one arm and exits threshold is only 20 μA.3 The threshold may be
through the other), because the mechanism of significantly lower in persons with preexisting
lethal shock is almost invariably the induction heart disease.
of ventricular fibrillation.
Key Points
Key Points • Electric currents below 0.5–1 mA cannot
• Current flow through the body requires a be perceived.
• Currents over 20 mA lead to involuntary
point of entry (current source) and a point
of exit (any ground connection). muscle contraction, making it impossible
• The current source is usually leakage from to let go of the source of current.
• Externally applied currents over 100 mA
internal circuitry to the chassis or other
metal contacts or terminals. may cause ventricular fibrillation.
• For an electric shock to be lethal, the cur- • Currents as small as 20–50 μA entering
rent must take a path through the body the heart through an intracardiac catheter
that includes the heart. may cause ventricular fibrillation.
Factors Reducing Risk of Electric

Physiologic Effect of Electric Shock
Current
The risk of electric shock or electrocution
For currents that enter and leave the body from appliances is reduced by several factors,
through the skin, the usual situation outside of including the following:
hospitals, Figure 2–2 shows the approximate
amounts of current associated with various 1. Leakage currents that are available from
physiologic effects, ranging from minimal per- most electric appliances are relatively
ception (0.5–1 mA) to severe burns and physi- small.
cal injury (6–10 A). Because current flow in a 2. People using appliances are often not
limb leads to involuntary muscle contraction connected to ground.
Figure 2–2. Effects of 60 Hz AC electric current flow of various magnitudes produced by a 1-second external contact
with the body. (From Cromwell, L., F. J. Weibell, and E. A. Pfeiffer.1980. Biomedical instrumentation and measurements,
2nd ed., 434. Englewood Cliffs, NJ: Prentice-Hall. By permission of the publisher.)
3. Contacts with the source of leakage significantly greater because of the following
current and with ground usually have factors:
high resistance, for example, dry, intact
skin. 1. Leakage currents that may be avail-
4. Healthy, alert people can withdraw from able from appliances are relatively large
a source of current in most cases. because patients may be attached to
5. The hearts of healthy people require sig- many instruments (thus providing mul-
nificant electric currents to induce ven- tiple current sources), conducting flu-
tricular fibrillation. ids may get into instruments through
spillage or leakage, and instruments may
be used by many persons or used in
Factors Increasing Risk of Electric many locations (or both), thus increasing
Shock in Hospitals the chance of fault caused by misuse or
wear. In the operating room, instruments
The risk of electric shock or electrocution such as electrosurgical units may present
from appliances in hospitalized patients is special risks to the patient if proper
precautions for electric safety are not varies by the class of the instrument and local
followed.4, 5 governmental regulations. Leakage current in
2. Through attached electric instruments, an electric apparatus may originate in several
patients are often grounded or they ways, including the following:
may easily contact grounded objects, for
example, metal parts of beds, lamps, 1. There is always a finite internal cir-
and instrument cases. cuit resistance between the power line
3. Contacts with the source of leakage cur- (hot wire) and the instrument chassis,
rent and with ground are often of low known as instrument ground; this may
resistance because connections to mon- be decreased by faults in the wiring or
itoring devices purposely minimize skin by breakdown of insulation. A resistance
resistance (e.g., electrodes applied with as large as 5 M still allows 24 μA to
conducting paste) or bypass it altogether flow between the “hot” conductor and
(e.g., indwelling catheters). Furthermore, ground, which may be enough to induce
patients with conductive intracardiac ventricular fibrillation in an “electrically
catheters, such as pacemaker leads and susceptible” patient.
saline-filled catheters, have a direct low- 2. The capacitance between the “hot” con-
resistance pathway to the heart. Because ductor and the chassis resulting from
only tiny currents flowing in such a path internal circuitry or external cabling may
may induce lethal ventricular fibrillation, provide a relatively low-impedance path-
such patients are called “electrically sus- way for alternating current. A capacitance
ceptible.” as small as 440 picofarad (pF) still allows
4. Weakened or comatose patients cannot 20 μA to flow between the “hot” conduc-
withdraw from a source of current. tor and ground.
5. Patients’ hearts may be more suscepti- 3. The inductive coupling between power-
ble (through disease) to electric current– line circuits and other circuit loops, such
induced ventricular fibrillation. as ground loops when there are multi-
ple ground connections to the patient,
Key Points can induce ground-path current flow
• The risk of electric shock from appliances as well. In addition to the leakage
is reduced by factors such as low leak- currents available from equipment-to-
age current, high resistance of contact patient ground connections, leakage cur-
with leakage current source, and ability to rents may be introduced by similar mech-
withdraw from source of current. anisms into other leads or connections to
• Factors increasing the risk of electric the patient.
shock include patient attachment to mul-
tiple instruments, contact with grounded
objects, low-resistance contacts such as Methods by Which Leakage
indwelling catheters, and patients with
cardiac disease or those unable to with- Current Reaches Patients
draw from current source.
Leakage currents may reach patients when
contact is made either directly or through
another person to exposed metal parts or to
LEAKAGE CURRENT the chassis of electric equipment. Leakage
currents may also reach patients through a
Origin direct connection of the chassis (equipment
ground) to the patient; in the past such a
One of the most important parameters in elec- direct patient ground connection was made
trical safety is leakage current. This is an easy to reduce noise in recording of physiologic
parameter to measure on biomedical instru- signals. Finally, leakage currents may reach
ments but is often misunderstood because patients through resistive or capacitive (or pos-
there are multiple kinds of leakage current and sibly inductive) coupling to leads other than the
the maximal value allowed for leakage current patient ground.
Methods to Reduce Leakage

Current Reaching Patients
Many methods are used in modern hospital
electric distribution systems and in biomedi-
cal instruments to decrease the risk of elec-
tric shock by reducing the available leakage
currents, including the following:
1. The chassis and all exposed metal parts of

electric appliances are grounded through
a separate ground wire, through the Figure 2–4. Excessive leakage current from a floor pol-
isher flows through the building wiring to an EEG machine
round pin of electric plugs. Any leak- and an ECG monitor attached to the patient through dif-
age currents that would otherwise flow ferent wall plugs. This ground loop was created by the use
to a subject in contact with the chas- of two outlets physically distant from one other, producing
sis are instead shunted through this low- a difference in the equipment ground potentials between
the two outlets. (From Seaba, P. 1980. Electrical safety.
resistance pathway to ground. Because American Journal of EEG Technology 20:1–13. By per-
the leakage currents in properly func- mission of the American Society of Electroneurodiagnostic
tioning equipment are small, the ground Technologies.)
wire in the building’s power distribution
system usually carries very little current,
unlike the neutral wire, which carries the points in one room should be connected
full operating current. Hence, the poten- to a single ground wire, an equipotential
tial drop between the equipment chassis grounding system. Also, all biomedical
and the earth ground connection located equipment connected to a patient should
at the electric distribution panel of the draw power from the same group of out-
building is minimal, and the equipment lets to avoid large ground loops (Fig. 2–4).
ground potential remains very close to the 3. When necessary, isolation transformers
earth ground potential (Fig. 2–3). may be used to eliminate the neutral-
2. Hospital rooms that have exposed metal to-ground connection entirely, thereby
parts, for example, window frames, bath- reducing the risk of shock when a patient
room plumbing, shelving, and door frames, connected to a biomedical instrument
may also connect these to earth ground comes in contact with an earth ground
through the same grounding system used (Fig. 2–5). However, isolation transform-
for electric outlets. All such grounded ers do not eliminate entirely the risk of
shock, because they may have significant
leakage currents.
4. Appliances can be constructed with non-
metallic cases to minimize the chances
of patients contacting the equipment
chassis.
5. Appliances should have short line cords,
and the use of extension cords should
be avoided to minimize capacitive and
resistive leakage currents between the hot
and the ground wires. Note that each
foot length of cord unavoidably intro-
duces about 1 μA of leakage current into
the ground connection.
Figure 2–3. The power cord ground wire conducts leak- 6. Direct connections of patients to earth
age current from an electric apparatus chassis to earth ground should be avoided. In particu-
ground. (From Seaba, P. 1980. Electrical safety. Amer-
ican Journal of EEG Technology 20:1–13. By permis-
lar, inadvertent electric paths between
sion of the American Society of Electroneurodiagnostic ground and patients that bypass the
Technologies.) normally high skin resistance (especially
Figure 2–5. An isolation transformer used to reduce equipment leakage current. The equipment is connected to the
secondary coil of the transformer, which is electrically isolated from the power line hot and neutral conductors. (From
Cromwell, L., F. J. Weibell, E. A. Pfeiffer, and L. B. Usselman. 1973. Biomedical instrumentation and measurements, 387.
Englewood Cliffs, NJ: Prentice-Hall. By permission of the publisher.)
paths provided by intracardiac catheters) • Reduce leakage currents to a patient by

should be avoided by using nonconduc- ◦ Battery-powered equipment
tive materials. ◦ Avoid connection of patient to ground
7. Current that may flow through all con- ◦ Short line cords (no extension cords)
nections between a patient and the equip- ◦ Ground all equipment metal chassis
ment, both signal and ground, should be through a separate ground wire
limited to no more than 50 μA (10 μA ◦ Connect all equipment to same group of
for “electrically susceptible” patients, i.e., outlets
Type CF) through the use of current ◦ Equipotential grounding system
limiters or inductive or optical cou- ◦ Limit current flow to patient to less than
pling devices. Alternatively, when practi- 10 μA
cal, battery-powered equipment that has ◦ Isolation transformer.
no direct connection to line voltage or to
ground can be used.
ELECTRIC SAFETY PRINCIPLES
Key Points AND IMPLEMENTATION
• Leakage currents reach a patient through
◦ Contact with equipment metal chassis Equipment Grounding
◦ Connection of equipment ground to
patient Proper grounding of electric equipment (i.e.,
◦ Resistive or capacitive coupling to providing a low-resistance pathway from the
patient leads. equipment chassis to an earth ground) is
usually accomplished through the grounding that damage through use and misuse does not
wire in the line cord that connects to the compromise safety, educating all those who
round pin in the plug and thence to the use the equipment (especially technicians) in
building’s electric grounding system. Failure electric safety principles, and ensuring that
of this ground connection may occur in sev- certain basic minimal safety tests are per-
eral ways. There can be a failure of attach- formed each time a biomedical apparatus is
ment of the grounding wire in the line cord plugged in, turned on, and connected to a
to the equipment chassis, a break in continu- patient.
ity of the grounding wire within the cord, or Tests that should be performed on building
a failure of connection of the grounding wire wiring at the time of installation include the
to the grounding pin. Also, the grounding pin following:
may make poor contact with the wall recepta-
cle because of a reduction in contact tension 1. Visually inspect the wiring of all wall
caused by mechanical wear. The grounding pin receptacles to ensure that it is correct.
can also be deliberately bypassed using a so- 2. Measure the resistance between each wall
called cheater (3-prong to 2-prong) adapter. receptacle ground (and other grounded
Defects also can occur in building wiring, objects in the room) and a ground
such as an improper or omitted connection of known to be adequate, such as a cold
the wall receptacle’s grounding terminal to a water pipe or an independent ground-
ground wire or an interruption of the ground ing bus. This resistance should be less
connection somewhere in the building’s wiring. than 0.1 .6
This is particularly likely if metal conduit, 3. Measure the contact tension provided by
rather than a wire, provides the ground con- the wall receptacle, that is, the force
nection since conduit is subject to corrosion required to withdraw the ground pin of a
and loss of mechanical contact. Particularly in test plug from the receptacle. This should
newly constructed or remodeled rooms and be at least 10 ounces.
buildings, it is advisable to visually inspect and 4. Test all outlets when first installed and
to electrically test the ground connection in periodically thereafter with an approved
all wall receptacles. Because the ground con- electrical tester that measures polarity
nection is only for electric safety purposes, the and ground resistance.
lack of it in no way affects operation of the
electric equipment and, therefore, will remain These tests, except the first, should also
undetected if not specifically checked. be performed periodically, for example, every
6–12 months, and the receptacles whose con-
Key Points tact tension has degraded below 10 ounces
should be replaced. Tests that should be per-
• Grounding of electric equipment is accom-
formed on each biomedical instrument at the
plished by the grounding wire in the line time of purchase and periodically thereafter,
cord and in the building wiring. for example, every 6–12 months, include the
• Failures of grounding can occur at any
following:
step along the path from the equipment
chassis to the earth ground. 1. Visually inspect the line cord and plug for
signs of damage, wear, or breakage.
2. Measure the resistance between the
Tests for Equipment Grounding ground pin of the plug and the instrument
and Leakage Current chassis. This should be less than 0.1 .
3. Measure the chassis-to-earth ground
Each hospital, laboratory, or clinic should (enclosure) leakage current using certi-
establish an electric safety program that fied leakage meter. This measurement
includes selecting equipment that meets should be made with the equipment’s
appropriate safety standards, testing new grounding pin disconnected (to ensure
equipment after purchase to verify that safety even if the building grounding
standards are met, inspecting and retesting system is faulty) and under four sep-
equipment periodically thereafter to ensure arate conditions. This should include
both normal and reverse polarity of the Rules for Electric Safety
hot and neutral wires (to ensure safety
even if the wall receptacle is erro- In addition to a program of periodic testing
neously wired with opposite polarity) and inspection, electric safety requires that all
and with the equipment power switch persons using electric equipment in the labora-
“on” and “off.” In all four conditions, tory or hospital are familiar with the following
this current should not exceed 100 μA rules:
for normal operation (Type B, BF, CF
applied parts) and 500 μA for single fault
conditions.6 1. Do not ever directly ground patients
4. Measure the leakage current from each or allow patients to come into contact
terminal that connects to a patient, with grounded objects while connected
including the patient ground to earth to a biomedical instrument. If an
ground, under the same four conditions instrument does not meet UL 60601-
(patient leakage from applied part to 1, EN/IEC 60601-1 Medical Electrical
earth ground). This is the maximal leak- Equipment—Part 1 General Require-
age current that the equipment can sup- ments for Safety and Essential Perform-
ply to a patient who is grounded through ance, then patient–ground connections
a second connection. For use with “elec- should be made to only one instrument
trically susceptible” patients, this cur- at a time.7
rent should not exceed 100 μA (Type B, 2. Ensure that every electric device or
BF applied parts) and 10 μA (Type CF appliance, for example, lamps, electric
applied parts) for normal operation and beds, electric shavers, and radios, that
500 μA (Type B, BF applied parts) and a patient might accidentally come in
50 μA (Type CF applied parts) for single contact with is connected to an ade-
fault conditions.6 quate earth ground, such as through use
5. Measure the leakage current from the of an approved three-prong or double-
power-line hot wire to each terminal insulated grounded plug.
that connects to a patient, including the 3. Use only safe, properly designed, and
patient ground, under the same four con- pretested electric equipment. All bio-
ditions (patient leakage via F-type applied medical devices directly connected to
part caused by external voltage on the patients must have isolation or current-
applied part). This is the maximal current limiting circuits if they are to be used
that can be absorbed by the equipment with “electrically susceptible” patients.
from a patient who accidentally comes in All line-powered equipment should have
contact with a 120 V power line. For use three-prong grounded plugs. In general,
with “electrically susceptible” patients, patients should not be allowed to bring
this current should not exceed 100 μA their own electric appliances from home
(Type B, BF applied parts) and 10 μA for use in a hospital room.
(Type CF applied parts) for normal oper- 4. Ensure that all electric equipment in use
ation and 500 μA (Type B, BF applied has had a safety inspection done recently
parts) and 50 μA (Type CF applied parts) (within 6–12 months), as indicated by a
for single fault conditions.6 dated electrical safety inspection tag or
sticker.
5. Connect all patient-connected equip-
Key Points
ment to outlets in the same area or
• Each hospital, laboratory, or clinic should cluster to avoid large ground loops.
establish an electric safety program that 6. Never use an extension cord on patient-
includes selecting equipment that meets connected equipment because this adds
appropriate safety standards and testing leakage current through its internal
new equipment after purchase to verify capacitance and resistance and, thus,
that standards are met. provides another chance for ground
• These tests should also be performed peri- connection failure. In the operating
odically, for example, every 6–12 months. room or situations with direct cardiac
connection, equipment should be tested that should be carried with all portable
before every use. equipment.
7. Cover all electric connections to intrac- 5. Turn on the instrument and calibrate it
ardiac catheters with insulation to before connecting it to the patient. Major
eliminate electric continuity between electric problems may show up during
external devices or ground and the calibration; furthermore, electric surges
catheter whenever possible. occur as the instrument is turned on and
8. Have a defibrillator available at all build- leakage currents may be higher while it is
ing locations where patients have cardiac starting up.
catheters in place. 6. Disconnect the patient from the instru-
9. Do not ignore the occurrence of any ment before turning it off or on.
electric shocks, however minor; investi-
gate their causes. Thoroughly test any
equipment that may have been involved ELECTRIC STIMULATION SAFETY
before putting it back into service. Also,
do not ignore any abnormal 60 Hz inter- In addition to the issues of electrical safety
ference or artifact in an electrophysio- of all biomedical equipment discussed above
logic recording; this finding may indicate that relate to the electrical supply voltage
that some device is leaking current into and leakage currents, clinical neurophysiology
the patient. studies such as evoked potentials, nerve con-
10. Follow certain safety procedures, includ- duction studies, and transcranial electrical and
ing routine safety checks, each time an magnetic stimulation studies, as well as ther-
electric device is to be connected to a apeutic devices such as nerve, spinal cord,
patient. cortical or deep brain stimulators, involve stim-
ulating neural tissue with electrical currents
(or strong magnetic fields), which introduces
Electric Safety Procedures for additional safety considerations related to tis-
Technicians sue damage from stimulation and effects on
nearby implanted electrical devices such as
The following procedures should be followed pacemakers.
by technicians while performing an electro-
physiologic test requiring line-powered equip-
ment on a patient, especially portable studies Stimulating Near Pacemakers and
performed in a patient’s room: Other Implanted Electrical
Devices
1. Check the physical condition of the
equipment. Is there any evidence of liq- Peripheral nerve stimulation has been in use
uid spills, cord wear, or damage? Is the for decades without risk or harm. Patients with
plug bent or broken? Is the equipment pacemakers or implanted cardiac defibrillators
labeled with a current electric safety are sometimes referred for nerve conduction
inspection sticker? studies and needle EMG. Potential risks, such
2. Inspect the patient area for any two- as induction of ventricular or atrial fibrilla-
wire ungrounded appliances. Have them tion, alteration of pacing mode, or cardiac tis-
unplugged and removed. sue damage from the electrical current, may
3. Inquire about any other instruments result in limitation of performance of studies in
attached to the patient. Are they labeled these patients. However, in a small study of 10
with a current electric safety inspection patients with pacemakers or implanted cardiac
sticker? defibrillators who underwent nerve conduc-
4. Choose an outlet in the same area or tion studies with stimulation at common sites,
cluster used by other patient-connected no abnormalities in the devices occurred and
devices. Before plugging in the equip- the electrical impulses generated during the
ment, check the contact tension of the studies were never detected by the sensing
chosen receptacle with a simple device amplifier.8
Transcranial Electrical and voltage, pulse duration (pulse width), and fre-
Magnetic Stimulation quency can be adjusted within limits to achieve
optimal effectiveness in treating symptoms
Stimulation of brain, spinal cord, peripheral with minimal side effects.
nerve, and muscle is being used increas- There are two recognized types of tissue
ingly for both diagnostic and therapeutic pur- response to implanted electrodes: (1) A passive
poses. Stimulators are isolated from ground tissue response can result from surgical trauma
and routinely allow less than 300 V or 100 mA. and the mechanical and chemical properties
Most stimulation is done with constant current of the implant. (2) An active tissue response
rather than constant voltage to better assure results from electrochemical reaction products
similar local current flows, despite varying tis- formed at the tissue–electrode interface and
sue impedance. Transcranial electric stimula- from physiologic changes associated with neu-
tion may be as high as 800 V to the scalp; ral activity induced by stimulation. The man-
the high impedance of the skull results in ner and degree to which neural injury occurs
only small current flow reaching the brain. depends on both the stimulation parameters
While current flow through neural tissue is and the neural substrate being stimulated.
low in all of these applications, safety consid- There is an interface between the metal
erations remain important.9 Direct stimulation stimulating electrodes and the ionic conduc-
of the cortex for localization of motor cor- tor of the body, which has an impedance Z(V)
tex and other eloquent brain regions uses at that is in general nonlinear (dependent on
most short trains of brief pulses that have been the voltage across the interface, V). Since the
shown in animal models to have no deleterious circuit used to deliver the electrical stimu-
effects.10 More recent chronic deep brain stim- lus is a constant-voltage device, a changing
ulation therapy uses much longer, rapid trains impedance of the electrode–tissue interface
of stimuli, but with very low current densities.11 leads to changes in the magnitude and time
Transcranial magnetic stimulation is widely course of the current that flows through the tis-
used in Europe for diagnosis of demyelinating sue, which is the primary determinate of the
diseases with no safety issues other than insti- effects of electrical stimulation on neurons.16
tutional requirements.12 Transcranial electric Tissue damage may result from products
stimulation with trains of three to five pulses at of electrochemical reactions at the electrode–
up to 800 V has also been found to be safe and tissue interface. These reactions depend upon
effective with only rare, minor complications the potential of the electrode and the time
such as scalp burns or masseter contraction.13 course of the stimulus pulse. Electrode geom-
etry and area may also be important, since the
current density on a metal electrode passing
current in an ionic conductor is nonuniform.
Therapeutic Cortical and Deep In addition to electrochemical mechanisms,
Brain Stimulation physiologic mechanisms involving synchronous
activation of populations of neurons also con-
Deep brain stimulators have been used to tribute to neural damage. Physiologic changes
treat a variety of movement disorders including associated with neural excitation are correlated
essential tremor, Parkinson’s Disease, and with the charge rather than the charge density
dystonia.14, 15 Such stimulators typically apply (charge divided by electrode area). Thus, both
100-200 Hz extracellular electrical stimulation charge and charge density must be limited to
to target brain areas such as thalamus, sub- avoid tissue damage.17 For cortical stimulation,
thalamic nucleus, or globus pallidus. There another concern is the threshold for kindling of
have also been research trials of deep brain brain tissue, which is the development of spon-
stimulation and cortical stimulation for treat- taneous epileptic seizures due to long-term
ment of epilepsy. The stimulator itself is a synaptic potentiation and other mechanisms.
pacemaker-like, battery-powered pulse gen- Kindling is usually thought to occur only if an
erator implanted subcutaneously with wires after-discharge is seen as a result of electri-
connecting to the implanted electrodes. The cal stimulation, since in experimental animals
output of the pulse generator can be no kindling is seen with subthreshold stimuli
programmed by a clinician. The stimulation (those that do not produce after-discharges).
Stimulation frequency may also have an 60-Hz intracardiac leakage current. Implications for
effect on tissue damage, since stimulus fre- electrical safety. Circulation 99:2559–64.
quency determines how often the activated 4. Litt, L., and J. Ehrenwerth. 1994. Electrical safety
in the operating room: Important old wine, disguised
neurons fire, and it is possible that cellular new bottles. Anesthesia and Analgesia 78:417–19.
respiration cannot keep up with the increased 5. McNulty, S. E., M. Cooper, and S. Staudt. 1994. Trans-
demands of neuronal ionic pumps for ATP dur- mitted radiofrequency current through a flow directed
ing high frequency firing. Thus, higher stim- pulmonary artery catheter. Anesthesia and Analgesia
78:587–9.
ulation frequencies may be associated with 6. UL 60601-1, EN/IEC 60601-1 Medical Electrical
a lower charge threshold for neural dam- Equipment—Part 1 General Requirements for Safety
age. Deep brain stimulation uses comparatively and Essential Performance.
high (100–200 Hz) frequencies. 7. NFPA99, ANSI/AAMI ES1 Safe Current Limits for
Electromedical Apparatus.
8. Schoeck, A. P., M. L. Mellion, J. M. Gilchrist, and
F. V. Christian. 2007. Safety of nerve conduction stud-
Key Points ies in patients with implanted cardiac devices. Muscle
& Nerve 35:521–4.
• Electric stimulation of neural tissue at 9. Shannon, R.V. 1992. A model of safe levels for elec-
the currents needed for activation is safe trical stimulation. IEEE Transactions on Biomedical
if kept away from pacemakers and other Engineering 39(4):424–6.
implanted electrical devices. 10. Swartz, B. E., J. R. Rich, P. S. Dwan, et al. 1996. The
• Care must be taken to prevent local tissue safety and efficacy of chronically implanted subdural
electrodes: A prospective study. Surgical Neurology
injury. 46(1):87–93.
11. Kenney, C., R. Simpson, C. Hunter, et al. 2007. Short-
term and long-term safety of deep brain stimulation
in the treatment of movement disorders. Journal of
SUMMARY Neurosurgery 106(4):621–5.
12. Foerster, A., J. M. Schmitz, S. Nouri, and D. Claus.
This chapter reviews the principles of electric 1997. Safety of rapid-rate transcranial magnetic stim-
safety that are relevant to clinical neurophys- ulation: Heart rate and blood pressure changes.
Electroencephalography and Clinical Neurophysiol-
iologic studies. Knowledge of these principles ogy 104(3):207–12.
is necessary both for those involved in evalu- 13. MacDonald, D. B. 2002. Safety of intraoperative tran-
ating and purchasing test instruments and for scranial electrical stimulation motor evoked poten-
those involved in maintaining and using them. tial monitoring. Journal of Clinical Neurophysiology
All those who order, perform, interpret, or 19(5):416–29.
14. Putzke, J. D., R. E. Wharen Jr., Z. K. Wszolek,
supervise electrophysiologic testing share the M. F. Turk, A. J. Strongosky, and R. J. Uitti. 2003. Tha-
legal responsibility for patient safety, including lamic deep brain stimulation for tremor-predominant
electric safety. Parkinson’s disease. Parkinsonism and Related Disor-
ders 10(2):81–8.
15. Sydow, O., S. Thobois, F. Alesch, and J. D. Speel-
man. 2003. Multicentre European study of thalamic
REFERENCES stimulation in essential tremor: A six year follow up.
Journal of Neurology, Neurosurgery, and Psychiatry
1. Anonymous. 1993. Risk of electric shock from patient 74:1387–91.
monitoring cables and electrode lead wires. Health 16. Grill, W. 2005. Safety considerations for deep brain
Devices 22:301–3. stimulation: Review and analysis. Expert Review of
2. Starmer, C. F., H. D. McIntos, and R. E. Whalen. Medical Devices 2(4):409–20.
1971. Electrical hazards and cardiovascular function. 17. Agnew, W. F., D. B. McCreery, T. G. H. Yuen, and
New England Journal of Medicine 284:181–6. L. A. Bullara. 1990. Effects of prolonged electrical
3. Swerdlow, C. D., W. H. Olson, M. E. O’Connor, D. M. stimulation of the central nervous system. In Neu-
Gallik, R. A. Malkin, and M. Laks. 1999. Cardiovas- ral prostheses fundamental studies, ed. W. F. Anew,
cular collapse caused by electrocardiographically silent and D. B. McCreery, 226–52. Englewood Cliffs, NJ:
Prentice-Hall.
Chapter 3
Volume Conduction
Terrence D. Lagerlund, Devon I. Rubin, and Jasper R. Daube
PRINCIPLES Distant Recordings in Volume

SOURCES OF ELECTRICAL Conductors
POTENTIALS
APPLICATIONS OF VOLUME
Cortically Generated Potentials in
CONDUCTION PRINCIPLES
Volume Conductors
SEP and Brain stem Auditory
Peripherally Generated Potentials
Evoked Potential Applications
in Volume Conductors
Effect of Volume Conduction on
CURRENT SOURCES: Electroencephalography (EEG)
MONOPOLES, DIPOLES, AND Applications
QUADRUPOLES Dipole Source Localization in EEG
Spatial Distributions of Potentials Nerve Conduction Study
Applications
VOLUME CONDUCTION Needle EMG Applications
THEORY: ELECTRIC
PROPERTIES IN VOLUME SUMMARY
CONDUCTORS APPENDIX
Effect of Volume Conductor Calculating Potentials in Infinite
Resistive–Capacitive Properties Homogeneous Media
on EEG Potentials in Nonhomogeneous
Effect of Volume Conductor Media
Resistive–Capacitive Properties Homogeneous Sphere Model
on Nerve Conduction Studies Multiplanar and Multiple Sphere
and Needle EMG Models
PRINCIPLES potentials may be either active or passive.

Active sources are ionic channels that open or
Electrophysiologic studies involve bioelectric close in response to changes in transmembrane
potentials generated by sources inside the potential, neurotransmitter binding, intracel-
body, such as the brain, peripheral nerve, and lular calcium, or second messengers, allowing
muscle, that may be some distance away from small currents to flow into or out of the cell
the recording electrodes. The sources of these body, dendrite, axon, or muscle fiber. Passive
33
sources are areas of neuronal membrane that in which multiple synapses may be active
permit current flow into or out of the cell simultaneously and in which multiple regions
by passive leakage or capacitive effects. These of active membrane are capable of generat-
current sources or sinks (active or passive) lead ing action potentials. In this case, the poten-
to widespread extracellular currents flowing in tials generated by each neuron are a sum of
the conducting medium throughout the body, the potentials generated by multiple active
called a volume conductor. The transfer of and passive areas of membrane. Only if the
electric potentials to a site a distance away from responsible neuronal generators are arranged
the generator is called volume conduction. regularly and activated more or less syn-
Volume conductors may be homogeneous, chronously is sufficient summation obtained to
such as a cylinder containing an electrolyte allow recording of potentials at a considerable
solution; however, in clinical neurophysiology distance from the generators.
the body acts as a nonhomogeneous volume In the cerebral cortex, the generator of spon-
conductor. taneous electroencephalographic (EEG) activ-
Some of the currents in the volume conduc- ity, the pyramidal neurons are arranged in a
tor reach the skin surface, where the current regular manner, with the main axes of the
causes a potential difference across the space dendritic trees parallel to one another and per-
between two recording electrodes. This differ- pendicular to the cortical surface. Thousands
ence in potential can be detected and amplified of these cortical pyramidal neurons are acti-
by a differential amplifier. The source of the vated more or less simultaneously by synapses
electrical potentials, the type of volume con- made by a single axon or small groups of axons,
ductor, the propagation of current through the producing significant extracellular current flow.
volume conductor, relationship between the Under these circumstances, the longitudinal
recording electrode montages, and distances components of current flow from different
of the electrodes from the electrical generator neurons add together, and the transverse com-
all have an effect on the potentials recorded ponents of flow cancel out, producing a laminar
in clinical neurophysiology. Volume conduction current along the main axes of the neurons.
theory describes the spread of electrical cur- Depending on whether the activated synapse is
rent throughout the body and plays an impor- excitatory or inhibitory, the direction of current
tant role in the responses recorded in clinical flow across the cell membrane is either inward
neurophysiology. or outward. The synaptic transmembrane cur-
rent flow is accompanied by an opposite out-
Key Points ward or inward current flow at another location
along the dendritic tree, called passive source
• Currents generated by sources in the body or sink, which produces a dipole, as described
flow in the conducting medium (volume in the following section. An excitatory postsy-
conductor) to reach electrodes. naptic potential (EPSP) occurs when positive
• Current flow causes potential differences ions flow intracellularly, called inward current
between electrodes which can be ampli- flow, and an inhibitory postsynaptic potential
fied and recorded (Ohm’s law). (IPSP) occurs when negative ions flow intra-
cellularly, called outward current flow. Thus,
the local extracellular potential produced by an
SOURCES OF ELECTRICAL EPSP is negative and that produced by an IPSP
POTENTIALS is positive.
The orientation of the dipole created by
synaptic activity in the cerebral cortex depends
Cortically Generated Potentials in on both the type of synaptic activity, whether
Volume Conductors an EPSP or IPSP, and the location of the
synapses, whether superficial or deep. An
The potential generated by a population of EPSP located superficially in the cerebral cor-
neurons is equal to the sum of the potentials tex, that is, along the distal branches of the
generated by the individual neurons. Cerebral pyramidal cells, produces a dipole with a
cortical neurons have extensive dendritic trees superficial negative and a deep positive pole.
Volume Conduction 35
A deep EPSP, for example, caused by a synapse surface, many dipole-like sources of EEG
near the cell body or on the basal dendrites, and evoked potential waveforms are radial in
produces a dipole with a superficial positive direction, that is, they are perpendicular to the
and a deep negative pole.1 The IPSPs and surface of the scalp. These generators typi-
EPSPs located at similar depths in the cerebral cally reside at the apex of cortical gyri. How-
cortex produce dipoles oriented opposite to ever, cortical generators located in the walls of
one another (Fig. 3–1). A deep IPSP produces sulci—where the cortical surface is perpendic-
an extracellular potential field similar to that of ular to the scalp surface—may create potential
a superficial EPSP. fields that correspond to a tangentially ori-
At a macroscopic level, the potential field ented dipole. A classic example of this is the
generated by synchronous activation of many potential field of the centrotemporal spike dis-
cortical pyramidal cells behaves like that of charges often seen in benign rolandic epilepsy
a dipole layer. This has been called an open of childhood.
field configuration, in contrast to the fields
generated by neurons with dendritic arboriza- Key Points
tions that are distributed radially around the
cell body and called closed fields. Closed-field • Currents generated by sources in the body
potentials are equivalent to the field produced flow in the conducting medium (volume
by a set of radially oriented dipoles at the sur- conductor) to reach electrodes.
face of a sphere; such a field is negligible at a • Potentials generated by a population of
distance because both the radial and tangential neurons are the sum of the potentials
components of current flow cancel each other generated by individual neurons.
in this configuration. • Regular arrangement and synchronous
Because the dendrites of cortical pyramidal activation of many neurons are required
cells are perpendicular to the cortical, or pial, for recording potentials at a distance.
• In EEG and cerebral evoked potentials,
the summated EPSPs and IPSPs from
cerebral cortical neurons are the source of
I the electrical recordings.
E • Parallel pyramidal cells activated simulta-
– neously in cerebral cortex allow effective
– + summation of EPSPs and IPSPs.
+ • These cortical sources assume the config-
uration of a dipole layer with the dipoles
perpendicular to the pial surface.
– • Radial dipoles are formed by activation
– of PSPs in gyral cortex, and tangential
+
+ dipoles by PSPs in cortex in sulci.
Figure 3–1. Patterns of current flow near a neuron Peripherally Generated Potentials
caused by synaptic activation. E, Current flow caused by
activation of an excitatory apical dendritic synapse depo-
in Volume Conductors
larizes the cell membrane, producing a current sink. The
extracellular potential, shown on the left, has a negative Spontaneous activity may arise from any cen-
polarity at the synapse. I, Current flow caused by activation tral or peripheral neuromuscular structure.
of an inhibitory synapse near the cell body hyperpolar- Voluntary potentials are initiated in the cor-
izes the cell membrane, producing a current source. The
extracellular potential, shown on the right, has a positive tex and bring about voluntary movement by
polarity at the synapse. (From Lopes da Silva, F., and traveling to muscle via the corticospinal tracts,
A. van Rotterdam. 1993. Biophysical aspects of EEG and motor neurons in the ventral horn of the
magnetoencephalogram generation. In Electroencephalog- spinal cord, and peripheral motor nerves to the
raphy: Basic principles, clinical applications, and related
fields, ed. E. Niedermeyer, and F. Lopes da Silva, 3rd ed.,
muscles. The synchrony and size of fibers in
78–91. Baltimore: Williams & Wilkins. By permission of spontaneous activity and voluntary activation is
the publisher.) sufficient to allow clinical recording of these
potentials only in muscle by recording individ- traveling potentials that have unique properties
ual muscle fiber potentials or the muscle fibers in volume conductors with distinctly different
in a motor unit with a needle electrode (motor appearances based on their location relative to
unit potential). the recording electrodes. The nerve and mus-
Peripheral evoked potentials are initiated cle fiber action potentials are typically recorded
by stimulation of peripheral motor and sen- from the overlying skin as close as possible
sory nerves or of the motor cortex. Potentials to the generating or propagating source; how-
from motor cortex stimulation travel peripher- ever, the recording electrodes are often located
ally to anterior horn cells and muscles. They some distance away from the nerve or muscle
can be recorded from the spinal cord (as motor generator with current passing through inter-
evoked potentials), peripheral nerve, and mus- vening tissue before reaching the recording
cle (as compound muscle action potentials). electrodes.
Peripherally activated sensory potentials travel Fibers in peripheral nerve and muscle can
peripherally to sensory nerve endings and be recorded individually, but most commonly
centrally to the cortex via the dorsal roots are recorded from the synchronous volley of
and the dorsal columns of the spinal cord. action potentials in multiple, closely grouped,
Summated sensory evoked potentials can be parallel fibers that produce the nerve and mus-
recorded in peripheral nerve (sensory nerve cle action potentials recorded clinically. The
action potentials, SNAPs), spinal cord, and waveforms from these groups of parallel fibers
cortex (as somatosensory evoked potentials, generate a nerve action potential. A traveling
SEPs). nerve action potential in a peripheral nerve
The peripheral and spinal cord motor and can be represented by two dipoles placed end-
sensory fibers serve to carry information from to-end. Figure 3–2 shows the current flow
one area of the nervous system to another. The and potential fields surrounding a nerve action
potentials in these structures are, therefore, all potential.
A B
cm n cm
20 cm
st n
oil ins –
t.e.
30 cm
mm
C 15
0
–
–4.2
+ 10
+ 0 – 7
12 5
+7 4
19 3
+3 29 14 2
40 28 1
n + 6.5 –48 +55 n 11-20-38

mm 0 1 2 3 4 5 10 15 20 26
Figure 3–2. Nerve action potential recorded from a nerve on a volume conductor. A, The volume conductor in cross
section showing isolated stimulation of the nerve in an oil bath. B, Nerve (n) recorded between testing electrode (t.e.)
moved to different locations on the volume conductor and reference electrode. C, Current flow (lines with arrows) and
potential fields (numbered lines) showing the decreasing positive and negative voltage at increasing distance from the
nerve. (From Lorente de No R. 1947. A study of nerve physiology, Vol. 2, 384–477. New York: The Rockefeller Institute
for Medical Research. By permission of the publisher.)
Figure 3–3. Schematic diagram of the cross section of the upper arm showing the major nerves and muscles and the
humerus. A, Median nerve action potentials with stimulation at the wrist were recorded from 24 locations with six surface
recordings around the arm and from a needle electrode inserted to three depths at the six points. B, Cross section of the
potential fields derived the recorded median nerve action potentials at the 24 points. Note the marked distortion of the
potential by the humerus.
The configuration and size of these poten- Volume conduction of nerve action poten-
tials depends on the relationship of the record- tials is shown in Figures 3–3 and 3–4.
ing electrode to the generator and may be Figure 3–3A illustrates the distribution of the
seen as positive waveforms, biphasic wave- potential field surrounding a median nerve
forms, or triphasic waveforms when recorded action potential in the upper arm. The distri-
from peripheral nerve, muscle, or dorsal col- bution of isopotential lines constructed from
umn axons. these recordings shows the distortion of the
Stimulate:
Median nerve Ulnar nerve Radial nerve A
B B E
C D
D
50 μV
E
+
5 ms
Figure 3–4. Surface recordings of the median, ulnar and radial SNAP from the skin at five sites around the upper arm.
Recordings all at the same amplifier sensitivity. Note that while the potential can be recorded around the arm, there is a
rapid fall off in amplitude away from the nerve.
potential field by the humerus (Fig. 3–3B). The the source can flow through the medium to
spread of nerve action potentials across the skin the sink where it is absorbed. In respect to
overlying the upper arm nerves is shown in the potentials they produce at distant record-
Figure 3–4. ing sites, many neuronal current generators
may be well described in terms of a cur-
Key Points rent dipole. For example, as noted above, the
main contributors to spontaneous EEG activity
• Peripheral EMG, motor and sensory are the excitatory and inhibitory postsynap-
nerve action potentials, and motor-evoked tic potentials in the dendritic trees of corti-
potentials are the summated activity of cal pyramidal neurons. The arrangement of
parallel bundles of individual nerve or synapses on the dendritic trees produces a
muscle fibers. current source and sink separated by a sig-
• Sensory evoked potentials at the cortex nificant distance, and this constitutes an elec-
are the summated EPSPs and IPSPs from tric dipole. The characteristic organization of
cerebral cortical neurons. the cortical pyramidal neurons, that is, ori-
• Potentials recorded from muscle may be ented parallel to each other and perpendicular
◦ Voluntary motor unit potentials (MUPs) to the cortical surface, allows the potentials
◦ Spontaneous potentials in muscle (e.g., from many such dipole sources to summate
fibrillation potentials) effectively.
◦ Evoked compound muscle action poten- The potentials of such dipoles fall off
tials (CMAPs). inversely with the square of the distance
• Potentials from nerve are from the source. The lines of current flow
◦ Sensory nerve action potentials (SNAPs) around a dipole form curved paths (Fig. 3–5B).
◦ Nerve action potentials (NAPs). The equipotential surfaces are perpendi-
cular to the lines of current flow and
have a figure-8 configuration around the
dipole.2 The zero potential surface is a plane
CURRENT SOURCES: halfway between the two poles of the dipole
MONOPOLES, DIPOLES, AND (Fig. 3–5B).
QUADRUPOLES Two adjacent current dipoles of opposite
orientation placed end-to-end effectively act
Every electrical potential has a source of cur- as two end-to-end dipoles or quadrupole
rent. A single source or sink of current is (Fig. 3–5C). The potential of a quadrupole
referred to as a monopole. The magnitude of falls off inversely with the cube of the dis-
the current density decreases with distance tance from the source, and the equipotential
away from the current source and can be mea- surfaces around the quadrupole have a clover-
sured along equipotential lines. Each equipo- leaf configuration (Fig. 3–5C). A quadrupole
tential line represents a constant potential is a fair approximation of the potential gener-
along the course of the line. In a monopole ated by an action potential propagating along
the equipotential lines form circles around the an axon: The axonal membrane has a neg-
current source or sink, and the magnitude of ative polarity outside and a positive polarity
current falls off inversely with distance from inside at the peak of the action potential. How-
the source (Fig. 3–5A). ever, on either side of this peak, the mem-
In the nervous system, adjacent current brane is positive outside and negative inside
monopoles of opposite polarity constitute a (Fig. 3–2).
current dipole (Fig. 3–5B). In a dipole, cur- In the simplest form of a volume conduc-
rent flows from the positive to the negative tor, a homogeneous medium without bound-
pole, and sets of potential lines are gener- aries in which generators and recording
ated away from the dipole. Similar to the electrodes are embedded, the recorded poten-
monopole, the magnitude of current falls off tial can be calculated from the configura-
inversely with distance from the source. This tion of source currents. The formulas used
is a more realistic generator than an isolated to calculate the potential are listed in the
monopole because the current emanating from Appendix.
A y B
–1.0 y
v = –1 –1.0
v = –0
–0.5 1 0.5 –1
–2 2 –2
5 –5
–5
x
– +– x
1.0 0.5 –0.5 –1.0 1.0 0.5 –0.5 –1.0
–0.5 –0.5
1.0 1.0
y
–1.0
–1
v = –0 –2 v = –0
C
1 –5 1
2 2
–0.5
5 5
+ –+ x
1.0 0.5 –0.5 –1.0
–5 0.5
–2
–1
1.0
Figure 3–5. Equipotential lines in a volume conductor for different current source distributions: A, A point source or
monopole; B, a dipole; C, a quadrupole (two oppositely directed dipoles); similar to the equipotential lines around an
action potential propagating along an axon. The arrows represent lines of current flow. The distribution of the potential
field in the volume conductor is shown in millimeters on the horizontal and vertical scales. (From Stein, R. B. ed. 1980.
Nerve and muscle: membranes, cells, and systems. New York: Plenum Press. By permission of the publisher.)
Key Points a monopole source increases with decreas-

ing distance (Fig. 3–6). The configuration of
• A single current source (monopole) in a
a dipole potential depends on the reference
conductor creates a potential that varies electrode location, distant or closely spaced
inversely with distance from source. (bipolar), and on the orientation of the elec-
• A current dipole’s potential varies inversely
trodes relative to the dipole (perpendicular or
as square of distance from source and parallel).
depends on the angle from the dipole axis. A dipole recorded relative to a distant ref-
• A quadrupole (two adjacent, end-to-
erence at points along a line perpendicular to
end dipoles) has a potential that varies the dipole axis is a single peak whose sharp-
inversely with the cube of distance from ness increases with decreasing distance from
source. the source (Fig. 3–7A). This situation applies to
scalp potentials produced by a radially oriented
Spatial Distributions of Potentials cortical dipole generator. For example, a radi-
ally oriented cortical dipole is seen frequently
The spatial distributions of potentials vary with in EEG recordings because of the radial orien-
the source. The sharpness of a potential from tation of the cortical pyramidal neurons.
1.0
A
y=1
(V 1)
0.5
2
5
0
0.5
B
y=1
Voltage (V 2)
0 2
–0.5
0.5
C
0
2
(V 3)
–0.5 Figure 3–6. Potentials recorded along a line located at

various distances from a current source (1, 2, and 5 cm), as
y=1 a function of position along the line: A, a current monopole,
B, a dipole, and C, an action potential source. (From Stein,
–1.0 R. B. ed. 1980. Nerve and muscle: Membranes, cells, and
–10 –5 0 5 10 systems. New York: Plenum Press. By permission of the
Distance (x) publisher.)
A v C v
–5 5
y, cm
–5 0 x, cm 5
v v
B D
–5 5
x, cm
–5 5
y, cm
Figure 3–7. Potentials recorded along a line located at various distances from a current dipole (solid curve, 1 cm; dashed
curve, 2 cm; dotted curve, 3 cm) as a function of position along the line. A, Referential recording, with the line perpendicular
to dipole axis; B, bipolar recording for line perpendicular to dipole axis; C, referential recording, with the line parallel to
dipole axis; D, bipolar recording for line parallel to dipole axis.
40
For the same dipole source with bipolar VOLUME CONDUCTION

recording, the potential differences at points THEORY: ELECTRIC PROPERTIES
along a line perpendicular to the dipole axis IN VOLUME CONDUCTORS
may also be plotted as a function of distance
along the line. The potential obtained with this In general, a volume conductor such as the
recording is biphasic, with increasing sharp- body can be characterized by its conductiv-
ness with decreasing distance (Fig. 3–7B). ity (or resistivity) and its capacitive properties
For a dipole source, the potential recorded (dielectric constant), which may vary from tis-
relative to a distant reference at points along sue to tissue. This may be modeled by divid-
a line parallel (tangential) to the dipole ing the entire volume conductor into many
axis is biphasic, and its sharpness increases small regions, each of which is assumed to
with decreasing distance from the source be homogeneous, that is, the conductivity and
(Figs. 3–6B and 3–7C). This occurs with a tan- capacitance are the same throughout.
gentially oriented cortical dipole generator that
is often present when the cortical generator
region is deep in a sulcus or fissure in which the Effect of Volume Conductor
cortical surface is perpendicular to the scalp Resistive–Capacitive Properties
surface, for example, the spikes with benign
rolandic epilepsy of childhood.
on EEG
For a dipole source with bipolar recording,
The capacitive properties of the volume con-
the potential difference at points parallel to the
ductor can be ignored at the frequencies of
dipole axis is a triphasic potential whose sharp-
interest in EEG recordings. Thus only a con-
ness increases with decreasing distance from
ductivity value for each region, together with
the source (Fig. 3–7D).
the geometry of the region, is needed to char-
The potential with a quadrupole source
acterize the volume conductor. In the nonca-
recorded relative to a distant reference along
pacitive (purely resistive) volume conductor,
a line parallel to the quadrupole axis is
EEG potentials are always in phase, or syn-
triphasic; its sharpness increases with decreas-
chronous, with the current sources, and the
ing distance from the source (Fig. 3–6C).
conductive properties of the medium are inde-
This situation approximates an action potential
pendent of frequency.
propagating along an axon parallel to the line
of the recording electrodes.
Effect of Volume Conductor

Key Points Resistive–Capacitive Properties on
• Spatial distributions of potentials gener- Nerve Conduction Studies and
ated by different sources can be displayed Needle EMG
by plotting potential against distance.
• Waveform configurations in a volume con- In contrast to the capacitive properties of the
ductor depend on location of the refer- volume conductor during EEG recordings, in
ence electrode, orientation of the elec- peripheral recordings in a resistive–capacitive
trodes to the generator, and distance of medium, volume conduction is frequency
the electrodes from the generator as sum- dependent. The result is that the source cur-
marized in the following table: rents and recorded potentials are out of phase,
Reference Orientation to dipole Waveform configuration Nearer generator
A Distant Perpendicular Single peak Sharper peak

B Bipolar Perpendicular Biphasic Sharper peak
C Distant Parallel Biphasic Sharper peak
D Bipolar Parallel Triphasic Sharper peak
Medial (Over nerve) Lateral (Away from nerve) some instances, the recording electrodes are
6.2 ms placed in close proximity to the electrical gen-
4.5 ms 6.7 ms erator; however, in many instances the gener-
ator is located at a longer distance away from
the recording electrodes. While two recording
5.0 ms electrodes on an equipotential line of a cur-
rent source will record no potential difference,
5.5 ms in most recordings the two electrodes will be
Figure 3–8. Median nerve action potentials with stimu- on different isopotential lines and a potential
lation at the wrist and an ipsilateral knee reference are difference will be recorded.
recorded medially over the course of the median nerve
and laterally, opposite the median nerve. Latency of the
components of the action potential over the nerve is signif-
icantly shorter than that recorded on the opposite side of A
the arm.
1 ms
100 μV
30 μ
–
a +
A1
2 mV
A2
350 μ
b
Figure 3–9. Compound muscle action potentials
recorded from the extensor digitorum brevis muscle at
the knee and ankle with peroneal nerve stimulation at
the ankle and knee. The additional positivity seen at the 460 μ
knee is contributed by activation of anterior compartment c
muscles at a distance, such as the extensor hallucis longus.
B
Low
which can result in erroneous latency mea-
surements of propagating nerve action poten-
tials with distant references (Fig. 3–8). The
latency recorded at a distance from the nerve High
is shorter than that recorded at the same level
over the nerve. A more common issue is the
presence of an initial positivity on a muscle a
action potential generated by a segment of
muscle at a distance from the recording site b
(Fig. 3–9).
c
Figure 3–10. A, Muscle fiber action potential recorded
Distant Recordings in Volume from a frog muscle fiber in a volume conductor at three dis-
tances (μ = micro) from the fiber (from Haakenson. 1951.
Conductors Acta Physica Slovaca 39:291–312). B, Schematic diagram
of the spatial current gradients in a volume conductor that
In clinical neurophysiology, the electrical cur- result in the change in muscle fiber action potential with
rent arising from neurons, axons, or muscle distance from the fiber (a, electrode pair with a high spa-
fibers are recorded with recording electrodes tial gradient; b, electrode pair with an intermediate spatial
gradient; c, electrode pair with a low spatial gradient) with
placed at different sites on the body. Two an exponential reduction in both amplitude and rise time
electrodes record the potential differences that from the initial positive peak to the negative peak due to
are generated from the neural structures. In filtering of fast frequency components.
When differences in a potential are large A B C

–
between two adjacent points, a high spatial 500 μV
– 500 μV
gradient (or high rate of change of a potential)
is present. In regions where the spatial gradient +
from the source is high at the site of record-
+
ing, the potential is called a near-field potential.
In contrast, when the potential is recorded far
from the current source, the spatial gradient is
low, and these recorded potentials are known
as far-field potentials.
When using a bipolar recording montage,
potentials near the generator and each record- 0.5 mm –
ing electrode located on equipotential lines 5 μV
with a larger spatial gradient between the two +
will be recorded as a large response, because
large potential differences will be recorded. In
contrast, if using a bipolar recording montage
where both electrodes are far from the gener-
ator source, both electrodes will be recording
current along isopotential lines with a low spa-
tial gradient and, therefore, a low amplitude
response will be recorded.
Volume conduction theoretically allows the
potentials from motor units to be recorded
anywhere in the body. For practical purposes,
it means that they can be recorded at some 0 5 10 ms
distance from the generator. The effect of
0 .25 .50 ms
spatial gradients are shown schematically in
Figure 3–10B and on a muscle fiber action Figure 3–11. Motor unit potential recordings at short
potential in Figure 3–10A. The capacitance of increments along a recording needle electrode. A, Iden-
tical amplification of all waveforms. B, Amplification
the intervening tissue results in filtering of the increased along the needle (as shown by the 500-μV
fast frequencies and distortion of the signal as bracket). C, Change in rise time at a fast sweep speed
shown in Figure 3–11. (compare the 5-mV response superimposed on all of the
smaller responses). The amplitude decreases markedly at
short distances from the site of the muscle fibers, but can
be recorded two millimeters away. At these sites not only
Key Points is the amplitude markedly reduced by resistance, but also
the rise time is slowed by tissue capacitance.
• A volume conductor has properties of
conductivity (or resistivity) and capacitive
(dielectric constant) properties. ◦ Rise time slows as distance from the
• Different body regions have different con- generator increases.
ductivities and dielectric constants which ◦ Nerve and muscle action potentials
influence volume conduction. recorded at a distance from the gener-
• In EEG recordings capacitive properties ator are positive.
can be ignored, so that each region need
be characterized only by conductivity.
• The frequency dependence for volume APPLICATIONS OF VOLUME
conduction in a resistive–capacitive medium CONDUCTION PRINCIPLES
alters peripheral recordings significantly.
◦ Nerve conduction latency measure-
ment becomes erroneous when recorded SEP and Brain stem Auditory
at a distance from the generator. Evoked Potential Applications
◦ Nerve and muscle action potential
amplitudes decrease rapidly as distance SEPs are action potentials elicited along the
from the generator increases. peripheral and central somatosensory pathways
that are induced by stimulation of the sensory described above, and localized potentials gen-
fibers of a peripheral nerve. Different types erated by groups of neurons in the dorsal
of potentials are recorded: action potentials horn of the spinal cord. The configuration of
that propagate along the peripheral nerve and these potentials depends on the location of the
spinal cord, and current that is generated at the recording electrodes.
sensory ganglion within the brain stem or tha-
lamus which eventually reaches the somatosen-
sory cortex. SEPs may be recorded at different STATIONARY POTENTIALS
points along the sensory conduction pathway, PRODUCED BY PROPAGATING
such as along the peripheral nerve at the knee GENERATORS
or elbow, along the lumbar or cervical spine, However, when a propagating generator
and at the scalp. passes through an interface between volume-
conducting regions of different sizes or
conductivities, a potential can be induced
EFFECT OF VOLUME simultaneously at all recording electrodes dur-
CONDUCTION ON POTENTIAL ing the time at which the generator is crossing
COMPONENTS DUE TO the boundaries between regions with differing
PROPAGATING GENERATORS properties.3 Such a potential, which does not
appear at different times in different recording
When recording from various electrodes locations, has been referred to as a station-
placed at different locations along the path of ary potential. This effect may be observed in
a complex propagating generator, such as an SNAPs and SEP recordings when recording at
action potential source, the time at which the a single site as a consequence of the change in
propagating potential is seen at each location is geometry of the volume conductor as a propa-
different because of the finite velocity of prop- gating nerve impulse travels from a limb to the
agation of the generator. (Note that volume trunk (Fig. 3–13). The artifactual occurrence
conduction of electric potentials from the gen- of stationary potentials seen at the interface of
erator to the recording electrode is essentially differently sized volume conducting regions is
instantaneous, because electric disturbances illustrated.
propagate at the speed of light in a conduct- The same effect may also influence the mor-
ing medium.). These differences in time of phology of a brain stem auditory evoked poten-
recording the electrode is what is typically tial (BAEP) recording because of changes in
seen and analyzed during SEP recordings with volume conductor properties along the central
a standard recording montage from different auditory pathways caused by the complicated
sites along the conduction pathway (Fig. 3–12). anatomy of the posterior fossa. These station-
SEPs at the cervical segments of the spinal ary potentials can be seen only in derivations in
cord are combinations of the synchronous, which the first and second electrodes between
traveling potentials in nerve fiber bundles as which the potential is measured are on oppo-
site sides of the boundary between the regions
with differing sizes or conductivities; generally,
–
this occurs only when recording with respect to
P11
P13/14 2 μV a relatively distant reference electrode.4
+
P9
Key Points
Cv – Fz
• Propagating action potentials are recorded
from various locations at different times
due to finite propagation velocity.
0 10 20 30 40
Milliseconds • A stationary potential occurs when a prop-
agating action potential passes an inter-
Figure 3–12. Median SEP recorded from the cervical
spinal cord (Cv–Fz). The origins of these typical peaks face between regions of different sizes
are P9 from the brachial plexus; P11 from the dorsal or conductivities, generally when record-
root entry zone; and P13/14 from dorsal horn and dorsal ing with respect to a relatively distant
column/medial lemniscus. reference.
• SEPs are complex waveforms made up of the head and the distant position would make
combinations of traveling waves in fiber the reference active; that is, the reference
pathways and cell groups in the spinal cord electrode would be electrically equivalent to
or cortex. a reference at the neck (still relatively close
• Evoked potentials from spinal cord are to intracranial generators), with a slight addi-
SEPs. tional resistance that is negligible in its effect
because of the very large input resistance of the
EEG amplifier (Fig. 3–14). In addition, such
a distant reference would have unacceptable
Effect of Volume Conduction on characteristics in that very large artifacts, for
Electroencephalography (EEG) example, those produced by the electrocardio-
Applications gram, movement, and muscle, would probably
be seen in the recording.
In recording spontaneous scalp EEG activity Properties of the volume-conducting medium
or the late components of the somatosensory between intracranial generators and scalp elec-
and visual evoked potential that are recorded trodes can have a major effect on the recorded
from scalp electrodes, it is desirable to measure potentials. When the poorly conducting skull
the potentials at each scalp electrode posi- is breached by openings, for example, natu-
tion with respect to a distant, totally inactive rally occurring openings such as the orbits or
reference electrode. In fact, it is not possi- external auditory meatus or iatrogenic open-
ble to find an inactive reference. Even if a ings such as craniotomy defects, long current
physically distant reference position were cho- paths through the opening may cause appre-
sen, as on a limb, volume conduction between ciable electric potentials to be recorded in
Figure 3–13. Traveling and stationary evoked potentials. A, Recording electrode sites from upper arm and shoulder with
median nerve stimulation. B, Median nerve traveling action potentials evoked at the wrist and recorded along the arm
and across the shoulder with bipolar (closely spaced electrodes). C, Elbow reference recordings showing fixed latencies
stationary potentials at the shoulder. D, Action potential propagation from a smaller to a larger volume conductor region
(two joined cylinders of unequal diameter but equal conductivities), illustrating the theoretical source of the stationary
potential.
~10 mg Ω input
resistance
EEG EEG
AMP AMP
A B
~5 K Ω
A
ref
Figure 3–14. The futility of using a distant reference for scalp EEG recording. The right arm reference (A) is electrically
equivalent to a neck reference (B), except for a slight additional resistance in series. AMP, amplifier. (From Nunez, P. L.
ed. 1981. Electric fields of the brain: The neurophysics of EEG. New York: Oxford University Press. By permission of the
publisher.)
with higher amplitudes occurring on the side

of the opening. A regional increase in the
thickness of the conducting medium between
intracranial generators and overlying elec-
trodes may lead to a significant focal attenu-
ation of electric activity, as in the case of a
subdural hematoma or a collection of fluid. For
extracranial generators such as the eyes, which
have a constant electric dipole movement that
induces changing electric potentials when they
move, the effect of skull openings is reversed;
that is, the amplitude of potentials caused by
these generators is usually attenuated in the
region of a large skull defect.
Key Points
• There is no totally inactive reference for
EEG; a distant body reference is electri-
Figure 3–15. The effect of skull openings on scalp- cally equivalent to a neck reference.
recorded potentials; the large skull resistivity—80 times • Naturally occurring or surgical skull open-
that of scalp or brain—leads to long current paths through
skull openings that may cause appreciable potentials to be ings affect volume-conducted potentials,
recorded far from the generators. (From Nunez, P. L. ed. causing amplitude asymmetries.
1981. Electric fields of the brain: The neurophysics of EEG.
New York: Oxford University Press. By permission of the
publisher.)
Dipole Source Localization in EEG

areas that are, in fact, far from the genera-
tors (Fig. 3–15). Amplitude asymmetries, that In many clinical neurophysiologic studies, par-
is, differences between homologous regions on ticularly EEG and evoked potential studies,
the opposite side of the head, are the most conducted for clinical or research purposes,
commonly observed effects of skull defects, localization of the generators of a particular
waveform or activity is of paramount impor-

tance. Volume conductor theory, as discussed “SQUID”
in this chapter, provides a way to calculate the measurement
system
surface potential distribution that would result
from a known configuration of intracranial gen-
erators, the forward problem. However, it is
also desirable to have a way to determine the
Gradiometer
type, location, strength, and orientation of all
of the generators of a given scalp surface-
recorded waveform or activity, the inverse
problem. Unfortunately, for any given potential
distribution recorded from the scalp surface,
the number of possible configurations of gen-
erators that could equally produce that distri- Figure 3–16. A magnetoencephalographic recording
device consists of a magnetic field gradiometer (two coils
bution well is infinite; in general, the inverse with opposite polarities connected in series) and a SQUID
problem does not have a unique solution. If amplifier. (From Nunez, P. L. ed. 1981. Electric fields of
the problem can be constrained by indepen- the brain: The neurophysics of EEG. New York: Oxford
dent anatomical or physiologic data, however, University Press. By permission of the publisher.)
then a solution to the inverse problem may
be possible.5 For example, if there is reason to record the magnetic equivalent of the EEG
to believe that a particular EEG waveform or evoked potentials known as the magnetoen-
or evoked potential peak is generated by a cephalogram (MEG) or the evoked magnetic
localized intracranial source that may be well fields (EMFs). This technique has been applied
represented as a single electric dipole, then a to research as well as clinical applications,
model of the volume conductor properties of and has allowed intracranial generators to be
the head, such as the three-sphere model dis- localized accurately and noninvasively.8–11 Clin-
cussed above under Multiplanar and Multiple ical applications include localizing the seizure
Sphere Models (see Appendix) can be coupled focus by determining the generators of inter-
with an appropriate mathematical algorithm ictal spikes and focal slow waves in patients
to find the location, orientation, and strength with intractable partial epilepsy who are being
of the single dipole whose predicted scalp- considered for epilepsy surgery and using
potential distribution best fits the observed somatosensory magnetic evoked fields to accu-
scalp potential.6, 7 rately locate the somatosensory cortex as part
The inherent uncertainties in the geom- of the planning process for a surgical proce-
etry and electric properties of the volume dure, for example, resection of a brain tumor.
conductor limit the accuracy with which dipole Middle- and long-latency auditory magnetic
localization based on scalp-recorded poten- evoked fields have also been used to accurately
tials can be performed. A new technique for locate auditory and speech reception cortical
improving source localization derives from the regions and to noninvasively lateralize cortical
magnetic field that any electrical current gen- language function.
erates. Thus, intracranial current sources gen-
erate magnetic fields that, with appropriate
Key Points
sensing devices, may be detected at various
locations outside the head. These magnetic • Finding the location, strength, and orien-
fields, unlike scalp-recorded potentials, are tation of all generators of scalp-recorded
unaffected by the intervening medium, and waveforms is the “inverse problem.”
calculations of source locations from magnetic • The “inverse problem” has no unique
field maps thus may be performed without solution for an arbitrary number and con-
a need for complex and possibly inaccurate figuration of generators.
volume conduction models. With a sensitive • Attempts to solve the “inverse problem”
magnetic detector, for example, a magnetome- often depend on assuming there is just one
ter or magnetic gradiometer (Fig. 3–16), and (or a few) dipole generators.
a special type of high-gain, low-noise ampli- • Magnetoencephalographic recordings may
fier, specifically, a superconducting quantum be more suitable than EEG for solving the
interference device, or SQUID, it is possible “inverse problem.”
Nerve Conduction Study atrophy of the muscle being recorded from,

Applications where some stimulation of a neighboring nerve
produces a recordable response. An example of
Volume conduction has an important effect this is in severe carpal tunnel syndrome, when
on the responses recorded during nerve con- the ulnar nerve is inadvertently stimulated and
duction studies. In motor conduction studies, the response from ulnar muscles are readily
the typical recording montage of G1 over the recorded over the thenar eminence.
motor endplate and G2 over the muscle ten- Other morphologic features of the wave-
don allows for recording of an initially nega- forms also result from volume conduction. As
tive (upward deflection from baseline) wave- mentioned previously, the G2 electrode is not
form with a maximal amplitude. A biphasic, truly “inactive.” The G2 electrode, even when
negative-positive waveform is recorded. When placed over the digit, contributes to the com-
a triphasic waveform with an initial positivity is pound muscle action potential, particularly in
recorded during motor nerve conduction stud- the ulnar and tibial motor nerve conduction
ies, the initial positive waveform component is studies.
generated from muscle fiber action potentials Sensory nerve conduction studies may pro-
that have an endplate region at some distance duce biphasic (negative-positive) or triphasic
away from the G1 recording electrode. These (positive-negative-positive) waveforms. Tripha-
distant muscle fiber action potentials propagate sic (initially positive) waveforms are com-
through the conducting medium to the record- mon with sensory nerve conductions since,
ing electrode and produce a positive deflec- in contrast to motor NCS where the record-
tion. This initial positivity may be the result ing electrodes are placed immediately over
of the G1 electrode placed at the wrong site the site of action potential generation, the
(off of the endplate zone) on the muscle being action potential is always generated at a dis-
recorded from. In this case, the initial positiv- tance from the G1 recording electrode and
ity will be present with stimulation at any site propagates toward and then away from the
along the nerve (Fig. 3–17). electrodes.
Alternatively, the initial positivity may be
generated from a muscle distant from that
being recorded, either from a physiologic phe- Key Points
nomenon or due to overstimulation during the
nerve conduction study (NCS) with stimulus • When a triphasic waveform with an ini-
spread to another nerve. The latter situation tial positivity is recorded during motor
is commonly seen with peroneal motor con- nerve conduction studies, the initial pos-
duction studies, recording from the extensor itive waveform component is generated
digitorum brevis, in which stimulation at the from muscle fiber action potentials that
knee causes depolarization of the anterior com- have an endplate region at some distance
partment muscles of the leg (such as the ante- away from the G1 recording electrode.
rior tibialis), which is recorded from the distant • An initial positivity on motor NCS may be
electrodes over the foot (Fig. 3–9). It is also generated from a muscle distant from that
often seen in situations where there is severe being recorded, either from a physiologic
phenomenon or due to overstimulation
during the NCS with stimulus spread to
another nerve.
• The G2 electrode contributes to the com-
pound muscle action potential, particu-
larly in the ulnar and tibial motor nerve
conduction studies.
• Triphasic, initially positive waveforms are
common with sensory nerve conductions,
since the action potential is always gener-
Figure 3–17. Initial positivity recorded on a median ated at a distance from the G1 recording
motor conduction study due to active electrode location off electrode and propagates toward and then
the site of generation (end plate region). away from the electrodes.
Needle EMG Applications travel past a stationary recording elec-

trode.
Volume conduction plays a role in the • With a concentric needle electrode, the
recording of EMG waveforms. The source “active” and “reference” recording sites
of all EMG waveforms are the action poten- are very close to each other and, as a
tials of individual muscle fibers. As previ- result, the recorded difference in current
ously discussed, these action potentials can that is generated at a distance from the
be considered quadrupoles that propagate electrode (from distant fibers) is minimal.
along the muscle fiber. These action potentials • Motor unit potentials recorded with a
are recorded as a triphasic (positive-negative- monopolar needle electrode, in which the
positive) potential as the current sources of “active” recording site is much closer to
the potential travel past a stationary recording the generator compared to the “refer-
electrode. Since the spatial gradient and cur- ence” electrode, are of higher amplitude
rent density are higher nearer the generator, a and longer duration than those recorded
triphasic muscle fiber action potential will have with a concentric needle electrode.
(1) a higher amplitude due to the recording of
higher density of current charge, (2) a more
rapid rise time in the peak of the potentials
due to the higher density of equipotential lines SUMMARY
and (3) more rapid rate of change in the cur-
rent density as the propagating action potential This chapter reviews the principles of vol-
passes, with proximity of the recording needle ume conduction as applied to the potentials
electrode to the generator (Fig. 3–11). recorded in clinical neurophysiologic studies.
The volume conduction concepts also help Knowledge of these principles is necessary for
to explain how the motor unit potentials differ proper interpretation of EEG, EMG, NCS,
in morphology when recorded with a concen- SEP, VEP, and BAEP recordings in order to
tric needle electrode compared to a monopolar extract information concerning the function
electrode. With a concentric needle electrode, and location of the neural structures that gen-
the “active” and “reference” recording sites are erate the recorded activity or waveforms.
very close to each other and, as a result, cur-
rent that is generated at a distance from the
electrode (such as action potentials of distant
muscle fibers in the motor unit) is recorded at REFERENCES
the point where the spatial gradient is low and 1. Niedermeyer, E., and F. Lopes da Silva, eds.
therefore the difference in current recorded 1993. Electroencephalography: Basic principles, clin-
at each site is minimal, essentially canceling ical applications, and related fields, 3rd ed. Baltimore:
out a recorded potential. In contrast, recording Williams & Wilkins.
the same motor unit potential recorded with 2. Stein, R. B. ed. 1980. Nerve and muscle: Membranes,
cells, and systems, 65–86. New York: Plenum Press.
a monopolar needle electrode, in which the 3. Dumitru, D., and J. A. DeLisa. 1991. AAEM mini-
“active” recording site is much closer to the monography #10: Volume conduction. Muscle Nerve
generator compared to the “reference” elec- 14:605–24.
trode (which is usually placed at a distance 4. Stegeman, D. F, A. van Oosterom, and E. J. Colon.
1987. Far-field evoked potential components induced
on the skin), will show a greater difference in by a propagating generator: Computational evidence.
the electrical potentials generated from distant Electroencephalography and Clinical Neurophysiol-
muscle fibers. This results in higher amplitude ogy 67:176–87.
and longer duration motor unit potentials when 5. van Oosterom, A. 1991. History and evolution of meth-
recorded with a monopolar electrode. ods for solving the inverse problem. Journal of Clinical
Neurophysiology 8:371–80.
6. Scherg, M. 1992. Functional imaging and localization
Key Points of electromagnetic brain activity. Brain Topography
5:103–11.
• Muscle fiber action potentials recorded 7. Wong, P. K. 1991. Source modelling of the rolandic
focus. Brain Topography 4:105–12.
during needle EMG are recorded as a 8. Cohen, D., and B. N. Cuffin. 1991. EEG versus MEG
triphasic (positive-negative-positive) poten- localization accuracy: Theory and experiment. Brain
tial as the current sources of the potential Topography 4:95–103.
9. Stok, C. J., H. J. Spekreijse, M. J. Peters, H. B. Boom,

and F. H. Lopes da Silva. 1990. A comparative
EEG/MEG equivalent dipole study of the pattern
onset visual response. Electroencephalography and
Clinical Neurophysiology Suppl. no. 41:34–50.
10. Papanicolaou, A. C., P. G. Simos, J. I. Breier,
et al. 1999. Magnetoencephalographic mapping of
the language-specific cortex. Journal of Neurosurgery
90:85–93.
11. Nakasato, N., and T. Yoshimoto. 2000. Somatosensory,
auditory, and visual evoked magnetic fields in patients
with brain diseases. Journal of Clinical Neurop-
hysiology 17:201–11.
12. Lorente de No R. 1947. Potential fields in homogenous
conductors. New York: Rockefeller Press.
APPENDIX
Figure 3–18. A current dipole. A, Coordinate system
showing definition of r and d. B, Lines of current flow
Calculating Potentials in Infinite (solid) and equipotential lines (dashed) in a volume con-
Homogeneous Media ductor. (From Nunez, P. L. ed. 1981. Electric fields of
the brain: The neurophysics of EEG. New York: Oxford
University Press. By permission of the publisher.)
The simplest form of a volume conductor is a homoge-
neous medium without boundaries in which generators
and recording electrodes are embedded. In this situa- potential generated by an action potential propagating
tion, the recorded potential can be calculated from the along an axon: the axonal membrane has a negative polar-
configuration of the source currents. ity outside and a positive polarity inside at the peak of the
action potential. However, on either side of this peak, the
membrane is positive outside and negative inside. Nerve
MONOPOLE, DIPOLE, AND action potentials have positive sources ahead of and behind
QUADRUPOLE SOURCES the depolarization, and a central negative sink (the area of
nerve depolarization).
A single source, or sink, of current is referred to as a
monopole. The potential relative to a distant reference at
distance r from a monopole source in an infinite homo-
geneous medium of conductivity σ (resistivity ρ = I/σ) is Potentials in Nonhomogeneous
given by
Media
I
V= In contrast to the volume conduction in a simple, homo-
4π σ r geneous conducting medium, when a monopolar source is
located in two hemi-infinite regions of differing conduc-
The potential relative to a distant reference measured at tivity with a planar interface between them, the lines of
distance r from a dipole source in an infinite homogeneous current flow change at the interface. This occurs because
medium of conductivity s is given by V = Id(cos θ )/4 σ π r2 , the current density (current per unit area) flowing in the
where I is the magnitude of the dipole current source, direction parallel to the interface is less in the region of
d is the pole separation, and θ is the angle between the higher conductivity. Consequently, if the source is located
dipole axis and the line from the dipole to the measure- in the region of lower conductivity, the lines of current
ment point (Fig. 3–18A); this formula is valid only for flow bend outward as they enter the region of higher con-
r >> d. Thus, the potential of a dipole falls off inversely ductivity. If the source is located in the region of higher
with the square of the distance from the source. The conductivity, the lines of current flow bend inward as they
lines of current flow around a dipole form curved paths enter the region of lower conductivity.
(Fig. 3–5B). The equipotential surfaces are perpendicular There are a number of inhomogeneities in the body,
to the lines of current flow and have a figure-8 configu- such as interfaces between limbs or body regions and inter-
ration around the dipole.2 The zero potential surface is a faces between the body and the external environment.
plane halfway between the two poles of the dipole, because Sources located a short distance under the skin surface, for
on this plane θ = 90◦ and cos θ = 0 (Fig. 3–18B). example, superficial nerve action potentials or EEG activ-
Two adjacent current dipoles of opposite orientation ity from cortical sources 2–3 cm deep, can be approximated
placed end-to-end constitute a current quadrupole. The as a plane with a volume conductor of high conductivity on
potential of a quadrupole falls off inversely with the cube one side of the region and of air on the other side with
of the distance from the source, and the equipotential essentially zero conductivity on the other side. No current
surfaces around the quadrupole have a “cloverleaf” con- flow penetrates from the high conductivity region to the
figuration. A quadrupole is a fair approximation of the zero conductivity region, and the lines of current flow are
completely reflected at the interface. Consequently, poten- brain, cerebrospinal fluid (CSF), skull, and scalp. Both
tials measured at the interface, that is, surface-recorded multiplanar and multiple sphere models have been used
potentials, caused by underlying generators are twice as to investigate the effects of these regions.
large as they would be for the same generators in an infi- For dipole sources located in the cerebral cortex and
nite homogeneous medium. Some extracellular currents for subdural recording electrodes, a model using two pla-
from a generator in a volume conductor reach the skin sur- nar interfaces (brain–CSF and CSF–skull) can be used.
face, where the current causes a potential drop across the This model predicts that the measured potentials would be
space between two electrodes (Ohm’s law). This potential approximately equal to those that would be recorded in an
difference can be measured by a recording system with a infinite homogeneous medium. For cortical dipole sources
differential amplifier. with scalp surface-recording electrodes, a model using five
regions (brain, CSF, skull, scalp, and air) and four planar
interfaces predicts that the measured potentials would be
approximately equal to one-fourth of those that would be
Homogeneous Sphere Model recorded in an infinite homogeneous medium. This may
be compared with the factor-of-2 augmentation of poten-
For sources located in the head, such as cortical and sub- tials predicted by the single planar interface model; the
cortical generators of EEG and evoked potentials, a spheri- predicted relative attenuation (by a factor of 8) is caused
cal volume conductor model is a reasonable approximation mainly by the effects of the poorly conducting skull, whose
to the actual geometry. The simplest model assumes a uni- conductivity is only about 1/80 that of brain or scalp. The
form conductivity within a sphere with a dipole source effect of the skull may be diminished markedly in subjects
located at the center of the sphere. At points near the cen- who have a skull defect, for example, because of previous
ter of the sphere the potential is the same as that expected surgery. The EEG activity in the vicinity of such a defect
in an infinite homogeneous medium, but at the surface of may be several times greater in amplitude than the EEG
the sphere the lines of current flow are confined to the activity in surrounding regions where the skull is intact.
spherical volume and the current density is greater. Thus, For deep dipole sources such as BAEP generators,
scalp surface recordings are three times greater in ampli- a multiple sphere model with four regions (brain, skull,
tude than intracerebral recordings. Brain stem generator of scalp, and air), three spherical interfaces, and a dipole in
short latency auditory evoked potential peaks are amplified the center is appropriate. For scalp surface-recording elec-
in that manner. trodes, this model predicts that the measured potentials
would be approximately equal to twice those that would be
recorded in an infinite homogeneous medium. This may be
compared with the factor-of-3 augmentation of potentials
Multiplanar and Multiple Sphere predicted by the homogeneous sphere model; the pre-
Models dicted relative attenuation (by a factor of 2/3) caused by
the poorly conducting skull is not nearly as great for deep
An air–body interface surface is only one of the inho- sources as it is for superficial, or cortical, generators.12
mogeneities that affects volume conduction. For EEG The effect of skull defects on potentials from deep
and scalp-recorded evoked potentials, the other inhomo- sources is correspondingly less than that from superficial
geneities of importance are the differing conductivities of sources.
Chapter 4
Digital Signal Processing

DIGITAL COMPUTERS IN Repetitive Transient Waveforms

CLINICAL NEUROPHYSIOLOGY Movement-Associated Potentials
Introduction
DIGITAL FILTERING
Digital Clinical Neurophysiology
Types of Digital Filters
Digital Electroencephalography
Characteristics of Digital Filters
Construction Of Digital Systems
TIME AND FREQUENCY
DIGITIZATION
DOMAIN ANALYSIS
Principles
Interval Analysis
Analog-to-Digital Conversion
Autocorrelation Analysis
COMMON USES OF DIGITAL Fourier (Spectral) Analysis
PROCESSING Statistical Analysis
AVERAGING Pattern Recognition
Evoked Potentials and Nerve
SUMMARY
Conduction Studies
DIGITAL COMPUTERS IN is not readily obtainable with visual analysis

alone. Furthermore, they are well suited for
CLINICAL NEUROPHYSIOLOGY quantification of key features of waveforms.
This may be useful in accurate clinical diag-
Introduction nosis of electroencephalographic (EEG), elec-
tromyographic (EMG), and evoked potential
Digital computers can perform types of signal studies, and it also lends itself to serial com-
processing not readily available with ana- parisons between studies performed on the
log devices such as ordinary electric cir- same subject at different times or between
cuits. Because of their large storage capaci- two groups of subjects in scientific investi-
ties and rapid, random-access retrieval, they gations. Digital computers may also partially
can make the process of obtaining, storing, automate the interpretation of clinical neuro-
retrieving, and viewing clinical neurophysiol- physiology studies. This chapter discusses the
ogy data easier. Also, because of their sophis- uses of digital signal processing and storage
ticated computational abilities, they may aid that are common to many types of physiologic
in extracting from waveforms information that studies.
53
Key Points may be more catastrophic with digital sys-

tems, with possible loss of an entire study
• Digital computers can make it easier to
because of system failure.
obtain, store, retrieve, and view clinical • Incompatible data formats—In marked
neurophysiology data. contrast to the relatively standard data
• They can be used to extract information
formats used in the personal computer
not readily obtainable with visual analysis industry that facilitate sharing of data, dig-
alone. ital instruments use data formats that are
• They can quantify key features of wave-
proprietary to each manufacturer, and in
forms. general, studies recorded on the instru-
• They can facilitate serial comparisons
ments of one manufacturer cannot be read
between studies on the same subject at on those of another. This limits the ability
different times. to share studies between laboratories. To
• They can paritially automate the int-
surmount this difficulty, some companies
erpretation of clinical neurophysiology now offer reader programs for personal
studies. computers that are capable of reading
the data formats used by many different
manufacturers, but these programs are an
Digital Clinical Neurophysiology additional expense.
• Obsolescence of data formats—As digi-
In recent years, digital instruments have largely tal systems evolve and new models are
replaced analog instruments. The advantages released, recording formats may change
of digital over the analog recordings that had over time even with the same manufac-
been used for the early work in each of the turer; thus, eventually, it may be impos-
fields of clinical neurophysiology derive from sible to use current systems to review
the unique capabilities of digital recording studies acquired on older instruments.
technology. These capabilities include:1
However, the advantages of digital recording
• Convenient storage and retrieval of records outweigh the disadvantages, and all fields of
• Montage reformatting clinical neurophysiology are moving steadily
• Filter, sensitivity, and time base changes toward digital technology. The greatest impact
• Reliability of interpretation of digital recordings has been in EEG, as dis-
• Rapid location of events and features of cussed in the following section.
interest
• Annotating recordings
• Quantitative analysis of background activ- Digital Electroencephalography
ity and transients
Although the accuracy of visual reproduction
The disadvantages of digital instruments include of the EEG waveforms on digital EEG instru-
the following: ments was limited in the past by the resolu-
tion of the screen display, this limitation has
• Cost—Digital instruments may be more been largely overcome by the ready availabil-
expensive, particularly in the long term, ity of personal computer graphics cards and
because with the rapid evolution of monitors with resolutions of 1280×1024 pixels
computer technology, digital instruments or higher. Furthermore, digital EEG may be
become obsolete more rapidly than their combined with digital video recording for the
analog counterparts did. evaluation of patients with seizures or spells.
• Maintenance—Repair of digital instru- With the combination of EEG and video, the
ments requires more knowledge than is EEG can be correlated with clinical behav-
required for analog machines, and trouble- ior during transient spells or seizures. More-
shooting is more complex. Maintenance over, by recording both the EEG and video
personnel must be knowledgeable about in a digital format, events of interest can be
computers and computer software as well located quickly during prolonged recordings
as hardware. Digital instruments may be and the video can be displayed nearly instanta-
less fault-tolerant, and equipment failures neously, compared to analog video recordings
Digital Signal Processing 55
that require time-consuming tape searches for referential data. In addition to the routine
the segment of interest. Also, digital recording bipolar and referential montages, special
of video significantly facilitates the editing and montages such as a common average ref-
copying of video segments. erence or a laplacian (source) montage
Applications of the unique capabilities of may be used.
digital recording technology are illustrated in • Filter, sensitivity, and time base changes—
the following discussion of digital EEG:1 In a similar fashion, the high- and low-
frequency filters and notch filter, the ver-
• Linear display—In contrast to pen-based tical display scale (sensitivity), and the
analog recordings on paper, in which horizontal display scale (time base) are
the movement of the pen along the selected at the time the EEG is reviewed,
arc of a circle (rather than perpendicu- rather than at the time of recording.
lar to the direction of paper movement) • Reliability of interpretation—A recent
causes a nonlinear distortion of waveform study comparing the accuracy of interpre-
morphology when high-amplitude pen tation of digital vs. analog EEG record-
excursions occur, a digital display accu- ings demonstrated a clear advantage of
rately represents the waveform morphol- digital EEG review,2 which most likely
ogy independently of signal amplitude. is related to the ability to view the
• Convenient storage and retrieval of same EEG segment using several dif-
records—Multiple digital recordings (typ- ferent montages, filters, and sensitivities.
ically hundreds of EEG studies) may be In this study, two experienced board-
kept online for quick retrieval, and larger certified electroencephalographers each
numbers of older recordings (thousands read 89 pediatric EEGs recorded dig-
of studies) may be archived on digital itally. The studies were read either in
media (such as CD-ROM, DVD-ROM, or conventional analog paper format, using
BD-ROM) that require very little storage a digital display but without use of dig-
space and from which they may be readily ital tools such as montage reformatting,
retrieved when needed. This significantly digital filtering, time base or sensitiv-
reduces storage space requirements com- ity adjustment at review time, or using
pared with analog recordings on paper all the features of a digital system. The
and eliminates the need for microfilm- inter-reader agreement (kappa) was calcu-
ing paper recordings. With standard com- lated for each reading condition. Kappa
puter networks, recordings (including dig- values of 0–0.39 represent poor agree-
ital video, when applicable) may be ment, 0.40–0.59 fair agreement, 0.60–0.74
viewed on appropriately configured per- good agreement, and 0.75–1.00 excellent
sonal computers located at sites remote agreement. As shown in Table 4–1, the
from the instruments used for recording inter-reader agreement in classification of
without a need to physically transport the records as normal vs. abnormal and focal
record. Wide area networks allow records vs. nonfocal was best when interpretation
to be accessed at essentially unlimited dis- was done using digital tools.
tance from the recording location, and • Rapid location of events and features
currently available high-speed network of interest—Typical digital EEG instru-
connections to homes allow reading of ments allow rapid paging or scrolling
emergency and after-hours recordings in through the record in the forward or
the reviewer’s home. reverse direction as well as skipping
• Montage reformatting—On digital instru- directly to specific times or specific events
ments, the EEG montage is selected at (marked by the technologist or another
the time the EEG is reviewed, rather person reviewing the EEG).
than at the time of recording. Digital • Annotating recordings—During the recor-
instruments record all data using a ref- ding, technologists may enter textual com-
erential montage with a single common ments about the recording conditions or
reference electrode (such as Cz or an the patient’s behavior; these replace the
average ear reference). All other mon- comments that would be written on a
tages then can be reconstructed by sim- paper record. Also, the reviewer may mark
ple arithmetic operations on the recorded entire “pages” of the EEG record or mark
Table 4–1 Inter-Reader Agreement (kappa) in Classification

of EEG Records
Reading condition Normal vs. abnormal Focal vs. nonfocal
Paper (analog) 0.69 0.46

Digital without reformatting 0.61 0.5
Digital with reformatting 0.81 0.65
individual waveforms or features and label DIGITIZATION

these with text descriptions.
• Quantitative analysis of background activ-
ity and transients—This may include
Principles
interval analysis, autocorrelation analysis,
Electric signals derived from an electrode or
spectral analysis, statistical analysis, and
some other type of transducer may be used
pattern recognition (such as automatic
to represent electric or nonelectric physiologic
spike or seizure detection) as well as
quantities (such as potential in microvolts, cur-
cross-correlation and cross-spectral anal-
rent in milliamperes, pressure in millimeters of
yses, interpolation, topographic displays,
mercury, or oxygen saturation in percentage)
multivariate statistical methods, cortical
in one of two ways. An analog signal takes on
projection techniques, and source local-
any potential (voltage) within a specific range
ization.
(e.g., −3 to 3 V). The potential generally is
directly proportional to the physiologic quan-
tity represented by the signal; therefore, that
Construction Of Digital Systems potential is an analog of the physiologic quan-
tity. Analog signals are generally continuous
A digital (computerized) system for acquisi- in the sense that the potential varies continu-
tion, storage, and display of physiologic wave- ously as a function of time. In contrast, a single
forms has the following key components: digital signal may take on only one of two pos-
sible potentials (e.g., 0 or 3 V); such a signal
may represent one of two possible states (on
• Electrodes or off; yes or no) or one of two possible dig-
• Amplifiers and filters its (0 or 1) and is said to represent one bit
• Analog-to-digital converters (binary digit) of information. Multiple digital
• Solid-state digital memory signals may be used to represent a physiologic
• Digital processor (central processing unit) quantity as a binary number (a series of 0s and
• Magnetic or optical disk (or tape) storage 1s forming a quantity in a base 2 number sys-
• Screen or printer for waveform display tem; that is, the rightmost digit has a value of
20 = 1, the second digit from the right has a
The electrodes, amplifiers, and filters in a digi- value of 21 = 2, the third digit has a value of
tal system are essentially identical to those in an 22 = 4, etc.). Digital signals are discrete and dis-
all-analog system. The amplified signal for each continuous (i.e., they have only two possible
channel is sent to an analog-to-digital converter states), and the nearly instantaneous transition
(ADC), which converts it by the process of from one state to another is made only at spe-
digitization to digital form and stores it in solid- cific times. This is the only format in which
state memory. A digital processor is capable of digital computers can store and process infor-
moving digital data around in memory and pro- mation, and it is most suited to performing
cessing or manipulating it; it may also send data complex and accurate arithmetic operations
to a magnetic or optical disk or tape storage (e.g., adding, subtracting, multiplying, divid-
media for permanent storage, or it may gen- ing) or logical operations (logical conjunction,
erate displays of waveforms and related textual disjunction, negation). Analog representations
annotations on a screen or printer. are more suited for human interpretation; for
example, a waveform display generally uses that can be detected by the ADC and cor-
vertical displacement as an analog to the phys- responds to a change of 1 in the least
iologic quantity, such as the potential being significant bit (20 ). A typical value might
displayed, and horizontal displacement as an be 1 mV (for the amplified signal reaching
analog to elapsed time. the ADC).
• Number of bits in ADC (n)—This deter-
Key Points mines the range of digitized (output) val-
ues. For an ADC that can accept positive
• An analog signal takes on any potential or negative inputs, 1 bit is required for
(voltage); the potential is directly propor- sign (+ or −), and the fractional resolu-
tional to the quantity measured. tion is then 1 part in 2n−1 . A typical value
• Analog signals are continuous (vary con- might be 9–16 bits (corresponding to ±1
tinuously as a function of time). part in 256 to 1 part in 32,768).
• A digital signal takes on only one of two • Input range—This determines the max-
possible potentials and represent one of imum and minimum input potentials.
two possible states or digits 0 or 1. Input potentials above or below the max-
• Digital signals are discrete and discon- imum or minimum are called overflow
tinuous (the transition from one state to or underflow, respectively. A typical value
another is made only at specific times). might be ±2 V.
Analog-to-Digital Conversion
Digitization, or analog-to-digital conversion, is
the process by which analog signals are con-
verted to digital signals. It is the transfor-
mation of continuous potential changes in an
analog signal representing a physiologic quan-
tity to a sequence of discrete digital numbers
(binary integers). Digitization is performed by
a complex circuit known as an analog-to-digital
converter (ADC). There are two aspects to
digitization: quantization and sampling.
QUANTIZATION
Quantization describes the assignment of a
digital number to the instantaneous value of
the potential input to the ADC. A simple exam-
ple is shown in Figure 4–1, which shows a 4-bit
ADC, whose input is an analog signal in the
range 0–16 V, and whose output is a 4-digit
binary number that can take on the values 0,
1, 2, . . ., 15. In this example, any input poten-
tial between 12 and 13 V will result in the same
output (12); thus, the resolution of the ADC
(also known as the quantum size) is 1 V. The
input range of the ADC is 0–16 V.
In general, the following three terms charac- Figure 4–1. Scheme of a 4-bit ADC. Inputs consist of the
terize quantization: continuous signal to be digitized (range 0–16 V) and a start
digitization pulse from a clock that is used to initiate digiti-
zation at appropriate times. Outputs consist of four digital
• Quantum size (ADC resolution)—This signals (+3 or 0 V representing “1” and “0”) that together
determines the minimum potential change can encode a 4-bit integer (range 0–15).
The three quantization parameters are related intervals. The following two terms characterize
by the formula: sampling:
Input Range = ±ADC Resolution × (2n−1 − 1) • Sampling interval—This determines the

temporal resolution of the digitizer. A typ-
where n = number of bits. Note that the input ical value may range from 0.01 ms (for
range of the ADC should match as closely brain stem auditory evoked potentials) to
as possible the expected range of amplified 5 ms or more (for EEG).
potentials. If the range of a signal exceeds • Sampling frequency—This is the recip-
the ADC range, the ADC will either overflow rocal of the sampling interval and is
or underflow and the signal will be distorted measured in hertz (Hz) (s−1 ).
(clipped), but if the range of a signal is too
small compared with the ADC range, much In addition to determining the temporal reso-
of the resolution of the ADC will be wasted lution of the digitizer, the sampling frequency
and the effective resolution may be insuffi- determines the maximum frequency in the
cient, again distorting the signal significantly signal to be digitized that can be adequately
(Fig. 4–2). represented. The sampling theorem (Nyquist
theorem) states that if a signal contains com-
Key Points ponent frequencies ranging from 0 to fN , then
the minimum sampling frequency that can be
• Quantization is the assignment of a digital used for the digitized data to adequately rep-
number to the instantaneous potential of resent the frequency content of the original
the signal. signal is 2fN , where fN is the Nyquist frequency.
• Important parameters of quantization are The Nyquist frequency can be calculated from
quantum size, number of bits, and input the sampling interval as fN = 1/(2 × sampling
range. interval). For example, if fN = 50 Hz, then the
sampling frequency must be at least 100 Hz
SAMPLING (sampling interval of 0.01 second or less). This
sampling frequency is the minimum necessary
In digitization, the conversion of the con- to avoid gross distortion of the input signal;
tinuous analog signal to digital form is usu- a larger sampling frequency (by a factor of
ally performed at discrete equidistant time 3–5) may be necessary in many applications to
Figure 4–2. Effect of quantization parameters, that is, ADC resolution and input range, on the fidelity with which an
analog signal can be represented digitally. In A, the signal exceeds the input range, so that its digital representation (D) is
clipped. In B, the signal uses more than 50% of the input range and is relatively well represented (E). In C, the signal uses
less than 15% of the input range and, because of the limited resolution of the ADC, it is poorly represented (F). (From
Spehlmann, R. 1985. Evoked potential primer: Visual, auditory, and somatosensory evoked potentials in clinical diagnosis,
35–52. Boston: Butterworth Publishers. By permission of the publisher.)
achieve adequate resolution of fine details in For example, if the sampling interval in
the waveforms being digitized. use is 5 ms, the Nyquist frequency is 100 Hz.
Sampling rates for clinical neurophysiologi- A 70-Hz low-pass filter with 6 dB per octave
cal recordings are: slope would attenuate frequencies of 100 Hz
to 0.57 of their original amplitude, which may
EMG–1,000,000 Hz (1.0 μs sampling interval) not be enough. A 50-Hz low-pass filter with
BAER–100,000 Hz (10 μs sampling interval) 12 dB per octave slope would attenuate fre-
EEG–200 Hz (5 ms) quencies of 100 Hz to 0.2 times their original
amplitude, which may be enough to prevent
Key Points significant contamination of the digitized sig-
• Conversion of an analog signal to digital nal by aliased frequency components, provided
that the amplitude of the faster components in
is performed at discrete equidistant time
the original signal is relatively small.
intervals (sampling).
• The Nyquist frequency is the maximum
frequency present in the signal. Key Points
• The sampling theorem states that the min- • Sampling at a frequency lower than twice
imum sampling frequency is twice the the Nyquist frequency produces aliasing
Nyquist frequency. (distortion of the signal).
• Aliasing is avoided by filtering the signal
ALIASING prior to digitization to remove frequencies
above the Nyquist frequency.
Sampling at a frequency lower than 2fN pro-
duces aliasing. Aliasing is distortion of a signal
caused by folding of frequency components in
the signal higher than fN onto lower frequen- COMMON USES OF DIGITAL
cies. For example, a sine wave of 75 Hz, if PROCESSING
sampled at 100 Hz, will appear in the digitized
data as a sine wave of frequency 25 Hz, not One common use of digital signal process-
75 Hz. Aliasing must always be avoided or else ing in clinical neurophysiology is signal aver-
the digitized data will be a gross misrepresen- aging, particularly in evoked potential and
tion of the true signal. In practice, aliasing is sensory nerve conduction studies. Averaging
avoided by filtering the input signal before dig- may also be applied to repetitive transient
itization to remove all frequencies above the waveforms and event-related potentials (such
Nyquist frequency (Fig. 4–3). as movement-associated potentials). A second
Figure 4–3. Effect of sampling interval and aliasing on the fidelity with which an analog signal can be represented dig-
itally. In A, the sampling frequency is 14 times that of the signal frequency and the signal is well represented (D). In B,
the sampling frequency is only six times the signal frequency, and the representation is less accurate but still acceptable
(E). In C, the sampling frequency is only 1.5 times the signal frequency, and thus less than the Nyquist frequency; the
consequent aliasing causes the digital representation (F) to be entirely misleading in that it appears to have a frequency
that is approximately half the true frequency. (From Spehlmann, R. 1985. Evoked potential primer: Visual, auditory, and
somatosensory evoked potentials in clinical diagnosis, 44. Boston: Butterworth Publishers. By permission of the publisher.)
major use of digital signal processing is for dig- signal averaged, although for different types
ital filtering. Less common but still important of studies the epoch length for averaging dif-
uses are in time–frequency analysis, includ- fers significantly. Epoch lengths of 200–500 ms
ing interval and Fourier (spectral) analysis, are typical for visual and long-latency auditory
autocorrelation analysis, statistical analysis, and evoked potentials. Epoch lengths of 30–100 ms
automated pattern recognition. Other uses are typical for middle-latency auditory evoked
tend to be more specialized to particular types potentials and for nerve conduction studies.
of clinical neurophysiologic studies; some of Epoch lengths of 10–20 ms are typical for brain
these are discussed elsewhere in this book. stem auditory evoked potentials and electro-
cochleograms.
Key Points The basic operation of an averager is shown
• Signal averaging is performed in evoked in Figure 4–4. After each stimulus, the input
potential studies and averaging of repeti- signal is digitized at several discrete sam-
tive transient waveforms. pling times within a fixed-length epoch that
• Digital filtering can be useful for many begins at the time of the stimulus. Digitized
types of clinical neurophysiology studies. values of potential at each discrete sample
• Spectral analysis, autocorrelation, statisti- time, each characterized by its latency (time
cal analysis, and pattern recognition are after the stimulus), are averaged for many
other uses of digital processing. stimuli; the resulting averaged signal may be
displayed on a screen or printed on paper.
The stimulus-dependent portions of the signal
(the evoked potential or nerve action poten-
AVERAGING tial) are similar in amplitude and latency in
each epoch averaged and appear in the aver-
Evoked Potentials and Nerve aged result, whereas the stimulus-independent
Conduction Studies (random) portions of the signal (noise and
background neuronal activity among others)
Digital averaging devices for nerve conduc- differ substantially from epoch to epoch and
tion studies and evoked potentials are used are suppressed by averaging. The suppression
routinely in clinical neurophysiology. Their factor, which often is called the signal-to-noise
√
function is similar regardless of the type of ratio, for truly random signals is n, where
Analog Digital
Stimulus
Response 1
A/D
Conversion
Response 2
Bins Epoch
R1 + R2
0 100 200 ms
2
D/A
Conversion
R1 + R2... +Rn
n
Figure 4–4. Operation of an averager. Analog signals recorded after each stimulus are digitized by an ADC during a fixed-
length time window, or epoch that begins at the time of the stimulus. The resulting digital representations are totaled and
divided by the number of epochs averaged. The digital result can be displayed by an analog device such as an oscilloscope
after conversion from digital to analog form. (From Spehlmann, R. 1985. Evoked potential primer: Visual, auditory, and
somatosensory evoked potentials in clinical diagnosis, 37. Boston: Butterworth Publishers. By permission of the publisher.)
n is the number of epochs averaged.3 For may be 50 μV or more, requiring a signal-to-

example, achieving a signal-to-noise ratio of 20 noise ratio of 100 (10,000 epochs averaged).
requires averaging 400 epochs. The required In contrast, sensory nerve action potentials are
signal-to-noise ratio and, hence, the number typically 10 μV or more in amplitude, with
of epochs depend on the type of signal being noise that is comparable, requiring a signal-to-
averaged and the amount of background activ- noise ratio of only 2–3 (4–9 epochs averaged).
ity or noise. For example, typical brain stem It is important to remember that there is a limit
auditory evoked potentials are about 0.5 μV in to the degree of improvement of the recorded
amplitude, whereas background EEG activity waveform with averaging (Fig. 4–5).
Figure 4–5. Averaged median sensory nerve action potential (SNAP) at three levels with increasing sample number (note
comparison of input signals with actual potential). A, Input signal to be averaged and the averaged SNAPs at the wrist
and elbow. B, Increasing level of contraction obscures the forearm waveform with low level contraction and makes it
unrecognizable at moderate levels, despite 6000 averages. Even larger numbers of averages at moderate contraction, did not
bring out the forearm waveform. C, The patient is flexing the elbow at a low level of activation showing improvement in
averages in the forearm and elbow with increasing numbers of averages.
A number of methods of averaging will • The reference time defining the epoch is
enhance the quality and reliability of the usually the peak of the waveform deter-
recording: mined manually or by software.
• Multiple runs to assure reproducibility of

waveforms Movement-Associated Potentials
• Filtering before averaging
• Artifact rejection by amplitude criteria
Movement-associated potentials—one class of
• Preamplifier block by extremely high volt-
event-related potentials—are a cerebral activ-
ages such as cautery in surgery ity associated with, and generally preceding,
• Stimulator trigger by known artifact
a movement (voluntary or involuntary). They
source, such as ECG are obtained by simultaneously recording sev-
eral EEG channels and one or more EMG
Key Points
channels—the latter to determine the time
• Epoch length for averaging differs among of occurrence and other characteristics of the
different types of evoked potentials and movement. An EMG channel may act as a
nerve conduction studies. “trigger” for the averager, but the epoch for
• Digitized values of potential at different averaging usually begins before the onset of the
times (latencies) after the stimulus are muscle activity as recorded by the EMG chan-
averaged for many stimuli. nel and may extend up to or beyond the time of
• Stimulus-dependent portions of the signal the muscle activity. Hence, this type of averag-
appear in the averaged result; stimulus- ing is often called back averaging. It requires
independent portions (noise) do not. somewhat more sophisticated processing than
• The degree of noise suppression depends ordinary forward averaging, because the signal
on the number of stimuli (epochs) being averaged must be digitized continuously
averaged. and stored in memory so that when a “trigger”
occurs, digitized data for the previous 0.5–1
second may be included in the average.
Repetitive Transient Waveforms
Key Points
Repetitive transient waveforms that are not
• Cerebral activity associated with a move-
stimulus-related may also be averaged, such as
epileptic spikes in an EEG or iterative EMG ment (movement-associated potentials) is
discharges. The epoch for averaging in this obtained by averaging EEG.
• An EMG channel may act as a “trigger” for
case is a time “window” around the waveform;
for example, from a specified time before to the averager.
• The epoch for averaging begins before the
a specified time after the peak of the wave-
form. The waveforms to be averaged and the muscle activity as recorded by the EMG
reference times defining the “windows” around channel (back averaging).
them may be determined manually by posi-
tioning a cursor over the peak of each suc-
cessive waveform to be averaged or automat- DIGITAL FILTERING
ically using sophisticated transient detection
programs capable of identifying all waveforms
of interest and locating their peaks and onsets.
Types of Digital Filters
After the epochs have been defined, averag-
A digital filter is a computer program or
ing proceeds in the same way as for evoked
algorithm that can remove unwanted frequency
potentials.
components from a signal.4 Just as for analog
Key Points filters, they may be classified as low-pass, high-
pass, band-pass, or notch filters. Most digital
• Averaging of repetitive transient wave- filters function by forming a linear combina-
forms use an epoch representing a time tion (weighted average) of signal amplitudes at
“window” around the waveform. the current time and various past times. The
two types of commonly used digital filters are • They need not introduce any time delay
the finite impulse response (FIR) filter and the (phase shift) in the signal, as invari-
infinite impulse response (IIR) filter. The FIR ably happens with ordinary analog filters;
filter output is a linear combination only of the thus, time relationships between differ-
input signal at the current time and past times. ent channels can be preserved even if
This type of filter has a property such that its different filters are used for each.5
output necessarily becomes zero within a finite
amount of time after the input signal goes to An example of a segment of EEG contami-
zero. The IIR filter output is a linear combina- nated by muscle artifact as it appears before
tion of both the input signal at the current time and after application of a digital filter is shown
and past times (feed-forward data flow) and in Figure 4–6.
the output signal at past times (feedback data
flow). This type of filter has the property that its
output may persist indefinitely in the absence TIME AND FREQUENCY
of any further input, because the output sig-
nal itself is fed back into the filter. IIR filters DOMAIN ANALYSIS
can be unstable and also have the undesirable
property of noise buildup, because noise terms Interval Analysis
created by arithmetic round-off errors are fed
back into the filter and amplified. For these Interval analysis is a method of determining
reasons, FIR filters are easier to design. How- the frequency or repetition rate of waveforms,
ever, IIR filters often require less computation which is similar to what is done by visual
than FIR for comparable sharpness in their fre- inspection. It is based on measuring the distri-
quency responses and, hence, are often used bution of intervals between either zero or other
for filtering signals in “real time.” level crossings or between maxima and minima
of a signal.6 A zero crossing occurs when the
potential in a channel changes from positive
Key Points to negative or vice versa (Fig. 4–7A). A level
• Digital filters include low-pass, high-pass, crossing (used less often than a zero crossing)
band-pass, or notch. occurs when the potential in a channel changes
• Two common types of digital filters are from greater than to less than a given value
finite impulse response (FIR) and infinite (e.g., 50 μV) or vice versa. The number of zero
impulse response (IIR). crossings or other level crossings per unit time
• FIR filters use a linear combination of the is related to the dominant frequency of the
input signal at multiple time points. signal (Figs. 4–7D and 4–7E). For example, a
• IIR filters use a linear combination of sinusoidal signal that crosses zero 120 times
the input signal and the output (filtered) every second has a frequency of ½ (120) =
signal at multiple time points. 60 Hz.
Key Points
Characteristics of Digital Filters • Interval analysis finds the frequency of
waveforms by measuring the distribution
Digital filters have several characteristics that of intervals between zero crossings.
distinguish them from analog filters. • Level crossings are sometimes used instead
of zero crossings.
• They can be constructed and modified
easily because they are software programs
rather than hardware devices. Autocorrelation Analysis
• They can easily be designed to have rel-
atively sharp frequency cutoffs if desired; Autocorrelation analysis may be used to recog-
for example, much sharper than the typ- nize the dominant rhythmic activity in a signal
ical 6 dB per octave roll-off of an analog and to determine its frequency. It is based
filter. on computing the degree of interdependence
Figure 4–6. Example of digital signal filtering. An electroencephalogram contaminated, A, by scalp EMG artifact was
filtered, B, using a low-pass digital filter. C, The frequency spectra of the unfiltered signals show the large high-frequency
(20 Hz) muscle activity components in the last two channels before filtering (From Gotman, J., J. R. Ives, and P. Gloor.
1981. Frequency content of EEG and EMG at seizure onset: Possibility of removal of EMG artifact by digital filtering.
Electroencephalography and Clinical Neurophysiology 52:626–39. By permission of Elsevier Science Ireland.)
(correlation) between successive values of a the dominant rhythmic component in the

signal. A signal that is truly random, such as signal.
white noise, will have no correlation between
successive values. In contrast, a signal with
rhythmic components has an autocorrelation
significantly different from zero. The auto- Fourier (Spectral) Analysis
correlation function (ACF) is defined as the
correlation between a signal and that same sig- Fourier analysis is the representation of a
nal delayed by time t, expressed as a function periodic function as a Fourier series—a sum
of t. The ACF at t = 0 is always 1, that is, of trigonometric functions, that is, sines and
100% correlation. For a periodic signal (one cosines. A Fourier series may be used to
with rhythmic components), it is an oscillating approximate any periodic function. The greater
function of t with a frequency like that of the the number of terms in the series, the greater
dominant rhythmic component in the original the accuracy of the approximation will be. The
signal (Fig. 4–7C). Fourier transform is a mathematical method to
analyze a periodic function into a sum of a large
number of cosine and sine waves with frequen-
Key Points cies of f, 2f, 3f, 4f, etc., where f is the lowest,
or fundamental, frequency in the function ana-
• Autocorrelation analysis can find the fre- lyzed, and 2f, 3f, etc., are harmonics. The input
quency of the dominant rhythmic activity of the Fourier transform is the periodic func-
in a signal. tion, or signal, to be analyzed. The output is
• It involves computing the correlation of the amplitude of each sine and cosine wave in
successive values of a signal to the same the series.
signal delayed by a time t. In working with discrete data, that is, a sig-
• The autocorrelation is an oscillating func- nal sampled at equal time intervals 0, T, 2T,
tion of t with a frequency like that of 3T, . . . , (N − 1)T, a discrete Fourier transform
Figure 4–7. Examples of several types of signal analysis: A, an EEG signal; B, its power spectrum; C, its autocorrelation
function; D, the distribution density function of intervals between any two successive zero crossings; and E, the distribution
density function of the intervals between successive zero crossings at which the signal changes in the same direction, that is,
from positive to negative or vice versa. (From Lopes da Silva, F. H. 1987. Computerized EEG analysis: A tutorial overview.
In A textbook of clinical neurophysiology, ed. A. M. Halliday, S. R. Butler, and R. Paul, 61–102. Chichester, England: John
Wiley & Sons. By permission of John Wiley & Sons.)
is used. In this case, the fundamental fre- By convention, both the cosine and sine
quency can be shown to be f = 1/NT. The input waves of a given frequency are often lumped
of the discrete Fourier transform is the N dig- together into one quantity describing the
itized values representing the signal at times amount of that frequency present in the signal.
0, T, 2T, . . . , (N − 1)T, where T is the sam- This quantity is called the spectral intensity
pling interval. The output is the amplitudes of or power, and a plot of this as a function of
N/2 − 1 sine waves at frequencies f, 2f, . . . , frequency is the power spectrum (Fig. 4–7B).
(N/2 − 1)f and of N/2 + 1 cosine waves at fre- The intensity, or power, is the square of the
quencies 0, f, 2f, . . ., (N/2)f. Note that there amplitude. The phase (phase angle) for any
is no “data reduction”; the number of input given frequency describes how much of that
values (N) equals the number of output values. frequency is in the form of a cosine wave
and how much is a sine wave. The following Fourth central moment m4 and kurtosis
formulas relate these quantities; here, C is excess β2 = m4 /(m2 )2 (peakedness or flat-
cosine wave amplitude and S is sine wave ness of distribution).
amplitude:
Key Points
• Statistical signal analysis involves finding
Intensity I = A2 = C2 + S2
√ √ the amplitude distribution (number of
Amplitude A = I = C2 + S2 samples having a given amplitude).
• The statistical moments of the distribution
S
Phase φ = Arctan can be calculated.
C
Note : φ = 0◦ for pure cosine wave
φ = 90◦ for pure sine wave Pattern Recognition
Pattern recognition algorithms are designed
Key Points to detect a specific waveform in a signal that
• Fourier (spectral) analysis analyzes a peri- has characteristic features, such as a motor
odic function into a sum of a large number unit potential in an EMG or a sharp wave in
of cosine and sine waves. an EEG. The characteristic features may be
• Applied to discrete (digitized) data, the defined in the time domain (e.g., durations,
Fourier transform gives the frequency slopes, and curvature of waveforms), in the
content (spectrum) of the signal. frequency domain (after filtering signal), or in
• Power (intensity) is the square of signal both. One common approach in developing
amplitude; phase represents the relative a pattern recognition algorithm is as follows:
amount of cosine and sine waves. (1) Define a set of candidate features and a
• The power spectrum is a plot of spectral method to calculate them; (2) Calculate the
intensity against frequency. chosen features for a visually selected collec-
tion of waveforms of the type to be detected,
the learning set, and for a collection of sim-
ilar waveforms determined not to be of the
required type, the controls; and (3) Determine
Statistical Analysis by statistical analysis whether it is possible
to reliably separate the two groups of wave-
Statistical analysis of a digitized signal may be forms on the basis of the calculated features.
a useful data reduction technique. In effect, it For example, one could calculate the rising
treats digitized values at successive time points and falling slope of candidate sharp waves,
as independent values of a random variable. compute these slopes for true epileptic sharp
In this technique, one may plot the amplitude waves and for other transients such as mus-
distribution—the number of digitized samples cle artifacts or nonepileptic sharp transients in
of the signal having a given amplitude value background activity, and determine whether a
vs. the amplitude itself—and visually inspect certain range of slopes characterizes the true
the shape of the distribution. Alternatively, one sharp waves. These techniques have been used
may calculate the moments of the probability with some success to detect spikes and sharp
distribution of signal amplitudes, including the waves, spike-and-wave bursts, sleep spindles
following:6 and K-complexes, and seizure discharges in
EEG recordings7 and to detect motor unit
First central moment, mean voltage m1 (cen- potentials, fibrillation potentials, and other iter-
ter of distribution) ative discharges in EMG recordings.
Second central moment, variance
√ m2 and Key Points
standard deviation σ = m2 2 (width of
distribution) • Pattern recognition algorithms detect
Third central moment m3 and skewness waveforms that have characteristic fea-
β1 = m3 /(m2 )3/2 (asymmetry of distribution) tures.
• Features may be defined in the time REFERENCES

domain, the frequency domain, or
both. 1. Fisch, B. J. 1999. Fisch and Spehlmann’s EEG primer:
• The algorithm is trained on a learning Basic principles of digital and analog EEG, 3rd ed.
Amsterdam: Elsevier.
set of chosen features and controls, then 2. Levy, S. R., A. T. Berg, F. M. Testa, E. J. Novotny Jr.,
applied to arbitrary signals. and K. H. Chiappa. 1998. Comparison of digital and
• Spike and sharp wave detection and conventional EEG interpretation. Journal of Clinical
seizure detection algorithms are examples Neurophysiology 15:476–80.
3. Misulis, K. E. 1994. Spehlmann’s evoked potential
with common clinical use. primer: Visual, auditory, and somatosensory evoked
potentials in clinical diagnosis, 2nd ed., 25–37. Boston:
Butterworth-Heinemann.
4. Maccabee, P. J., and N. F. Hassan. 1992. AAEM
minimonograph #39: Digital filtering: Basic concepts
SUMMARY and application to evoked potentials. Muscle & Nerve
15:865–75.
This chapter reviews the principles of digi- 5. Gotman, J., J. R. Ives, and P. Gloor. 1981. Frequency
tization, the design of digitally based instru- content of EEG and EMG at seizure onset: Possi-
bility of removal of EMG artefact by digital filtering.
ments for clinical neurophysiology, and several Electroencephalography and Clinical Neurophysiology
common uses of digital processing, including 52:626–39.
averaging, digital filtering, and some types of 6. Lopes da Silva, F. H. 1987. Computerized EEG anal-
time-domain and frequency-domain analysis. ysis: A tutorial overview. In A textbook of clinical
neurophysiology, ed. A. M. Halliday, S. R. Butler, and
An understanding of these principles is nec- R. Paul, 61–102. Chichester, England: John Wiley &
essary to select and use digitally based instru- Sons.
ments appropriately and to understand their 7. Gotman J. 1999. Autonomic detection of seizures and
unique features. spikes. Journal of Clinical Neurophysiology 16:130–40.
Chapter 5
Basics of Neurophysiology
Jasper R. Daube and Squire M. Stead
INTRODUCTION Excitability
CELL MEMBRANE Propagation
Transmembrane Ion Gradients Patterns of Activity
Active Transport
SYNAPTIC TRANSMISSION
Equilibrium Potential
Biosynthesis, Storage, Release, and
Ion Channels
Reuptake of Neurochemical
Neuronal Excitability
Transmitters
RESTING POTENTIAL Postsynaptic Effects of
Steady State Neurochemical Transmitters
Sodium Pump Classic Neurotransmission
Role of Extracellular Calcium Neuromodulation
Role of Glial Cells Electrical Synapses
LOCAL POTENTIALS CLINICAL CORRELATIONS
Ionic Basis Pathophysiologic Mechanisms
Characteristics of Local Potentials Energy Failure
Ion Channel Blockade
ACTION POTENTIALS
Threshold SUMMARY
Ionic Basis of Action Potentials
Revision of “Transient Disorders and Neu- activity of excitable cells. The following discus-
rophysiology.” Chapter 5 in Medical Neuro- sion applies to both myocytes and neurons.
sciences. (Ed. E. E. Benarroch, B. F. West-
moreland, J. R. Daube, T. J. Reagan, and B.
A. Sandok.) Lippincott Williams & Wilkins
1999—with permission.
CELL MEMBRANE
Transmembrane Ion Gradients

INTRODUCTION
The plasma membrane is a phospholipid lipid
To become proficient in the practice of clinical bilayer with the polar phosphate (hydrophilic)
neurophysiology, one should have an under- heads abutting the extracellular matrix and
standing of the basic principles underlying the cytoplasm. The nonpolar lipid (hydrophobic)
69
Table 5–1 Relative Ionic Concentrations in Mammalian Neurons

Sodium Potassium Chloride Calcium
Internal concentration Low High Low Low

External concentration High Low High High
Resting permeability Low High Moderate Low
tails constitute the middle of the bilayer. and (2) active, energy adenosine triphosphate
Embedded in the lipid bilayer are protein (ATP)-dependent transport of ions against
macromolecules, including ion channels, lig- their concentration gradient, via ATP-driven
and receptors, and ionic pumps, that are in ion pumps.
contact with both the extracellular fluid and the In the central nervous system, astrocytes
cytoplasm. The lipid bilayer is relatively imper- provide a buffer system to prevent excessive
meable to water soluble molecules, including accumulation of extracellular potassium ions.
ions such as sodium (Na+ ), potassium (K+ ),
chloride (C1− ), and calcium (Ca2+ ). These ions
are involved in electrophysiologic activity and Active Transport
signal transmission (Table 5–1). The concen-
trations of sodium, chloride, and calcium are Nerve and muscle cells obtain energy from glu-
higher extracellularly, and the concentrations cose and oxygen via the glycolytic pathways, the
of potassium and impermeable anions, largely Krebs cycle, and the electron transport system.
protein molecules (A− ), are higher intracel- These pathways provide the energy for normal
lularly (Fig. 5–1). Maintenance of transmem- cell function in the form of ATP. ATP is partly
brane ion concentration depends on the bal- consumed in generating the resting poten-
ance between (1) passive diffusion of ions tial by a mechanism in the membrane, which
across ion channels, or “pores,” of the mem- moves potassium in and sodium out of the
brane, driven by their concentration gradient cell, with slightly more sodium being moved
Vm = –60 to –75 mV
A– (25) E ion (mV)

A– (162)
K+ (3 – 5) K+ (140) E K+ = –100
Na+ (30)
Na+ (150) E Na+ = +40
Cl– (8)
Cl– (130) Ca2+ (10–7 M)
E Cl– = –75
Ca2+ (12) E Ca2+ = +124

2 K+
ATP
ase
3 Na+
Figure 5–1. Transmembrane ion concentrations, equilibrium potential, and resting membrane potential. The semiper-
meable cell membrane determines a differential distribution of ions in the intracellular and extracellular compartments.
Sodium and chloride predominate extracellularly, and potassium and nondiffusible (A− ) ions predominate intracellularly.
Transmembrane ion composition is maintained by the activity of ATP-dependent pumps, particularly sodium/potassium
adenosine tri-phosphatase (ATPase). The different transmembrane concentrations of diffusible ions determine the equilib-
rium potential of each ion (Eion ). The contribution of each ion to the membrane potential depends on the permeability of
the membrane for that particular ion. Increased permeability to an ion brings the membrane potential toward the equilib-
rium potential of that ion. At rest, the membrane is predominantly, but not exclusively, permeable to potassium. There is
continuous leakage of potassium out of the cell and of sodium into the cell, driven by both their concentration gradient and
electrical gradient. The ion gradient is restituted by the activity of sodium/potassium ATPase, which is electrogenic (as it
exchanges two potassium for three sodium ions) and contributes to the maintenance of the resting potential.
Basics of Neurophysiology 71
than potassium. This movement is referred to Electrical pressure is defined by

as active transport, and the system through
which it occurs is the sodium pump. The We = Em × Zi × F
sodium pump moves sodium out of the cell and
potassium into the cell against their concentra- in which We = electrical pressure (work requi-
tion gradients. Chloride moves out of the cell red to move an ion against a voltage); Em =
passively with sodium. absolute membrane potential; Zi = valence
(sign and number of charges on the ion);
F = Faraday (amount of charge per mole of
electrons, in Coulombs).
Equilibrium Potential Diffusion pressure is defined by
The diffusible ions (sodium, potassium, and Wd = R × T × (ln[Ce ]/ ln[Ci ])

chloride, but not calcium) tend to move across
the cell membrane according to their concen- in which Wd = diffusion pressure (work requi-
tration gradients. The molecular motion of ions red to move an ion against a concentration
is a source of energy known as the diffusion gradient); R = universal gas constant (energy
pressure. For example, the intracellular con- per degree per mole); T = absolute tempera-
centration of potassium is 30 times higher than ture; ln = natural logarithm; [Ce ] = extracellu-
the extracellular concentration, [Ko + ]; there- lar ion concentration; [Ci ] = intracellular ion
fore, potassium tends to diffuse from intracel- concentration.
lular to extracellular fluid. The opposite occurs At equilibrium
with sodium. As ions diffuse across the cell
membrane, a separation of charges develops We = Wd
because the nondiffusible negatively charged
intracellular ions (principally proteins) have a and therefore
charge opposite that of the diffusible ions. Two
regions that accumulate different charges have Em = (R × T)/(ln[Ce ]/ln[Ci ])
an electrical potential difference. The volt-
age that develops as a diffusible ion moves By rearrangement, the equilibrium potential is
across the membrane and produces an elec-
trical pressure that opposes the movement of Em = ((R × T)/(Zi × F)) × ln([Ce ]/[Ci ])
the ion. The net ionic movement continues
until the electrical pressure equals the diffu- This equation is known as the Nernst equation
sion pressure. At this time, the system is in and is an important relationship that defines
equilibrium. the equilibrium potential, Em , across the cell
At equilibrium, random ionic movement membrane for any ion in terms of its concen-
continues, but no net movement of ions occurs. tration on the two sides of a membrane. By
The electrical potential that develops across substituting for the constants at room temper-
the membrane at equilibrium is called the ature, converting to a base 10 logarithm, and
equilibrium potential, and this potential is dif- converting to millivolts, we get a useful form of
ferent for each ion. The equilibrium poten- the equation:
tial of an ion (Eion ) is the voltage difference
across the membrane that exactly offsets the Em = (61.5 mV/Zi ) log10 ([Ce ]/[Ci ])
diffusion pressure of an ion to move down its
concentration gradient. Therefore, the equilib- We may use these equations to calculate the
rium potential is proportional to the difference equilibrium potential for any ion if we know the
between the concentration of the ion in the concentrations of that ion on the two sides of
extracellular fluid and the concentration in the the membrane. The Nernst potential assumes
intracellular fluid. An algebraic representation 100% permeability of the selected ion. Lower
of the equilibrium potential can be derived ionic conductances decrease the magnitude of
because the physical determinants of the diffu- this potential. In the resting state, the approx-
sion pressure and electrical pressure expressed imate neuronal equilibrium potentials of the
are known. The final equation is the Nernst major ions are K+ = −100 mV, Na+ = +40 mV,
−
equation. Cl = −75 mV, Ca2+ = + 124 mV (Fig. 5–1).
The contribution of a given ion to the actual equilibrium potential of that ion, and it may
voltage developed across the membrane with increase or decrease, depending on whether
unequal concentrations of that ion depends the membrane potential is above or below the
not only on its concentration gradient but also equilibrium potential.
on the permeability (P) of the membrane to The movements of ions that occur with
that ion. Permeability is the ease with which normal cellular activity are not sufficient
an ion diffuses across the membrane and is a to produce significant concentration changes;
reflection of the probability that the membrane therefore, membrane potential fluctuations are
channel that conducts the ion will open. For normally due to permeability changes caused
example, an ion with a high concentration gra- by channel opening and closing. Increased per-
dient that has very low permeability (e.g., cal- meability (i.e., opening of the channel) to a
cium) does not contribute to the resting mem- particular ion brings the membrane potential
brane potential. If a membrane is permeable toward the equilibrium potential of that ion.
to multiple ions that are present in differing In an electrical model of the membrane, the
concentrations on either side of the membrane, concentration ratios of the different ions are
the resultant membrane potential is a function represented by their respective equilibrium
of the concentrations of each of the ions and of potentials (ENa , EK , ECl ); their ionic permeabil-
their relative permeabilities (Fig. 5–2). ities are represented by their respective con-
The Goldman equation combines these fac- ductances (Table 5–2). The conductance (the
tors for the major ions that influence the mem- reciprocal of the resistance) for a particular ion
brane potential in nerve and muscle cells. Such is the sum of the conductances of all the open
calculations, on the basis of the actual ionic channels permeable to that ion. The move-
concentrations and ionic permeabilities, agree ment of ions across the membrane is expressed
with measurements of these values in living as an ion current. By Ohm’s law, this cur-
cells. These equations also show that a change rent depends on two factors: the conductance
in either ionic permeability or ionic concen- of the ion and the driving force for the ion.
trations can alter membrane potential. If the The driving force is the difference between
concentration gradient of an ion is reduced, the membrane potential and the equilibrium
there will be a lower equilibrium potential for potential of that ion.
that ion. If the resting membrane potential
is determined by the equilibrium potential of
that ion, the resting potential will decrease. Ion Channels
In contrast, if the permeability for an ion
is increased by opening of channels for that Ion channels are intrinsic membrane proteins
ion, the membrane potential will approach the that form hydrophilic pores (aqueous pathways)
Ion Ion Ion

channels pumps concentrations
Membrane Transmembrane
permeability ion gradients
Membrane potential
Figure 5–2. Variables that determine the membrane potential. Transmembrane ion gradients determine the equilibrium
potential of a particular ion. The transmembrane gradients depend on the activity of ATP-driven ion pumps and the buffer-
ing effects of the astrocytes on extracellular fluid composition. Membrane permeability to a particular ion depends on
the opening of specific ion channels. This opening can be triggered by voltage (voltage-gated channels), neurotransmitters
(ligand-gated channels), or intracellular chemicals such as calcium, ATP, or cyclic nucleotides (chemically gated channels).
Increased membrane permeability to a given ion (the opening of the ion channel) brings the membrane potential toward
the equilibrium potential of this ion.
Table 5–2 Characteristics of Different Membrane Potentials

Local potentials
Generator Synaptic Electrotonic

Characteristic potential potential potential Action potential
Graded and + + + −
localized
All-or-none spread − − − +
Active membrane Na+ , K+ Na+ , Ca2+ , K+ , None Na+ , K+ ,
channel Cl− sometimes Ca2+
Initiated by Sensory Neurotransmitter Generator, Electrotonic
stimulus synaptic, or potential
action
potentials
through the lipid bilayer membrane. They nucleotide-gated channels found in many sen-
allow the passive flow of selected ions across sory receptors (e.g., photoreceptors in the
the membrane on the basis of the electro- retina).
chemical gradients of the ion and the physical Mechanoreceptors are activated by mechan-
properties of the ion channel. Most channels ical distortion of the cell membrane and
belong to one of several families of homologous are sometimes referred to as stretch-activated
proteins with great heterogeneity in amino acid channels. Gating stimuli may interact in some
composition. They are defined on the basis channels. For example, the ion permeabil-
of their ion selectivity, conductance, gating, ity of some ligand-gated channels is affected
kinetics, and pharmacology. by membrane voltage or intracellular factors
In general, the transmembrane portion of (or both). Voltage-gated channels are critical
the protein forms the “pore,” and the specific for neuronal function. They control excitabil-
amino acids in the region of the pore deter- ity, spontaneous neuronal activity, generation
mine ion selectivity, conductance, and voltage and conduction of action potentials, and neu-
sensitivity of the channel. Amino acids in the rotransmitter release. Sensitivity to voltage is
extracellular or intracellular portion (or both) due to a voltage sensor at the pore. A region
of the protein channel determine the gating of the pore acts as a selectivity filter, which
mechanism and the kinetics of inactivation. Ion regulates ion permeability according to the
channels vary in their selectivity; some are per- size and molecular structure of the ion. The
meable to cations (sodium, potassium, and cal- range of voltage for activation and the rate
cium) and others to anions (primarily chloride). of activation (opening) and inactivation (clos-
The open state predominates in the resting ing) are important variables in voltage-gated
membrane for a few channels; these are mostly channels.
the potassium channels responsible for the Voltage-gated cation channels are responsi-
resting membrane potential (see below). Most ble for the maintenance of neuronal excitabil-
ion channels are gated; that is, they open in ity, generation of action potentials, and
response to specific stimuli. According to their neurotransmitter release (Table 5–3). They are
gating stimuli, ion channels can be subdivided members of a family of proteins with a com-
into (1) voltage-gated channels, which respond mon basic structure consisting of a principal
to changes in membrane potential; (2) ligand- subunit and one or more auxiliary subunits.
gated channels, which respond to the binding The amino acid composition of a subunit deter-
of a neurotransmitter to the channel molec- mines ion selectivity, voltage sensitivity, and
ular complex; and (3) chemically gated chan- inactivation kinetics of the channel. Voltage-
nels, which respond to intracellular molecules gated sodium channels are critical for the gen-
such as ATP, ions (particularly calcium), and eration and transmission of information in the
cyclic nucleotides. Important examples of nervous system by action potentials. In neu-
chemically gated channels include the cyclic rons, sodium channels are concentrated in the
Table 5–3 Examples of Ion Channels

Ion channel Equilibrium potential Location Function
Voltage-gated
Na+ +35 Nodes of Ranvier Initiation and
Entire unmyelinated conduction of action
axon potential
Axon hillock
K+ −90 Diffuse along Repolarization of
internode action potential
Diffuse in neurons Decrease neuronal
excitability and
discharge
Ca2+ +200 Dendrite Slow depolarization
Soma Burst firing
Oscillatory firing
Axon terminal Neurotransmitter
release
Chemically gated −75 Dendrite Synaptic inhibition
Cl− (GABA) Soma
Cation channel 0 Dendrite Synaptic excitation
(l-glutamate,
acetylcholine)
Note: GABA, γ -aminobutyric acid.
initial segment of the axon (the site of gen- gene expression. Intracellular calcium is also
eration of action potentials) and in the nodes necessary for muscle contraction and glan-
of Ranvier (involved in rapid conduction of dular secretion. These functions depend on
action potentials). In muscle, these channels levels of calcium in the cytosol that are deter-
participate in excitation–contraction coupling. mined by the calcium influx through vari-
There are several varieties of voltage-gated ous channels, release from intracellular stores
calcium channels, and they have different (particularly the sarcoplasmic reticulum), and
distributions, physiology, pharmacology, and counterbalancing active mechanisms of reup-
functions (Table 5–4). Calcium influx occurs take and extrusion.
not only through voltage-gated channels but Large numbers of voltage-gated potassium
also through ligand-gated and cyclic nucleotide- channels determine much of the pattern of
gated channels. Calcium ions are important activity generated by neurons. They are pri-
in the regulation of numerous processes in marily responsible for the resting membrane
neurons, including modulation of neuronal potential, repolarization of the action potential,
firing pattern, neurotransmitter release, sig- and control of the probability of generation
nal transduction, enzyme activation, intracel- of repetitive action potentials. Ligand-gated
lular transport, intermediate metabolism, and channels open in response to the binding
Table 5–4 Voltage-Gated Calcium Channels

Channel type Location Function
T Apical dendrites Modal bursting or tonic firing

L Soma, dendritic shafts, and spines Slow action potentials
Skeletal muscle Excitation–contraction coupling
N Synaptic terminals Neurotransmitter release
P/Q Synaptic terminals Neurotransmitter release
Dendritic shafts and spines Persistent depolarization
Neurotransmitter
Ion+
e
Outsid
Inside
Binding site
Cell
Neurotransmitter
membrane
Ion+
Outside
Inside
Figure 5–3. The plasma membrane consists of a phospholipid bilayer that provides a barrier to the passage of water-
soluble molecules, including ions. Passage of ions across the membrane depends on the presence of transmembrane
proteins, including ion channels and ion pumps. Ion channels provide an aqueous pore for the passage of ions across
the membrane, according to their concentration gradients. The opening of an ion channel, or pore, may be triggered, or
gated, by several stimuli, such as voltage (voltage-gated channel) or neurotransmitters (ligand-gated channel). In the exam-
ple shown here, a neurotransmitter (such as glutamate) binds to a specific ligand-gated cation channel, and this produces
a change in the spatial configuration of the channel protein, allowing the pore to open and the cation to pass through the
membrane. Changes in the amino acid composition of the ion channel protein affects its ion selectivity, gating mechanism,
and kinetics of channel opening (activation) and closing (inactivation).
of neurotransmitters (Fig. 5–3). They include • Ions with greater permeability have a
(1) nonselective cation channels permeable to greater influence on the membrane
sodium, potassium, and, in some cases, cal- potential.
cium; and (2) anion channels permeable to • Opening an ion channel moves the mem-
chloride. These channels are discussed in rela- brane potential toward that ion’s equilib-
tion to synaptic transmission. rium potential.
Key Points
• Permeability of membrane ion channels Neuronal Excitability
and activity of ATP-driven ion pumps
determine ion concentration gradients Neuronal excitability is defined as the abil-
across the membrane. ity of the neuron to generate and transmit
• Ion channels are transmembrane proteins action potentials. It depends on the mem-
that form a pore for the passive movement brane potential, which determines the gating of
of ions. the sodium channels. The membrane potential
• Different ion channels provide selective depends on the transmembrane ion concentra-
permeability to different ions. tion (which determines the equilibrium poten-
• Ion channels are opened by a voltage tial) and ion permeability. Increased perme-
change or neurotransmitter binding. ability to an ion moves the membrane potential
• Most of the rapid communication signals toward the equilibrium potential of that ion.
in the nervous system are handled by In the absence of a stimulus, the membrane
voltage-gated ion channels. potential of the neuron, or resting membrane
• An ion’s equilibrium potential is the elec- potential, is dominated by its permeability to
trical potential that exactly balances the potassium, whose channels are open; there-
opposing concentration gradient-driven fore, this potential varies between −60 and
movement of an ion across the membrane. −80 mV. Because the threshold for opening
• Diffusion of an ion across the mem- voltage-gated sodium channels that are needed
brane (permeability) depends on the ion to trigger and propagate action potentials is
channel. approximately −50 to −55 mV, any change
Table 5–5 Ionic Basis of Local Potentials

Ion Equilibrium Effect of increased Examples of local
potential, mV permeability on potentials
membrane potential
Na+ +40 Depolarization Generator potentials

Excitatory postsynaptic
potential
Ca2+ +200 Depolarization Excitatory postsynaptic
potential
K+ −90 Hyperpolarization Inhibitory postsynaptic
potential
Cl− −75 Hyperpolarization, Inhibitory postsynaptic
depolarization, or no potential
change
of the membrane potential in this direction flow of current across the membrane. Its value
will increase the probability of triggering an determines spontaneous neuronal activity and
action potential. An increase in membrane neuronal activity in response to extrinsic input.
permeability to sodium or calcium increases Because the resting potential is the absolute
excitability, and an increase in permeability difference in potential between the inside and
to potassium or chloride decreases excitability the outside of the cell, it represents trans-
(Table 5–5). membrane polarity. A decrease in the value
of the resting membrane potential means less
negativity inside the cell and the membrane
RESTING POTENTIAL potential moves toward zero; this constitutes
depolarization. When the membrane poten-
The resting potential is the absolute difference tial becomes more negative than the value
in electrical potential between the inside and of the resting potential, the potential moves
the outside of an inactive neuron, axon, or mus- away from zero; this is hyperpolarization. The
cle fiber. If an electrical connection is made resting membrane potential depends on two
between the inside and the outside of a neu- factors:
ron, the cell acts as a battery and an electrical
current will flow. The potential is generally 1. Leak ion channels open at rest with mar-
between −60 and −80 mV, with the inside of kedly different permeabilities to sodium
the cell negative with respect to the outside. and potassium, making the cell mem-
The resting potential can be measured directly brane a semipermeable membrane.
by using a microelectrode. The tip of such an 2. Energy-dependent pumps, particularly
electrode must be less than 1 μm in diameter the sodium/potassium pump.
to be inserted into a nerve or muscle cell. By
connecting the microelectrode to an appropri- At rest, there is a continuous “leak” of potas-
ate amplifier, the membrane potential can be sium outward and of sodium inward across the
recorded and displayed. The machine registers membrane. Cells at rest have a permeability
the potential difference between the two elec- to sodium ions that ranges from 1% to 10%
trical inputs, which is displayed as a vertical of their permeability to potassium. Thus, in
deflection of a spot of light that moves contin- the absence of synaptic activity, the membrane
uously from left to right across the screen. A potential is dominated by its high permeabil-
negative membrane potential is registered as a ity to potassium, and the membrane potential
downward deflection; thus, when a microelec- is drawn toward the equilibrium potential of
trode enters a neuron or muscle fiber, the oscil- this ion (90 mV). However, the membrane at
loscope beam moves down to a new position. rest is also permeable to sodium and chloride,
The resting membrane potential is the trans- so that the membrane potential is also pulled
membrane voltage at which there is no net toward the equilibrium potential of these ions.
The resting potential varies among different more open and potassium conductance is
types of neurons, but it is typically −60 to much higher than that of other ions. Therefore,
−80 mV. The continuous leaking of potassium potassium is the largest source of separation
outward and sodium inward is balanced by the of positive and negative charges (voltage) as
activity of the sodium/potassium pump. it diffuses out and leaves the large anions
behind. This is illustrated schematically in
Figure 5–4. Small amounts of sodium entering
Steady State the cells, driven by both electrical and chem-
ical forces, tend to depolarize the membrane.
Potassium diffuses through the membrane As a result, potassium is no longer in equilib-
most readily because potassium channels are rium and leaves the cell. Thus, the cell is not in
Voltmeter
Semipermeable
membrane
Protein salt
solution
Salt water
Inside Outside
A Initial state
0
– +
B Redistribution by
diffusion alone
0
– +
C Redistribution by voltage
(charge separation)
0
– +
Figure 5–4. A theoretical model of the generation of a membrane potential by diffusion across a semipermeable mem-
brane. A, Equal amounts of anions and cations are dissolved on each side of the membrane, producing no voltage gradient.
The membrane is permeable to all ions except large anions (A− ). B, K+ , Na+ , and Cl− redistribute themselves solely by
diffusion; this results in a charge separation, with greater negativity inside. C, Electrical pressure due to charge separation
and diffusion pressure due to concentration differences are balanced at the resting membrane potential.
equilibrium but in a steady state dependent on Role of Glial Cells

metabolic energy. In this steady state, the small
outward potassium leak must be exactly equal Astrocytes are important in controlling the
in magnitude to the rate at which potassium is extracellular concentration of potassium. Astro-
transported into the cell. The same is true also cytes are highly permeable to potassium and
for sodium. In this condition, the net move- are interconnected with each other by gap
ment of each ion across the membrane is zero, junctions. When the extracellular concentra-
an exact description of the resting membrane tion of potassium increases because of neu-
potential. ronal activity, astrocytes incorporate potassium
and transfer it from one cell to another through
gap junctions. This prevents the extracellular
accumulation of potassium and maintains neu-
Sodium Pump ronal excitability. This is referred to as spatial
buffering of extracellular potassium.
The sodium pump (Na/K ATPase) maintains
the intracellular concentrations of sodium
and potassium despite their constant leaking Key Points
through the membrane. The sodium/potassium • At the resting potential there is no net
pump transports three sodium ions out of flow of current between the inside and the
the cell for every two potassium ions carried outside of a cell.
into the cell. Because the pump is not elec- • Hyperpolarization moves the resting
trically neutral, it contributes directly to the potential away from zero; depolarization
resting potential; that is, it is electrogenic. moves it toward zero.
The contribution of the sodium/potassium • The high permeability of the membrane
pump steady state to the resting potential to potassium is the primary determinant
is approximately 11 mV. The cell membrane of the resting potential.
at rest is permeable also to chloride ions. • Outward leakage of potassium and inward
In most membranes, chloride reaches equi- leakage of sodium at rest are balanced
librium simply by adjustment of its inter- by the reverse action of the sodium/
nal concentration to maintain electroneutrality, potassium pump.
without affecting the steady state membrane • Calcium stabilizes the membrane.
potential. • Glial cells buffer the extracellular concen-
tration of potassium.
Role of Extracellular Calcium

LOCAL POTENTIALS
The external surface of the cell membrane
contains a high density of negative charges In a normal nerve cell or muscle cell with
because of the presence of glycoprotein adequate sources of oxygen and glucose, the
residues. This produces a local negative poten- resting potential is maintained at a stable, rel-
tial that contributes to the resting membrane atively unchanging level. However, the resting
potential. Divalent ions, such as extracellu- potential readily changes in response to stimuli.
lar calcium, alter the transmembrane potential The membrane potential can change from the
by neutralizing this local, negative surface resting state in only two ways. It can become
potential. Neutralizing the surface poten- either more negative inside, hyperpolarization,
tial increases the contribution of the trans- or less negative inside, depolarization. Even if
membrane potential to the resting potential, the membrane potential reverses, so that the
and this increases the threshold for opening inside becomes positive with respect to the
voltage-gated sodium channels. This explains outside, it is still referred to as depolarization,
the stabilizing effect of extracellular cal- because the potential is less negative than the
cium on membrane excitability and the pres- resting potential.
ence of increased spontaneous activity (tetany) The changes in the membrane potential that
that occurs in patients with hypocalcemia or occur with anoxia or a change in the concentra-
alkalosis. tion of the ions on either side of the membrane
are relatively long lasting (minutes to hours). 4. Opening of chloride channels increases
In contrast, rapid changes (seconds or less) chloride conductance, resulting in rapid
can occur in response to electrical, mechanical, stabilization or hyperpolarization of the
or chemical stimuli. These changes occur as a membrane voltage.
result of current flow through the membrane. 5. Closing of potassium channels, resulting
Transient currents in living tissues are due to in a slow depolarization.
the movement of charged ions and can flow 6. Opening of calcium channels, resulting in
through the membrane as a result of an applied a slow depolarization.
voltage or of a change in membrane conduc-
tance. A local potential is a transient depolar- Generator potentials occur primarily by open-
izing or hyperpolarizing shift of the membrane ing of both sodium and potassium channels and
potential in a localized area of the cell. Local increasing conductance of both ions. This pro-
potentials result from the current flow due to duces depolarization. Generator potentials also
localized change in ion channels that alter the occur in response to specific molecules that
permeability to one or more ions. Ion channel activate olfactory receptors and to photic stim-
opening or closing may result from: uli that activate photoreceptors in the retina of
the eye. Electrotonic potentials occur in one of
1. Synaptic potential from a chemical agent two ways:
acting on the channel.
2. Receptor potential from a stimulus acting 1. Opening of sodium channels by a current
on a sensory receptor channel. arising from a voltage in an adjacent area
3. Electrotonic potential from an eternally of membrane. This produces depolariza-
applied voltage. tion.
2. Opening or closing of several different
Synaptic potentials are the response to infor- ion channels by an externally applied volt-
mation carried by a neurotransmitter released age. Application of a negative voltage to
by an adjacent neuron. Receptor potentials are the outside of the membrane causes out-
the response to external stimuli. Electrotonic ward current flow and depolarization of
potentials participate in the transfer of infor- the membrane. Application of a positive
mation throughout a cell by action potentials. voltage to the outside of the membrane
causes inward current flow and hyperpo-
larization of the membrane. When a volt-
Ionic Basis age is applied to the outside of the axonal
membrane, the negative pole is com-
Local potentials result from the flow of current monly referred to as the cathode and the
through the membrane with a change in chan- positive pole as the anode. The current
nels that are open or closed in response to a flow at the cathode depolarizes, whereas
chemical agent, mechanical deformation, or an that at the anode hyperpolarizes a mem-
applied voltage. Neurotransmitters and neuro- brane. This is because the negative charge
modulators produce synaptic potentials by one at the cathode is extracellular and hence
of six mechanisms: decreases the transmembrane potential,
conversely for the anode.
1. Opening of potassium channels increases
potassium conductance, resulting in hyper-
polarization, a relatively slow process. Characteristics of Local Potentials
2. Opening of sodium channels increases
sodium conductance, resulting in depo- All local potentials have certain characteristics
larization, a relatively fast process. in common (Table 5–6). Importantly, the local
3. Opening of both potassium and sodium potential is a graded potential; that is, its ampli-
channels increases the conductance of tude is proportional to the size of the stimulus
both ions, resulting in a depolarization (Fig. 5–5). Measurement of a local potential
but to a lesser degree than in item 2 uses the resting potential as its baseline. If
above. the membrane’s resting potential is depolarized
Table 5–6 Comparison of Local Potentials and Action Potentials

Characteristic Local potentials Action potentials
Example Generator Nerve impulse

Synaptic
Electrotonic
Duration, ms 5–100 1–10
Amplitude, mV 0.1–10 70–110
Ionic mechanism Local changes in Transient increase in
permeability to Na+ , K+ , permeability to Na+ ,
Ca2+ , or Cl− followed by increase in
permeability to K+
Threshold No Yes
Spatial and temporal Yes No (all-or-none)
summation
Refractory period No Yes
Propagation Passive and decremental Active and
nondecremental
from −80 to −70 mV during the local poten- hyperpolarization). Because the local potential
tial, the local potential has an amplitude of is a graded response proportional to the size
10 mV. This potential change is one of decreas- of the stimulus, the occurrence of a second
ing negativity (or of depolarization), but it stimulus before the first one subsides results
could also be one of increasing negativity (or of in a larger local potential. Therefore, local
0
Membrane potential
–20 E
(mV)
–40
D C
–60 B
0 2 4 (ms)
D
Stimulating
C
current
B
0
A
Figure 5–5. Local potentials. These potentials are shown as an upward deflection if they are a depolarization and as a
downward deflection if they are a hyperpolarization. Resting potential is 70 mV. At time zero, electrical currents of varied
polarities and voltages are applied to the membrane (bottom). A is an anodal current; B, C, and D are cathodal currents. A
produces a transient hyperpolarization; B, C, and D produce a transient depolarization that is graded and proportional to
the size of the stimulus. All of these are local potentials. D produces an action potential, E.
potentials can be summated. They are sum- subsided (Fig. 5–6). This summation of local
mated algebraically, so that similar potentials potentials occurring near each other in time
are additive and hyperpolarizing and depolar- is called temporal summation (Fig. 5–6B). Dif-
izing potentials tend to cancel out each other. ferent synaptic potentials have different time
Summated potentials may reach threshold and courses. Most synaptic potentials range from
produce an action potential when single poten- 10 to 15 ms in duration; however, some are
tials individually are subthreshold. When a very brief, lasting less than 1 ms, but others
stimulus is applied in a localized area of the may last several seconds or several minutes.
membrane, the change in membrane potential Longer duration synaptic potentials have a
has both a temporal and a spatial distribution. greater chance for temporal summation. By
A study of the temporal course of the local means of temporal summation, the cell can
potential (Fig. 5–6) shows that the increase in integrate signals that arrive at different times.
the potential is not instantaneous, but develops Study of the spatial distribution of local
over a few milliseconds. After the stimulus potentials reveals another of their character-
ends, the potential subsides over a few istics. As their name implies, they remain
milliseconds as well. Therefore, local potentials localized in the region where the stimulus is
have a temporal course that outlasts the stim- applied; they do not spread throughout the
ulus. The occurrence of a second stimulus at entire cell. However, the locally applied stim-
the same site shortly after the first produces ulus, because of local current flow, has an
another local potential, which summates with effect on the nearby membrane. The potential
the residual of the earlier one that has not yet change is not sharply confined to the area of
A Spatial summation
Microelectrode
Action
Axon potential
0 mV
Neuron
cell body EPSP
Threshold
80 mV
Dendrites
Nerve
terminal
Temporal summation
0 mV
Threshold
Figure 5–6. Summation of local potentials in a neuron. A, Spatial summation occurs when increasing numbers of nerve
terminals release more neurotransmitter to produce larger EPSPs. B, Temporal summation occurs when a single terminal
discharges repetitively more rapidly to produce larger EPSPs.
Synaptic
Sensory stimulus potential Neurotransmitter
Generator Electrotonic Action

potential potential potential
Figure 5–7. Local potentials and triggering of the action potential. Three types of local potentials are (1) receptor (or
generator) potential, triggered by the action of a sensory stimulus on a sensory receptor; (2) synaptic potential, triggered by
the action of a neurotransmitter; and (3) electrotonic potential, which consists of the passive movement of charges according
to the cable properties of a membrane. Both the generator and synaptic potentials give rise to electrotonic potentials, which
depolarize the membrane to threshold for triggering an action potential. The action potential is a regenerating depolarizing
stimulus that, via electrotonic potentials, propagates over a distance without decrement in its amplitude.
the stimulus but falls off over a finite distance accommodation require several milliseconds,
along the membrane, usually a few millimeters. both to develop and to subside. As a result, if
The application of a simultaneous second an electrical stimulus is applied gradually so
stimulus near the first (but not at the same that accommodation prevents a change in the
site) results in summation of the potentials in membrane potential, no effect is observed.
the border zones; this is called spatial sum-
mation (Fig. 5–6A). Thus, the membrane of Key Points
the cell can act as an integrator of stimuli that
arrive from different sources and impinge on • Amplitude of depolarizing and hyperpo-
areas of membrane near one another. Spatial larizing local potentials depend on stimu-
summation and temporal summation are lus intensity.
important mechanisms in the processing of • Local changes in membrane potential can
information by single neurons; when sum- be triggered by synaptic transmitters, volt-
mated local potentials reach threshold, they age changes, or sensory stimuli.
initiate an action potential (Fig. 5–7). If a cur- • Local potentials summate spatially and
rent or voltage is applied to a membrane for temporally.
more than a few milliseconds, the ion channels
revert to their resting state, changing ionic con-
ductances of the membrane in a direction to ACTION POTENTIALS
restore the resting potential to baseline value.
This phenomenon is known as accommodation Action potentials have several advantages for
(Fig. 5–8). Therefore, if an electrical stimulus the rapid transfer of information in the ner-
is increased slowly, accommodation can occur vous system. Because action potentials are
and no change will be seen in the membrane all-or-none (i.e., they are not graded responses;
potential. The changes in conductance during they either occur or do not occur), they can
transfer information without loss over rela-
Accommodation tively long distances. Their all-or-none fea-
ture also allows coding of information as fre-
Membrane quency rather than the less stable measure of
potential amplitude. Also, their threshold eliminates the
effects of small, random changes in membrane
Stimulus potential.
On Off
Figure 5–8. Accommodation of the membrane potential
to applied stimulus of constant strength. Note the response Threshold
to sudden cessation of the stimulus when the stimulus is
turned off. The residual change in conductance produces
a transient change in resting potential. Thus, accommoda-
The membranes of neurons, axons, and muscle
tion can result in a cell responding to the cessation of a cells have another characteristic that is basic to
stimulus. their ability to transmit information from one
area to another—their excitability. If a mem- which is approximately +60 mV. This depo-
brane is depolarized by a stimulus, there is a larization locally reverses the polarity of the
point at which suddenly many sodium chan- membrane, the inside becoming positive with
nels open. This point is known as the threshold respect to the outside. With the opening of
for excitation (Fig. 5–5). If the depolarization the sodium channels and increased sodium
does not reach threshold, the evoked activity is conductance, there is a flow of current with
a local potential. Threshold may be reached by the inward movement of sodium ions. The
a single local potential or by summated local change in sodium conductance is usually tran-
potentials. When threshold is reached, there is sient, lasting only a few milliseconds, and is
a sudden increase in the membrane’s perme- followed by the opening of potassium chan-
ability to sodium. This change in conductance nels, an increase in the potassium conductance,
results in the action potential. and an outward movement of potassium ions.
These three changes overlap, and the poten-
tial of the membrane during these changes is
Ionic Basis of Action Potentials a function of the ratios of the conductances
(Fig. 5–9).
In the resting state, many more potassium Sodium conductance increases several
channels are open, the conductance of sodium thousand-fold early in the process, whereas
is much less than that of potassium, and the potassium conductance increases less, does so
resting potential is near the equilibrium poten- later, and persists longer. The conductance
tial of potassium. At threshold, many sodium changes for these two ions result in ionic shifts
channels open so that the conductance of and current flows that are associated with a
sodium suddenly becomes greater than that of membrane potential change: the action poten-
potassium, and the membrane potential shifts tial (Fig. 5–9). The action potential is a sud-
toward the equilibrium potential of sodium, den, short duration, all-or-none change in the
A +60 B
Membrane potential
+20
100 0 mV
(mV)
–20
Na+ Conductance
K+ Conductance
10
–60
1
–100
0.1
0.01
Direction
2)
0.001
Conductance (mmho/cm
100 of
+
Na propagation
0 mV
10
K+
1
+
K exit Na+ entry
Resting
0.1 potential
0.01 Axon
0.005
0 0.5 1.0 1.5
ms
Figure 5–9. Conductance changes during action potential. A, Temporal sequence at a single site along an axon. Changes
in conductances (permeabilities) of sodium and potassium are plotted against time as they change with associated changes in
membrane potential. Note that sodium conductance changes several thousand-fold early in the process, whereas potassium
conductance changes only about 30-fold during later stages and persists longer than sodium conductance changes. B, Spatial
distribution of an action potential over a length of axon at a single instant.
+30
Membrane potential (mV)

0 C
Repolarization
–50 B Threshold
D
A
–70 Resting potential
E
1.0 2.0 3.0 4.0

Time (ms)
Figure 5–10. Components of an action potential with a resting potential of 70 mV. A, Local electrotonic potential. B,
Threshold level. C, Spike. D, Negative (depolarizing) afterpotential. E, Positive (hyperpolarizing) afterpotential.
membrane potential that occurs if the mem- an extracellular electrode. The increase in
brane potential reaches threshold (Table 5–5). potassium conductance persists and results
Its components are shown in Figure 5–10. in a hyperpolarization after the spike com-
The initial portion of the membrane potential ponent of the action potential—the after-
change is the local potential. At threshold, the hyperpolarization. The after-hyperpolarization
rising phase of the action potential suddenly is due to continued efflux of potassium ions,
changes because of the influx of positive ions. with a greater potential difference between the
In most nerve cells and skeletal muscle cells, inside and the outside of the cell than the
the inward current during the rising phase typical resting potential difference. The after-
of the action potential is carried by sodium hyperpolarization is positive when measured
ions, because sodium conductance is markedly with extracellular electrodes and therefore is
increased. The action potential also could be called a positive afterpotential. During the pos-
carried by calcium ions if the calcium conduc- itive afterpotential, the membrane potential
tance increased sufficiently, as occurs in some is near the potassium equilibrium potential,
dendrites. and is increased with increased activity of the
Repolarization begins as sodium conduc- sodium pump. The amounts of sodium and
tance decreases or potassium conductance potassium that move across the membrane dur-
increases (usually both). The decreased flow ing the action potential are small, buffered by
of sodium ions is followed by an efflux of surrounding astrocytes, and do not change the
potassium ions. The rate of return of the concentration enough to result in a change in
membrane potential to the baseline slows the resting potential. In addition, the sodium
after sodium conductance has returned to that moves in during the action potential is
baseline, producing a small residual on the continually removed by the sodium pump dur-
negative component of the action potential, ing the relatively long intervals between action
which is called the negative afterpotential. In potentials.
some myelinated axons, repolarization occurs
by a decrease in sodium conductance with
no change in potassium conductance. The Excitability
afterpotential is positive when the membrane
potential is recorded with a microelectrode The excitability of a membrane is the ease
within the cell, but it is called negative with which an action potential can be gener-
because it is negative when recorded with ated and is usually measured in terms of the
Axon membrane
potential 0 mV
Threshold
80 resting potential
Supernormal period
Axon excitability 100%

Subnormal period
0
Relative refractory period
Absolute refractory period
Figure 5–11. Excitability changes during an action potential. The lower portion of the illustration shows the ease with
which another action potential can be elicited (change in threshold). During absolute and relative refractory periods, the
amplitude of the action potential evoked is low. Subsequently, it is normal.
voltage required to initiate an action poten- the membrane potential is very near thresh-
tial. During increased sodium conductance, old, the cell may fire spontaneously. If the
the membrane cannot be stimulated to dis- membrane potential remains more depolar-
charge again. A second stimulus at this time ized than threshold, however, the membrane
is without effect; therefore, action poten- cannot be stimulated to fire another action
tials, unlike local potentials, cannot summate. potential (Fig. 5–12). The term threshold is
This period of unresponsiveness is the abso- also used to describe the voltage required to
lute refractory period (Fig. 5–11). As sodium excite an action potential with an eternally
conductance returns to normal, the mem- applied stimulus. When threshold is used in
brane again becomes excitable, but for a short this sense, an axon with an increased excitabil-
period, termed the relative refractory period, it ity due to partial depolarization may be said
requires a larger stimulus to produce an action to have a lower threshold for stimulation, even
potential. After the relative refractory period, though the actual threshold is unchanged. The
while the negative afterpotential is subsiding,
the membrane is partially depolarized, is closer A mV B C
to threshold, and has an increased excitability. +20
This period is the supernormal period. Finally,
during the positive afterpotential, the mem- 0
brane is hyperpolarized, and stronger stimuli –20
are required. This period is the subnormal
period. –40
Up to now, the term threshold has been –60 Threshold
used to refer to the membrane potential at
which sodium channels open and an action –80
potential is generated. The threshold of a
Figure 5–12. The effect of stimulation of a neuron at dif-
membrane remains relatively constant. If the ferent resting potentials as recorded with a microelectrode.
membrane potential becomes hyperpolarized, A, The membrane is hyperpolarized, and a stimulus pro-
it moves away from threshold, and the mem- duces a subthreshold local potential. B, The membrane
brane is less excitable. If the membrane poten- is normally polarized at 65 mV, and a stimulus produces
a local potential that reaches threshold and results in an
tial moves closer to threshold, the membrane action potential. C, The membrane is depolarized beyond
becomes more excitable and will generate an threshold, and a stimulus produces only a small local
action potential with a smaller stimulus. If potential.
first meaning of threshold is used when intra- to shift the membrane potential to thresh-
cellular recordings are considered, and the old and thereby generate an action potential
second is used in reference to extracellular in the immediately adjacent membrane. Thus,
stimulation and recording. The threshold of the action potential spreads away from its site
the nerve membrane differs in different parts of initiation along an axon or muscle fiber.
of the neuron: it is high in the dendrite and Because of the refractory period, the poten-
soma and lowest at the initial segment. Thus, tial cannot reverse and spread back into an area
an action potential is usually generated in the just depolarized.
area of the axon hillock. The rate of conduction of the action poten-
tial along the membrane depends on the
amount of longitudinal current flow and on
the amount of current needed to produce
Propagation depolarization in the adjacent membrane. The
longitudinal current flow can be increased by
Another important characteristic of action increasing the diameter of an axon or mus-
potentials is their propagation. If an action cle fiber, because this increase reduces the
potential is initiated in an axon in the tip of the internal resistance, just as a larger electrical
finger, for instance, the potential spreads along wire has a lower electrical resistance. How-
the entire length of that axon to synapse on ever, many axons in the central and peripheral
its second-order sensory neuron in the spinal nervous systems have an increased conduc-
cord. This characteristic permits the nervous tion velocity because they are insulated with
system to transmit information from one area a myelin sheath. A myelinated axon has its
to another. When an area of membrane is membrane bared only at the nodes of Ranvier,
depolarized during an action potential, ionic so that transmembrane current flow occurs
currents flow (Fig. 5–13). In the area of depo- almost exclusively at the nodal area. When
larization, sodium ions carry positive charges the current flow opens enough sodium chan-
inward. There is also a longitudinal flow of cur- nels to reach threshold in the nodal area, it
rent both inside and outside the membrane. results in many more sodium channels opening
This flow of positive charges (current) toward and an influx of sodium ions with a genera-
nondepolarized regions internally and toward tion of an action potential. The nodal area in
depolarized regions externally tends to depo- the mammalian nervous system is unique in
larize the membrane in the areas that sur- that it consists almost exclusively of sodium
round the region of the action potential. This channels, with an almost complete absence of
depolarization is an electrotonic potential. In potassium channels. The potassium channels
normal tissue, this depolarization is sufficient are located at the paranodal regions (adjacent
Action potential
0 mV
Local
potential
Resting
potential
Axon
Figure 5–13. Current flow and voltage changes in an axon in the region of an action potential. The voltage changes along
the membrane are shown in the upper part of the figure and the spatial distribution of current flow is shown in the lower
part as arrows through the axon membrane.
Direction
Current
of
flow
propagation
I1 I2 I3
N1 N2 N3
A
Fiber Axon
Myelin
Action potential
Repolarization Electronic potential
B Inward
N1 N2 N3
Instantaneous transmembrane current flow Outward
N1 Resting potential
I1 RP
I = Internode
N2 RP N = Node of Ranvier
I2 RP
N3 RP
I3 RP
Time
Figure 5–14. Saltatory conduction along an axon from left to right. A, The charge distribution along the axon is shown
with an action potential (depolarization) at the second node of Ranvier (N2 ). Current flow spreads to the net node (N3 ). B,
Membrane current flow along the axon. C, The portion of the action potential found at each node is indicated by dotted
lines.
to the node), which are covered by myelin. The • Excitability is the probability of triggering
action potential generated at the node con- an action potential.
sists predominantly of inward sodium currents • The density of voltage-gated sodium
with little outward potassium currents, and potentials determines the threshold for
repolarization is achieved by means of sodium triggering an action potential.
inactivation and leakage currents. An action • Membrane depolarization (toward thresh-
potential at one node of Ranvier produces suf- old) increases excitability.
ficient longitudinal current flow to depolarize • A refractory period occurs with mem-
adjacent nodes to threshold, thereby propa- brane depolarization above threshold
gating the action potential along the nerve in and inactivation of voltage-gated sodium
a skipping manner called saltatory conduction channels.
(Fig. 5–14).
Key Points Patterns of Activity

• An action potential consists of two compo- Information in the nervous system is coded
nents: by the number and type of axons that are
◦ Rapid membrane depolarization due to active and by the firing pattern of action poten-
opening of voltage-gated sodium chan- tials. This activity is initiated in peripheral
nels. receptors or in neurons. Neuronal firing of
◦ Repolarization due to delayed opening action potentials may occur spontaneously or
of voltage-gated potassium channels. in response to external stimulation. Beating,
or pacing, neurons fire repetitively at a con- the encoding of information by cortical neu-
stant frequency; their intrinsic firing rate may rons and interferes with the transmission of
be increased or decreased by external stimula- sensory information. Inactive states of the cere-
tion. Bursting neurons generate regular bursts bral cortex occur during deep sleep and in
of action potentials separated by hyperpo- some types of seizures.
larization of the membrane. Such neurons
are important for rhythmic behavior such as
breathing, walking, and chewing. Neurons that Key Points
fire in response to external stimulation may do • The action potential is an all-or-none sig-
so in one of three ways: nal that is transmitted without decrement
along an axon.
• The frequency of discharge of action
1. A sustained response neuron shows
repeated action potentials with a constant potentials is determined by the amplitude
firing frequency that reflects the strength of the stimulus.
• Action potential conduction velocity
of the stimulus.
2. A delayed response neuron fires action depends on axon diameter and the myelin
potentials only after stimulation of suffi- sheath.
• Voltage-gated sodium channels are clus-
cient intensity.
3. An accommodation response neuron fires tered at the nodes of Ranvier in a myeli-
only a single potential at the onset of nated axon.
• Potassium channels are covered by the
stimulation and remains silent thereafter.
myelin sheath.
Some neurons (e.g., in the thalamus) have the

ability to discharge either in rhythmic bursts
or with typical action potentials. The firing SYNAPTIC TRANSMISSION
pattern depends on the level of the resting
membrane potential. An important property A synapse is a specialized contact zone where
of this type of neuron is the presence of a one neuron communicates with another neu-
particular class of calcium channel, the T chan- ron. The contact zone between a neuron and
nel. This channel can be activated only if the a nonneural effector element (e.g., a muscle
membrane potential is relatively hyperpolar- fiber) is referred to as a neuron-effector junc-
ized (e.g., −80 mV). Under this condition, a tion. The two types of synapses are chemical
stimulus opens the T channel and calcium and electrical. The most common form of com-
enters the cell and produces a small, brief munication in the nervous system is through
calcium-based depolarizing potential change chemical synapses. A chemical synapse con-
called the low-threshold calcium spike. This sists of a presynaptic component (containing
calcium spike triggers the opening of sodium synaptic vesicles), a postsynaptic component
channels, which produces a burst of repetitive (dendrite, soma, or axon), and an interven-
action potentials. ing space called the synaptic cleft (Fig. 5–15).
As calcium accumulates in the cell, it Many of the drugs used in clinical medicine
opens a calcium-activated potassium chan- have their pharmacologic site of action at the
nel that allows the efflux of potassium. synapse. The mechanism underlying chemical
The resulting hyperpolarization (called after- synaptic transmission should make it appar-
hyperpolarization) allows reactivation of the T ent that synaptic transmission has three unique
channel, entry of sodium, and recurrence of characteristics:
the cycle. This sequence generates rhythmic
burst firing of the neuron. Thus, T channels are 1. First, conduction at a synapse is delayed
an exception to the general rule of neuronal because of the brief interval of time
excitability: hyperpolarization “deinactivates” required for the chemical events to occur.
T channels and increases the likelihood that 2. Second, because the two sides of the
the neuron will discharge in rhythmic bursts of synapse are specialized to perform only
activity. Rhythmic burst firing in thalamic neu- one function, transmission can occur in
rons that project to the cerebral cortex impairs only one direction across the synapse.
Thus, neurons are polarized in the direc- Biosynthesis, Storage, Release,

tion of impulse transmission. and Reuptake of Neurochemical
3. Third, because nerve impulses from many Transmitters
sources impinge on single cells in the
central and peripheral nervous systems, Neurochemical transmitters include amino
synaptic potentials summate both tempo- acids, acetylcholine, monoamines (catechol-
rally and spatially. amines, serotonin, and histamine), neuropep-
tides, and purines (ATP and adenosine). Amino
The membrane of a cell is continually bom- acids include l-glutamate, the most abundant
barded with neuromodulators and neuro- excitatory neurotransmitter in the central ner-
transmitters, which produce either excitatory vous system, and γ -aminobutyric acid (GABA),
postsynaptic potentials (EPSPs) or inhibitory the most abundant inhibitory neurotransmit-
postsynaptic potentials (IPSPs) of varying ter. Both of these neurotransmitters are syn-
duration. When the membrane potential thesized from intermediate metabolites of the
reaches threshold, an action potential is gen- Krebs cycle. Acetylcholine and monoamines
erated. Thus, a single neuron can integrate are synthesized by specific enzymes from pre-
activity from many sources. A summary of the cursors that are actively taken up by the
electrical events in a single cell underlying the presynaptic terminal. Neuropeptides are syn-
transmission, integration, and conduction of thesized from messenger RNA in the cell
information is shown in Figure 5–16. body and transported to the synaptic terminal.
A Resting synapse
Synaptic
Presynaptic cleft Postsynaptic
neuron cell
B Active synapse
Action potential Action potential Effector activity
Depolarization
Electrical synaptic potential
events
Ca2+ Diffusion of Action potential and

transmitter release of neurotransmitter
Active zone Nerve
Muscle contraction
Muscle
Gland
Secretion of
Chemical
hormones
events Postsynaptic
Transmitter Transmitter membrane
released reacts with activation
postsynaptic initiates
receptor function
Figure 5–15. Synaptic transmission. A, In a resting synapse, both the presynaptic axon terminal and the postsynaptic
membrane are normally polarized. B, In an active synapse, an action potential invades the axon terminal (from left in
the diagram) and depolarizes it. Depolarization of the axon terminal of a presynaptic neuron results in the release of
neurotransmitter from the terminal. The neurotransmitter diffuses across the synaptic cleft and produces local current
flow and a synaptic potential in the postsynaptic membrane, which initiates the effector activity (neuronal transmission,
neurotransmitter release, hormonal secretion, or muscle contraction).
Presynaptic terminals Myelinated axon

Axon hillock
Cell body
Dendrites
Graded EPSPs Action potential Action potential

initiated conducted to next cell
Figure 5–16. Neuronal electrical activity from its initiation by excitatory postsynaptic potentials (EPSPs) to its transmis-
sion as an action potential to another area.
Neurochemical transmitters are stored in spe- other presynaptic mechanisms also regulate
cial intracellular organelles called synaptic neurotransmitter release:
vesicles. Small clear synaptic vesicles store the
classic neurotransmitters (amino acids, acetyl-
choline, monoamines), and large dense core 1. The neurotransmitter inhibits its own
secretory granules store neuropeptides. release, acting via presynaptic inhibitory
Neurotransmitter release is triggered by the autoreceptors.
influx of calcium through voltage-gated chan- 2. Inhibitory neurons (generally contain-
nels that open in response to the arrival of ing GABA) form axoaxonic synapses that
an action potential in the presynaptic termi- inhibit the release of neurotransmitter
nal. These channels are clustered in specific from the postsynaptic axon, a process
regions of the presynaptic membrane called called presynaptic inhibition (Fig. 5–17).
active zones (Fig. 5–15B). The synaptic vesi-
cles are mobilized in the presynaptic terminal The synaptic action of neurotransmitters is
and dock close to the active zones. In response terminated by several mechanisms. Presynap-
to the influx of calcium, the vesicle membrane tic reuptake, mediated by specific sodium-
fuses with the presynaptic membrane, which dependent and ATP-dependent neurotrans-
allows the release of the neurotransmitter into mitter transporters, is the primary mechanism
the synaptic cleft; this process is called exo- of inactivation of glutamate, GABA, and
cytosis. The mobilization, docking, and fusion monoamines. Monoamines are metabolized
of synaptic vesicles depend on the interactions after reuptake by monoamine oxidases and
of various synaptic vesicle proteins with other methyltransferases. Acetylcholine and neu-
components of the presynaptic terminal. ropeptides do not undergo reuptake but are
A neuron can produce and release dif- rapidly inactivated by enzymatic hydrolysis in
ferent neurotransmitters. Neurons frequently the synaptic cleft.
contain a classic neurotransmitter (an amino
acid or acetylcholine) and one or more neu-
ropeptides. The neuron can release a variable
mixture of these neurotransmitters according Postsynaptic Effects of
to its firing pattern, a process referred to as Neurochemical Transmitters
frequency-dependent chemical coding. Classic
neurotransmitters can be released after a single Postsynaptic effects are mediated by two
action potential; neuropeptides are released in main classes of receptors: (1) Ligand-gated
response to rapid, burst firing of a neuron. Two receptors or ion channels mediate rapid
A
1
Axon
3
Axon 2
Axon
Presynaptic terminal (2) Presynaptic (2) Postsynaptic

resting potential action potential EPSP amplitude (3)
amplitude
Normal 110 mV 20 mV
–70 mV
2 active alone
–60 mV 100 mV 10 mV
Presynaptic inhibition
1 and 2 active
Figure 5–17. A, Presynaptic inhibition of neuron 3 when axon 1 partially depolarizes axon 2. B, Response to axon 2 acting
alone. C, Response to axon 2 after depolarization of axon 1. In the latter case, there is less neurotransmitter and a smaller
excitatory postsynaptic potential (EPSP).
changes in ionic conductance (ionotropic the electrical behavior of the neuron but also
effect) and (2) G-protein-coupled receptors may produce long-term effects, such as use-
produce slower changes in neuronal excitabil- dependent modification of synaptic efficacy,
ity and metabolism (metabotropic effect) cytoskeletal changes during development and
(Table 5–7). These changes not only modify repair, and control of genetic transcription.
Table 5–7 Comparison of Classic Neurotransmission and Neuromodulation

Classic neurotransmission Neuromodulation
Function Rapid synaptic excitation or Modulation of neural

inhibition excitability
Receptor mechanism Ion channel receptors G-protein-coupled receptors
Ionic mechanism Opening of either cation Opening or closing of
channel (fast EPSP) or Cl− voltage-gated K+ or Ca2+
channel (fast IPSP) channels (slow IPSP and slow
EPSP)
Example l-Glutamate (ionotropic) l-Glutamate (metabotropic)
GABA (GABAa ) GABA (GABAa )
Acetylcholine (nicotinic) Acetylcholine (muscarinic)
Monamines
Neuropeptides
Adenosine
Systems Relay systems, direct Diffuse systems, indirect
Sensory Internal regulation
Motor Consciousness
Note: EPSR, excitatory postsynaptic potential; GABA, γ -aminobutyric acid; IPSP, inhibitory postsynaptic potential.
Table 5–8 Postsynaptic Potentials

Receptor Ionic mechanism Effect Kinetics
Nicotinic Increased cation (Na+ , Ca2+ ) Excitatory Fast

Glutamate conductance
GABAa Increased Cl− conductance Inhibitory Fast
Glycine
G-protein-coupled Decreased K+ conductance Excitatory Slow
receptors Increased K+ conductance Inhibitory Slow
Classic Neurotransmission 10 mV less negative than the resting poten-

tial. Ion currents that increase the net positive
Classic neurotransmission is used for fast, pre- charge of the membrane produce depolariza-
cise, point-to-point transmission of excitatory tion EPSPs because they bring the membrane
or inhibitory signals. It involves rapid, brief potential toward the threshold for triggering
opening of ligand-gated ion channel recep- an action potential. In classic neurotransmis-
tors. Neurotransmitters produce a transient sion, fast EPSPs result from the opening of
increase or decrease in ion channel conduc- cation channels (conducting sodium ions and,
tance to the passive flow of a specific ion in some cases, calcium ions). Ligand-gated
current. These ionic currents produce local cation channel receptors that produce fast
changes in the membrane potential called EPSPs include nicotinic acetylcholine recep-
postsynaptic potentials. In most mammalian tors and several ionotropic glutamate recep-
neurons, the resting membrane potential is tors. Ion currents that increase the net negative
approximately 60 to 80 mV. The threshold charge of the membrane produce IPSPs. Fast
for opening voltage-gated sodium channels IPSPs are produced by the opening of chloride
that trigger an action potential is reached channels. GABA (via GABAA receptors) and
when the postsynaptic potentials drive the glycine act via this mechanism (Table 5–8 and
membrane potential to a value that is about Fig. 5–18).
Action
potential
Threshold
EPSP EPSP
Resting
potential IPSP
Inhibitory
endings (IPSP)
Excitatory
endings (EPSP)
Figure 5–18. Postsynaptic inhibition in the neuron on the left occurs when the inhibitory and excitatory endings are active
simultaneously. On the right, a microelectrode recording shows two excitatory postsynaptic potentials (EPSPs) summating
to initiate an action potential. When there is a simultaneous occurrence of an inhibitory postsynaptic potential (IPSP),
depolarization is too low to reach threshold, and no action potential occurs.
Neuromodulation synapses, which transmit signals in only one

direction.
Neurotransmitters acting through G-protein-
coupled receptors, second messengers, and Key Points
protein phosphorylation cascades control the • Presynaptic activities include
excitability and responsiveness of neurons ◦ Synthesis and storage of neurotransmit-
to rapid synaptic signals, a process called ters in synaptic vesicles
neuromodulation. G-protein-coupled recep- ◦ Vesicle mobilization
tors include metabotropic glutamate receptors, ◦ Exocytic release of neurotransmitter.
GABAB receptor, and receptors for cate- • Action potentials trigger the opening of
cholamines, serotonin and histamine, neu- presynaptic, voltage-gated calcium chan-
ropeptides, and adenosine. Potassium channels nels and exocytosis of neurotransmitter.
are an important target of neuromodulatory • Fast excitatory or inhibitory postsynaptic
signals. Potassium currents determine the pat- potentials are elicited by neurotransmitters
tern of activity generated by neurons through acting on ligand-gated chloride or cation
control of the resting membrane potential, channels.
repolarization of the action potential, and prob- • Neuromodulation (changes in neuronal
ability of generation of repetitive action poten- excitability due to permeability changes
tials. The opening of potassium channels brings of voltage-gated potassium or calcium
the membrane potential toward the equilib- channels) is brought about by neuro-
rium potential of potassium (100 mV) and transmitters acting on G-protein-coupled
thus away from the threshold for triggering receptors.
an action potential. Closure of the potassium
channels moves the membrane potential away
from the equilibrium potential of potassium
and thus closer to the threshold.
Neuromodulation involves the production CLINICAL CORRELATIONS
of slow potentials. Activation of G-protein
receptors that lead to closure of potas- Pathophysiologic Mechanisms
sium channels produces slow depolarization
and increased neuronal excitability. G-protein The mechanisms responsible for neuronal
receptor mechanisms that increase potassium excitability, impulse conduction, and synap-
permeability lead to membrane hyperpolar- tic transmission in the central and peripheral
ization and reduce neuronal excitability. The nervous system may be altered transiently to
same neurotransmitter may act via differ- produce either loss of activity or overactiv-
ent receptor subtypes, each coupled to a ity of neurons. A loss of activity results in a
distinct transduction pathway. Also, differ- clinical deficit of relatively short duration (sec-
ent neurotransmitters, via their respective onds to hours), whereas overactivity results
receptors, may activate a similar transduction in extra movements or sensations. Both types
pathway. of transient alteration are usually reversible.
These transient disorders may be focal or
generalized (Table 5–9) and may be due
Electrical Synapses to different mechanisms (Table 5–10). Tran-
sient disorders reflect disturbances in neuronal
Although most synapses in the nervous system excitability due to abnormalities in membrane
use chemical transmitters, neurons with junc- potential.
tions that contain channels extending from the
cytoplasm of the presynaptic neuron to that
of the postsynaptic neuron interact electrically. Energy Failure
In these electrical synapses, the bridging chan-
nels mediate ionic current flow from one cell Energy metabolism is necessary for main-
to the other. Transmission across the electri- tenance of the membrane potential by the
cal synapse is very rapid, without the synaptic ATP-coupled sodium/potassium pump. Most
delay of chemical synapses. Electrical synapses of the ATP produced in the nervous sys-
are also bidirectional, in contrast to chemical tem by aerobic metabolism of glucose is
Table 5–9 Transient Disorders of Neuronal Function

Neuronal excitability Focal disorder Generalized disorder
Increased Focal seizure Generalized seizure

Tonic spasms Tetany
Muscle cramp
Paresthesia
Paroxysmal pain
Decreased Transient ischemic attack Syncope
Migraine Concussion
Transient mononeuropathy Cataplexy
Periodic paralysis
Table 5–10 Mechanisms of paralysis or loss of consciousness (Fig. 5–19).

Transient Disorders The neuron also uses ATP to maintain ion gra-
dients that allow active presynaptic reuptake
Energy failure of neurotransmitters, such as the excitatory
Hypoxia-ischemia amino acid l-glutamate. Under conditions of
Hypoglycemia energy failure, glutamate accumulates in the
Seizures
synapse and produces prolonged activation of
Spreading cortical depression
Trauma its postsynaptic receptors, leading to neuronal
Ion channel disorders depolarization and the accumulation of cal-
Mutation of channel protein (channelopathies) cium in the cytosol. Because the lack of ATP
Immune blockade also impairs active transport of calcium into
Drugs the endoplasmic reticulum or toward the extra-
Toxins cellular fluid, calcium accumulates. Essentially,
Electrolyte disorders all forms of neuronal injury involve to vari-
Demyelination ous extents
1. Accumulation of glutamate and activation

used to maintain the activity of the sodium of glutamate receptors
pump. Conditions such as hypoxia, ischemia, 2. Accumulation of cytosolic calcium and
hypoglycemia, or seizures affect the balance activation of calcium-triggered cascades
between energy production and energy con- 3. Generation of free radicals
sumption of neurons and cause energy failure 4. Mitochondrial failure
and thus impaired activity of sodium/potassium
ATPase.
The consequences are functional and poten-
If the active transport process stops, the
tially reversible (e.g., cell depolarization, pump
cell accumulates sodium and loses potassium
failure, and accumulation of intracellular
and the membrane potential progressively
sodium). If the cause is not corrected, cal-
decreases. This depolarization has two con-
cium accumulates and triggers irreversible
sequences. First, there may be a transient
changes, including destruction of cellular and
increase in neuronal excitability as the mem-
mitochondrial membranes, disorganization of
brane potential moves closer to threshold for
the cytoskeleton, and degradation of DNA by
opening voltage-gated sodium channels and
nucleases, and these effects lead to cell death
triggering the action potential. This may pro-
by necrosis or apoptosis.
duce a paroxysmal discharge of the neuron cell
body or axon. Second, if depolarization per-
sists, the sodium channel remains inactivated
and the neuron becomes inexcitable. This is Ion Channel Blockade
known as depolarization blockade and results
in a focal deficit, such as focal paralysis or Voltage-gated sodium channels mediate the
anesthesia, or a generalized deficit, such as initiation and conduction of action potentials.
ATP-driven [K+] out
pumps
[Na+] in
[Ca2+] in
Effect
Increasing severity of energy failure
Partial depolarization Neuronal

Normal
depolarization blockade death
Normal Increased Decreased Loss of

function excitability excitability function
Figure 5–19. Effects of increasing severity of energy failure (and ATP depletion) on activity of ATP-driven pumps, ionic
concentrations in the intracellular and extracellular fluid, and neuronal electrical activity. With progressive failure of ATP-
driven pumps, potassium accumulates in the extracellular fluid, and sodium and calcium accumulate inside the neuron. This
produces progressive neuronal depolarization. With partial depolarization, the resting potential moves closer to the thresh-
old for triggering an action potential; this results in a transient increase in neuronal excitability, which may be manifested by
paresthesias or seizures. With further depolarization, the membrane potential is at a level that maintains inactivation of the
sodium channel, preventing further generation of action potentials and, thus, reducing neuronal excitability. This consti-
tutes a depolarization block, which manifests with transient and reversible deficits such as paralysis or loss of consciousness.
If the energy failure is severe and prolonged, the excessive accumulation of intracellular calcium triggers various enzymatic
cascades that lead eventually to neuronal death and irreversible loss of function.
A B Excitatory synapses
Presynaptic Inhibitory synapses

inhibitory synapse
Presynaptic
terminal
Transmitter
substance
Block
Depolarizing Excitatory synapse

block
Competitive
inhibition
EPSP Nothing Maintained

depolarization
Figure 5–20. Abnormalities of synaptic transmission may occur (A). Types of transmission block include block of transmit-
ter release (block), block of transmitter binding to postsynaptic membrane (competitive inhibition), and binding of another
depolarizing agent to the membrane (depolarizing block). B, These types of abnormalities may occur at each neuronal
synapse shown.
95
Voltage-gated calcium channels mediate neu- threshold. Information is moved from one area
rotransmitter release, and ligand-gated cation to another as action potentials conducted by
(sodium and calcium) channels mediate excita- single cells. The information is integrated in
tory postsynaptic potentials. All these channels neurons by the interaction of local potentials
may be blocked by autoantibodies, drugs, or generated in response to the neurotransmitters
toxins. Examples of the types of transmission released from depolarized nerve terminals. In
block are illustrated in Figure 5–20. There may this system, information can be coded either as
be presynaptic block of transmitter release, the rate of discharge in individual cells or axons
or postsynaptic block by competitive or non- or as the number and combination of active
competitive inhibition of postsynaptic recep- cells. Both of these are important mechanisms,
tors, or by depolarizing substances. Blockade for although the activity of the nervous system
of sodium channels at the node of Ranvier can be conveniently described in terms of the
slows conduction velocity or causes conduction electrical activity of single cells, the combined
block; this produces a reversible focal deficit activity of large numbers of cells and axons
(weakness or anesthesia). For example, the determines the behavior of the organism.
blockade of sodium channels in sensory axons Each type of alteration in neuronal or mus-
by local anesthetic agents produces anesthesia, cle cell physiology can produce symptoms or
and antibodies against ganglioside GM1 (asso- signs of short duration, transient disorders.
ciated with sodium channels) in the nodes of The particular findings in a patient depend
Ranvier of motor axons produce focal paraly- on which cells are altered. If the changes
sis. Autoantibodies may also block ion channels are in neurons that subserve sensation, there
involved in neuromuscular transmission and may be a loss of sensation or an abnormal
produce reversible muscle fatigue or paralysis. sensation such as tingling, loss of vision, or
“seeing stars.” In other systems, there might
be loss of strength, twitching in muscles,
SUMMARY loss of intellect, or abnormal behavior. In all
these cases, the physiologic alterations are
The transmission of information in the nervous not specific and may be the result of any
system depends on the generation of a rest- one of a number of diseases. Transient dis-
ing potential that acts as a reserve of energy orders do not permit a pathologic or etio-
poised for release when the valve is turned logic diagnosis. Any type of disease (vascular,
on. Ionic channels act as the valve, controlling neoplastic, inflammatory) may be associated
the energy in the ionic concentration gradi- with transient changes. Therefore, the pathol-
ent. The release of energy is seen either as ogy of a disorder cannot be deduced when
local graded potentials or as propagated action its temporal profile is solely that of transient
potentials that arise when local potentials reach episodes.
Chapter 6
Electrophysiologic Generators in
Clinical Neurophysiology
STRUCTURAL GENERATORS Brain Stem

Peripheral Nerves Special Sensory Receptors
Muscles Optic and Auditory Pathways
Sweat Glands Cerebral Cortex
Spinal Cord
SUMMARY
STRUCTURAL GENERATORS Schwann cells, connective tissue, and blood

vessels. They contain three types of axons.
The variety of clinical neurophysiologic studies Motor axons originate from neurons in the
corresponds to a variety of structural gener- spinal cord (spinal motor neurons) or brain
ators in the body, including muscles, sweat stem (cranial nerve motor neurons) and
glands, peripheral nerves, and various com- synapse on muscle fibers. Sensory axons orig-
ponents of the central nervous system. Each inate from neurons in the spinal dorsal root
structural generator may have associated with ganglia or cranial nerve ganglia; these axons
it several different types of physiologic poten- terminate in skin, muscle, or other organs
tial. In some cases, the activity resulting from in specialized sensory receptors, for example,
different physiologic potentials can be easily pacinian corpuscles or muscle spindles, or as
distinguished, but in many cases (particularly “bare” nerve terminals. Autonomic axons orig-
sensory pathways in the central nervous system inate either in neurons of the spinal cord
assessed by evoked potential studies), com- or brain stem (preganglionic neurons) or in
plete knowledge of the physiologic generator neurons in autonomic ganglia (postganglionic
underlying any recorded waveform is lacking. neurons located in the sympathetic trunk gan-
glia for many sympathetic neurons or in vis-
ceral ganglia near the end organ innervated
for all parasympathetic and some sympathetic
Peripheral Nerves neurons). The sum of the propagating action
potentials of all stimulated sensory axons in a
Peripheral nerves consist of axons and sup- motor or mixed nerve can be recorded as a
porting structures, including myelin-producing sensory nerve action potential during sensory
97
nerve conduction studies (see Chapter 17). for example, by the electromyographic nee-
These primarily involve the large diameter, or dle, of nerve terminals in the end plate region
IA and IB, sensory axons in the nerve, because and are mediated by normal neuromuscular
only they are stimulated by conventional elec- synaptic transmission.
trical stimuli. Generally, motor and autonomic Muscle action potentials are similar to
axons are tested only indirectly by stimulat- nerve action potentials but have a gener-
ing the nerve and observing the postsynaptic ally slower propagation velocity. They propa-
effects in muscle and sweat glands. gate outward—often in two opposite directions
simultaneously—from the vicinity of the neu-
romuscular junction to the ends of the muscle
Key Points fiber. Muscle action potentials may occur spon-
• Peripheral nerves contain motor axons, taneously in individual muscle fibers (fibrilla-
sensory axons, and autonomic axons. tion potentials), simultaneously in all muscle
• Sensory nerve action potentials are the fibers that are part of the same motor unit (e.g.,
sum of propagating action potentials of voluntary motor units or involuntary fascicula-
sensory axons, mainly IA and IB. tion potentials), or nearly simultaneously in all
muscle fibers supplied by one motor or mixed
nerve (leading to the surface-recorded com-
pound muscle action potential in motor nerve
Muscles conduction studies) (see Chapters 23 and 26).
Muscles consist of muscle fibers and connec-

tive tissue; both motor and sensory axons tra- Key Points
verse muscles. Contraction of muscle fibers is • Muscle end plate potentials are muscle
initiated by neuromuscular synaptic transmis- fiber EPSPs that originate at the end plate.
sion (with acetylcholine as the neurotransmit- • Miniature end plate potentials are caused
ter), which leads to a propagated muscle action by spontaneous, random release of a single
potential. This in turn causes calcium to enter packet of acetylcholine.
the muscle fibers. Calcium is the intracellular • End plate noise is caused by miniature
trigger for the contractile process. end plate potentials.
Muscle end plate potentials are muscle fiber • End plate spikes are action potentials of
excitatory postsynaptic potentials (EPSPs) that muscle fibers caused by mechanical acti-
originate at the end plate, or neuromuscu- vation.
lar junction. Both sodium and potassium ions • Fibrillation potentials are spontaneous
flow through the channel opened by acetyl- muscle action potentials in individual
choline receptor binding so that the reversal muscle fibers.
potential, that is, the membrane potential in • Voluntary motor units and fasciculation
a voltage clamp experiment at which the ionic potentials are summed action potentials
current flow reverses from net inward to net from all fibers in one motor unit.
outward, for this channel is near 0 V, interme- • Compound muscle action potentials are
diate between the Nernst potentials of sodium summed action potentials from all simul-
and potassium. Normal end plate potentials are taneously activated fibers.
caused by the simultaneous release of hun-
dreds of quanta, or packets, of acetylcholine,
which occurs when an action potential reaches
the nerve terminal. However, miniature end Sweat Glands
plate potentials are much smaller and are
caused by the spontaneous, random release of Sweat glands are end organs innervated by
a single packet of acetylcholine from the nerve sympathetic postganglionic neurons that have
terminal. End plate noise recorded from nee- acetylcholine as their neurotransmitter. Many
dle electrodes in the vicinity of the muscle end of these axons have multiple branches in
plate is caused by miniature end plate poten- the skin, and a single axon may innervate
tials. End plate spikes are action potentials of many sweat glands. Release of acetylcholine
muscle fibers caused by mechanical activation, by sympathetic nerve terminals in response to
Electrophysiologic Generators in Clinical Neurophysiology 99
various stimuli, for example, an electric stim- Spinal Cord

ulus applied to a contralateral limb, leads to a
prolonged depolarization of sweat glands that The spinal cord contains cell bodies of motor,
can be recorded diffusely from the skin sur- autonomic, and other neurons and axons of
face in certain areas. This recorded response is central and peripheral neurons arranged in var-
called the peripheral autonomic surface poten- ious ascending and descending tracts. Many
tial. Function of peripheral sympathetic axons synapses occur on spinal cord neurons. Thus,
and sweat glands is also assessed with the potentials originating from generators in the
quantitative sudomotor axon reflex test (see spinal cord include EPSPs, IPSPs, and action
Chapter 36). In this test, a controlled release potentials. Although the quantity and integrity
of acetylcholine into one skin region leads to of spinal motor neurons in the ventral horn can
reflex depolarization of autonomic axons that be assessed to some degree by standard elec-
propagate antidromically to a “Y” branch point tromyographic and nerve conduction studies,
and orthodromically from there to activate specific potentials produced by spinal motor
sweat glands synaptically in a nearby region of neurons themselves are not routinely mea-
skin, thus producing sweat. Sweat gland activity sured. During somatosensory evoked poten-
(sweat production) is quantified by measuring tial (SEP) studies, recordings from cervical or
the water output of the glands1 (Fig. 6–1). cervical–cranial derivations may demonstrate
an N11 waveform that is thought to be gen-
erated primarily at the root entry zone in the
Key Points dorsal horn at the level of spinal segments
• Sweat glands are innervated by sympa- C5 and C6 (see Chapter 18). This waveform
may be the propagated action potential in the
thetic postganglionic neurons that use
sensory axons ascending in the dorsal column
acetylcholine as their neurotransmitter.
• Activation of sympathetic nerve terminals rather than a response recorded from the roots
entering the cord, as indicated by the latency
releases acetylcholine causing prolonged
shift that occurs with recording at progressively
depolarization of sweat glands.
• The quantitative sudomotor axon reflex higher levels in the cord.
test provides a quantitative measure of
distal sudomotor axon function.
Key Points
• Generators in the spinal cord include
EPSPs, IPSPs, and action potentials.
• The N11 peak in SEP studies is generated
at the root entry zone in the dorsal horn of
the C5–C6 spinal segments.
Brain Stem
The brain stem somatosensory pathways include
the dorsal column nuclei in the lower medulla
and the medial lemniscus. Recordings during
Figure 6–1. Physiology of two postganglionic axon SEP studies demonstrate an N13 waveform
reflexes: The sudomotor axon reflex and the veno-arteriolar thought to be generated by the upper cer-
reflex. The former is the basis of the quantitative sudomo- vical cord dorsal columns (presynaptic action
tor axon reflex test, but the latter is not used in any clinical potential) or by the dorsal column nuclei (post-
test. (From Low, P. A. 1993. Quantitation of autonomic
function. In Peripheral neuropathy, ed. P. J. Dyck, P. K. synaptic potential) or by both. A P14 waveform
Thomas, J. W. Griffin, P. A. Low, and J. F. Poduslo, 3rd may also be seen; it is thought to be generated
ed., 731. Philadelphia: WB Saunders. By permission of the by the medial lemniscus (action potential) (see
publisher.) Chapter 18).
Key Points Optic and Auditory Pathways

• The N13 peak in SEP studies is generated
The optic and auditory pathways are gener-
by the upper cervical cord dorsal columns
ators of later components of the visual and
or by the dorsal column nuclei or by both. auditory evoked potentials. Propagated action
• The P14 peak in SEP studies is likely gen-
potentials in the optic tracts or optic radia-
erated by action potentials in the medial tions may contribute to some variable early
lemniscus. components of diffuse light-flash visual evoked
potentials, but these have little clinical useful-
ness (see Chapter 22). Wave II of the BAEP
is thought to be generated either by propa-
Special Sensory Receptors gated action potentials in the auditory nerve
as it enters the brain stem or by postsynap-
Sensory receptors in the visual and audi- tic potentials in the cochlear nucleus. Wave III
tory systems generate characteristic potentials may be generated by postsynaptic potentials in
that can be recorded with appropriate evoked the superior olivary nucleus and waves IV and
potential studies. V by the lateral lemniscus (propagated action
The retina contains photoreceptors, the rods potentials) or the inferior colliculus (postsynap-
and cones, and other types of neurons, includ- tic potentials).2 Later waves (VI and VII) may
ing bipolar, horizontal, amacrine, and gan- arise from the medial geniculate body and the
glion cells. Retinal visual evoked potentials, auditory radiations (see Chapter 19).
the electroretinogram, can be recorded from
electrodes placed near or on the eye and are Key Points
thought to be caused by summed postsynaptic • BAEP wave II is due to propagated audi-
potential activity in retinal neurons.
tory nerve action potentials entering brain
Hair cells are sensory receptors in the
stem or postsynaptic potentials in the
cochlea. They release a neurotransmitter that
cochlear nucleus.
activates the peripheral axons of the bipolar • BAEP wave III may be due to postsy-
cells of the spiral ganglion, which in turn con-
naptic potentials in the superior olivary
duct action potentials to the central axons that
nucleus.
make up the auditory nerve and synapse in • BAEP waves IV and V are due to propa-
the cochlear nucleus in the lower pons. Some
gated action potentials in the lateral lem-
components of the electrocochleogram, that
niscus or postsynaptic potentials in the
is, the cochlear microphonic and the summat-
inferior colliculus.
ing potential, are thought to be produced by • BAEPs waves VI and VII may arise from
sensory receptor potentials in the hair cells.
the medial geniculate body and the audi-
Wave I of the brain stem auditory evoked
tory radiations.
potential (BAEP) is caused by the propagated
action potentials in the auditory nerve (see
Chapters 19 and 21).
Cerebral Cortex
The cerebral cortex is the generator of
Key Points essentially all electroencephalographic (EEG)
activity recorded without averaging as well
• Retinal visual evoked potentials (elec- as the late components of evoked poten-
troretinogram) can be recorded from elec- tials, including the P100 component of the
trodes on or near the eye. pattern-reversal visual evoked potential (see
• The cochlear microphonic and summating Chapters 8, 18, 19, and 22). Cortical neu-
potential are summated sensory receptor rons include both pyramidal cells (excitatory
potentials of cochlear hair cells. neurons that provide the major output of the
• Wave I of the brain stem auditory evoked cerebral cortex) and stellate cells (excitatory or
potential is caused by propagated action inhibitory interneurons). The long apical den-
potentials in the auditory nerve. drites of pyramidal cells are perpendicular to
Electrophysiologic Generators in Clinical Neurophysiology 101
the cortical surface. Each pyramidal cell has oriented in many directions, all of which make
an extensive dendritic tree on which 1000– effective summation impossible.
100,000 synapses may occur.
The neocortex consists of six cellular layers. Key Points
Layer I is the most superficial and layer VI is • Cerebral cortex generates essentially all
the deepest layer. Layer I contains mainly glial
EEG activity and late components of
cells and axonal and dendritic processes. Layer
evoked potentials like the VEP P100.
IV is the most developed in sensory areas of the • Neocortex has six cellular layers, with
cortex and receives much of the specific tha-
layer I most superficial and layer VI most
lamocortical projections. Layer V is the most
deep.
developed in motor areas of the cortex in which • Primary visual cortex and visual associa-
many of the pyramidal cells are exceptionally
tion cortex (Brodmann areas 17, 18, and
large and project particularly to distant sites,
19) are generators of the VEP P100.
including the brain stem and spinal cord. • Postsynaptic potentials, not action poten-
Brodmann divided the cortex into 52 cortical
tials, in cortical neurons are responsible
areas on the basis of cell size, neuron density,
for all scalp-recorded electrical activity.
myelinated axon density, and number of layers.
The primary somatosensory cortex (areas 1, 2,
and 3) is the likely generator of the scalp com-
ponents of the SEP. Primary visual cortex (area SUMMARY
17) and visual association cortex (areas 18 and
19) are the likely generators of the P100 com- This chapter reviews the generators of electro-
ponent of the visual evoked potential. Auditory physiologic potentials in terms of basic cellular
cortex (areas 41 and 42) may be the genera- electrophysiology and the anatomical struc-
tor of the late components of the long-latency tures that generate electrophysiologic poten-
auditory evoked potential. tials of clinical interest. Knowledge of the gen-
It is known that postsynaptic potentials— erators of the potentials recorded in clinical
not action potentials—in cortical neurons are neurophysiologic studies is helpful in under-
responsible for all scalp-recorded electrical standing the characteristics and distribution of
activity. This is because postsynaptic poten- the recorded potentials and is the first step in
tials are of long duration (tens to hundreds correlating the alterations seen in disease states
of milliseconds), involve large areas of mem- with the pathologic changes demonstrated in
brane surface, occur nearly simultaneously in the underlying generators.
thousands of cortical pyramidal cells, and occur
especially in pyramidal cell dendrites that are
uniformly perpendicular to the cortical sur- REFERENCES
face. These properties all allow postsynaptic
potentials to summate effectively to produce a 1. Dyck, P. J., P. K. Thomas, J. W. Griffin, P. A. Low, and
detectable scalp potential. In contrast, action J. F. Poduslo, eds. 1993. Peripheral neuropathy, 3rd ed.
potentials are brief (1 ms), involve small sur- Philadelphia: WB Saunders.
2. Melcher, J. R., J. J. Guinan Jr., I. M. Knudson, and N.
face areas of membrane (axons), occur at ran- Y. Kiang. 1996. Generators of the brain stem auditory
dom and widely spaced intervals in various evoked potential in cat. II. Correlating lesion sites with
neurons, and propagate along axons that are waveform changes. Hearing Research 93: 28–51.
Chapter 7
Waveforms and Artifacts

Jasper R. Daube
INTRODUCTION PHYSIOLOGIC ALTERATION OF

CONTINUOUS WAVEFORMS WAVEFORMS
EVENT RECORDING Single Potential
Frequency Continuous Waves
Configuration Signal Display
COMBINATIONS OF ARTIFACTUAL WAVEFORMS
CONTINUOUS WAVEFORMS Physiologic Artifacts
AND EVENTS Nonphysiologic Artifacts
WAVEFORM ALTERATIONS
SUMMARY
INTRODUCTION commonly seen in electromyogram (EMG). It

must be noted that there is overlap in these
The waveforms that make up the electric sig- patterns such that an EEG can have superim-
nals generated by the central, peripheral, and posed events, for example, the spike discharges
autonomic nervous systems and muscle are in epilepsy, and an EMG can have a continu-
classified according to the variables that char- ously varying pattern when multiple waveforms
acterize them. These variables identify the sig- are superimposed.
nals and demonstrate the abnormalities that
occur in disease. Each waveform is a change
over time in the potential difference between CONTINUOUS WAVEFORMS
two recording points. If no change occurs in
the potential difference, a flat line with no sig- A continuously varying signal is described by
nal is recorded, even if there is a voltage differ- the rate of change of the signal (cycles per sec-
ence between the points. Changes in potential ond), the signal amplitude (peak-to-peak), the
difference are broadly classified as continuous shape of the waveform, and the consistency of
and intermittent. Continuous waveforms are the signal over time. Continuous waveforms
characterized by a relatively smooth, continu- recorded from living tissue are usually sinu-
ously varying appearance. These are most com- soidal, as shown in Figure 7–1. The rate of
monly seen in electroencephalogram (EEG) change of a signal is known as its frequency
and related studies. Intermittent waveforms and is measured in cycles per second. The
are single, discrete events on a flat background, term more commonly used to describe this rate
especially motor unit potential. These are most of alteration is hertz (Hz). A 10-Hz signal is
103
Figure 7–1. Simple, regular, sinusoidal waves of different frequencies combine to form complex waveforms. A, A 3-Hz
and a 20-Hz waveform summate into a waveform in which both are still recognizable. B, A 3-Hz and a 10-Hz waveform
summate to form a more complicated waveform in which the components are less recognizable. C, Summation of a 3-Hz
and a 6-Hz waveform results in an apparently regular, more complex waveform. D, Summation of a 3-Hz and a 3.5-Hz
waveform results in fluctuation of the waveform as the components go in and out of phase.
a waveform that has continuous variation 10 time characterizes the signal at a pair of
times per second. The traces in Figure 7–1 electrodes. The ability to dissect a waveform
have waveforms varying at five basic rates (3, into its component frequencies does not mean
3.5, 6, 10, and 20 Hz). The continuous wave- that distinct structures generate each of the
forms generated by physiologic generators can frequencies. For example, a neuron with
be described by their component frequencies.
The change of a continuous waveform from
one frequency to another may occur because
a single generator changes its rate of activ-
ity or because one generator working at 3 Hz
becomes inactive while another working at
6 Hz becomes active.
If the recording electrodes are located near
two structures that simultaneously generate
signals of different frequencies, a more com-
plex signal is recorded. Signals illustrating the
combination of two frequencies are shown
in Figure 7–1. Note that the combination
of signals of both frequencies is still recog-
nized if the frequencies are widely differ-
ent. Combinations of waveforms with sim-
ilar frequencies result in less recognizable
waveforms and even temporary obliteration
of the signal (3 and 3.5 Hz). Such complex
physiologic waveforms can still be described
in terms of the component frequencies of
the signal. Waveforms become even more
complex when differently shaped waves of
the same frequency summate, as shown in
Figure 7–2.
Continuously varying signals in physiologic
recordings are usually described in terms of
their frequency components, as illustrated in
Figure 7–2. Summation of waveforms with different
Figure 7–3. Many of them, such as elec- shapes but the same frequency (3 Hz) results in more com-
troencephalographic waveforms, often have plex waveforms. A sine wave is combined with (A) a square
a predominant frequency that at any given wave, (B) a triangular wave, and (C) a ramp wave.
Waveforms and Artifacts 105
ECG
Heart
EMG
muscle
EEG Muscle Nerve EMG
action
action action Fibrillation
potentials potentials potentials potentials potentials
R
P T α
QS βγ
Δθ
S–W
100
90 s
Ne
nd
d
le
co
on
on
rve 0.1
sc
80 se
ec
ec
mu
t=
2
G
mu ms
2s
0
5s
Amplitude, %
EC
t=
70
0.
EE s
t=
rve
sc
EC
EE
0.0
0.0
T=
20
le
6m
G
Ne
G
60
T=
ms
T=
50
Cathode-ray tube
40
30
Usual limit of Usual limit of
20
recording pens galvanometers
10
0
0.01 0.1 1.0 10 100 1000 5000 10,000
Frequency, cycles per second
Figure 7–3. The frequency components of electric activity recorded in clinical neurophysiology are shown on a log-
arithmic scale. T and t are the upper and lower time constants of each frequency cutoff. ECG, electrocardiogram;
EEG, electroencephalogram; EMG, electromyogram. (From Geddes L. A., and L. E. Baker. 1968. Principles of applied
biomedical instrumentation. 317. New York: John Wiley & Sons. By permission of the publisher.)
synaptic input from multiple sources displays single generator would be active at any of
postsynaptic potentials that summate at the cell the component frequencies. Frequency anal-
body and produce a complex, varying intra- ysis with automated electronic systems can
cellular potential. Many frequencies would define the frequency components of any sig-
be identified by frequency analysis, but no nal as a histogram, as shown in Figure 7–4.
1a 50 μV
2a 10 μV
1 second
10 pW
1b
1 pW
2b
0 5 10 15 20 Hz
Figure 7–4. Frequency analysis of, 1a, normal and, 1b, abnormal electroencephalographic recordings. The frequency
components of 1a are shown in 1b as a predominantly 10-Hz signal. Tracing 2a is a combination of the 10-Hz activity, with
the abnormal 1–4 Hz activity shown in 2b. (From Fisch B. J. 1991. Spehlmann’s EEG primer, 2nd ed. 132. Amsterdam:
Elsevier Science Publishers. By permission of the publisher.)
Frequency analysis of either the first or the EVENT RECORDING

second trace in A, B, C, and D in Figure 7–1
would show only one frequency component, Frequency
whereas frequency analysis of the third trace
in each panel would show two frequency com-
A second description of the waveforms seen in
ponents. A frequency analysis of this type
clinical neurophysiology is in recordings made
can be made on signals of different time
up of a sequence of well-defined events, as
duration.
shown in Figure 7–5. Recurrent, brief spikes
Therefore, a continuously varying signal is
in a recording that includes no other activity
described by its frequency components, their
are shown in Figure 7–5A. The description and
amplitudes, and the intervals of time over
classification of events use terminology similar
which they occur. This approach to describing
to that used for continuous recording, but with
signals is used most commonly with electroen-
some differences in meaning. In Figure 7–5A,
cephalographic potentials.1 New methods of
the spikes occurring in the first 5 seconds of
analysis of continuous potentials continue to the recording are described as occurring at a
be developed for signals from both EEG and frequency of 3 per second; those in the next
EMG.2–4 4 seconds have a frequency of 6 per second.
A description of neurophysiologic signals The entire recording of 37 events in 10 sec-
always includes the location at which the signal onds could be described as spikes occurring 3.7
is recorded. For example, the frontal region of times per second. However, this is an average
the head may have a 30 μV, 20 Hz signal occur- of their occurrences instead of the actual recur-
ring for 10 seconds of a 20-second recording, rence of 3 and 6 times per second. Therefore,
whereas the occipital region simultaneously events are better described as (1) the number
has a 10 Hz signal throughout the 20-second per unit time, (2) the regularity of their occur-
recording epoch. rence in time, and (3) the pattern of recur-
rence. In Figure 7–5C, groups of eight spikes
recur at 2-second intervals. This is described
Key Points
as a burst pattern of 10 per second spikes
• Continuous waveforms are mixtures of recurring regularly at 2-second intervals.
smoothly changing sinusoidal wave- A recording from a pair of electrodes often
forms. includes the activity of multiple generators,
• Continuous waveforms change appear- each of which may be generating events in
ance by different patterns. A single generator may gen-
◦ Change in the underlying generator’s erate regularly recurring events that occur in a
pattern changing but definable pattern or events that
◦ Addition of waveforms from other recur in an unpredictable, or random, pattern.
generators. In Figure 7–5B, a second independent genera-
• Continuous waveforms are recorded pri- tor becomes active after 6 seconds. Such super-
marily in electroencephalography, sleep imposition gives a complex pattern of events
studies, and autonomic studies. that can be separated into the recurrence of
Figure 7–5. Single spikes occur as events at different rates. A, a spike changes from 3 per second to 6 per second. B,
a spike occurring at 1 per second combines with another, independent spike occurring at 2 per second. C, a single spike
occurs in bursts of eight spikes at 10 per second every 2 seconds.
distinct events coming from different genera- monophasic, biphasic, triphasic, or more com-
tors. This distinction is made most accurately plex with multiple phases.11 If there is more
by analyzing the pattern of events.5 Also, recog- than one phase, each phase can be described
nizing the unique appearance of the individual according to its amplitude, duration, and con-
events can help to distinguish them, but this figuration. Each component of a discharge has
becomes unreliable when the individual wave- a rise time, or a rate of rise, from the positive
forms are similar. New methods continue to peak to the negative peak. The rise time is a
be developed for identifying and characteriz- direct function of the distance of the recording
ing intermittent events in both EEG and EMG electrodes from the generator and can be used
recordings.6–9 to determine how close a generator is to the
In summary, characterizing the occurrence recording electrodes. Short duration or rapid
of events requires (1) identifying the patterns rise times occur when the recording electrodes
of individual events as bursting or nonburst- are close to the generator.
ing, (2) describing those that occur in bursts A typical discharge is triphasic when it is
according to their rate of firing during a burst recorded from a nearby generator. If the wave-
and the recurrence rate of the burst, and (3) form is moving, the initial positive portion
describing those that are nonbursting by the of the discharge is recorded from the poten-
rate and pattern of firing. tial when the potential is distant from the
recording electrode. A negative component is
obtained when the potential is adjacent to the
Configuration active electrode, and the late positive compo-
nent when the potential leaves the electrode.
Each event is characterized further on the basis The complexity of a discharge is a function of
of its own variables.10 The event, often called the number of generators that contribute to
a discharge or spike, has an amplitude that is the discharge and the synchrony of their firing.
measured either from the baseline-to-peak or Greater synchrony of firing produces a simpler
from peak-to-peak (Fig. 7–6). The discharge waveform of larger size.
has a duration from onset to termination. Individual events can also be characterized
Discharges have configurations that may be according to their frequency spectrum. The
Turns
Amplitude
Satellite
Baseline crossing
Phase
Rise
time
Duration
Figure 7–6. Each single event, such as the spikes in Figure 7–5, can be described quantitatively by the variables illustrated
here. The same event could also be analyzed into its component frequencies, which could range from fewer than 1 Hz to
more than 10,000 Hz.
analysis of an event—regardless of its size, signal when repeated through one cycle could
shape, or configuration—can break it down be considered an event and characterized as
into a summation of the activity of different fre- such. In contrast, the designation of events and
quencies. Event recording is used in recording their patterns of recurrence and complexity of
from single axons and single neurons and in appearance occasionally require characterizing
clinical electromyography. In these settings, them in terms of frequency, amplitude, and
the measurements are generally those of the rate of recurrence.
pattern of firing and the characteristics of the
individual discharges. The frequency compo- Key Points
nent of such potentials is not a useful measure. • Continuous and intermittent patterns may
Key Points be superimposed.
• An EEG with spike discharges is mea-
• Intermittent waveforms are mixtures of sured by both the underlying sinusoidal
discrete waveforms recurring in a defined frequencies and the number, character,
pattern. and distribution of the spikes.
• Intermittent waveforms change in appear- • An EMG interference pattern is mea-
ance by sured by extracting the basic features of
◦ Alteration in the configuration of indi- the underlying events as turns (sudden
vidual events reversal of potential) and amplitude.
◦ Alteration in their rate and pattern of
recurrence
◦ Superimposition of multiple events. WAVEFORM ALTERATIONS
• Intermittent waveforms are recorded pri-
marily in electromyography. Abnormalities of the waveforms generated by
the central nervous system, nerves, or muscles
can be assessed only in terms of a change in
COMBINATIONS OF the waveform of a specific generator. No wave-
CONTINUOUS WAVEFORMS form itself can be defined as abnormal without
AND EVENTS reference to the generator. Normal waveforms
arising from one generator would be abnormal
The division of electric activity into continuous if they arose from a different generator. For
waveforms and events is somewhat arbitrary, example, the spike activity normally generated
and they may be found together, such as spike by muscle has features that are similar to those
discharges in a sinusoidal electroencephalo- of an epileptic spike generated by the cere-
graphic waveform. Such spikes are intermittent bral cortex. Therefore, alterations in waveform
and described using the criteria for intermit- must be considered in relation to the categories
tent events. The superimposition of multiple described above. In contrast, electric artifacts
events in an EMG discharge gives rise to a con- often have distinct waveforms that do not arise
tinuous signal that can be measured in terms from any physiologic generator. Artifacts are
of its frequency components called an interfer- best defined as electric activity of no clinical
ence pattern since individual components are significance originating from nonphysiologic
no longer recognizable. While an EMG pattern sources. Artifacts are considered separately in
can be described by its frequency components, the section Artifactual Waveforms.
it has not been helpful in clinical diagnosis. In
contrast, analysis of the amplitude and num-
ber of spike potential reversals per unit time PHYSIOLOGIC ALTERATION OF
in the interference pattern has provided useful WAVEFORMS
clinical information.
Similar measurements can be made with Changes in single potentials, such as a sin-
either spikes or continuous waveforms. The gle well-defined electromyographic (EMG)
frequency of a continuous waveform is the spike, are described by the characteristics
inverse of the duration of a cycle. For example, of the single event. In contrast, changes in
a 10-Hz signal has a 100-ms interval between continuously varying signals, such as electroen-
recurrences. The broad sine wave of the 10-Hz cephalographic (EEG) waves, are described by
the characteristics of a series of waves. The If a single potential recurs over time, another
distinction between a single event and contin- set of variables is measured, including stabil-
uous waves is not always clear, but usually can ity, rate, pattern, and the type of change that
be readily separated. The alterations of single occurs with time. The alteration of waveforms
potentials and waves are considered separately with disease is defined by the variable which is
below. outside the normal range. In disease, each vari-
able should be considered for measurement.
The methods for measuring these variables
Single Potential must be defined because the results can vary
with the method of measurement.
Electric activity generated by nerve or mus-
Key Points
cle tissue often appears as a single, discrete
event, a single potential, with no activity or • Single potentials (events) are described by
only unrelated activity around it. Single poten- their variation from normal in
tials may be normal or abnormal. Changes in ◦ Size (amplitude, area, and duration)
individual potentials are described by measur- ◦ Configuration (phases and turns) and
ing the variables of the potentials to determine changes in configuration over time
whether they are outside the normal range ◦ Frequency and pattern of recurrence
(Table 7–1). To describe single potentials, four ◦ Distribution of the potential and change
sets of variables are measured. in characteristics with location.
The first set describes the size and includes
amplitude (peak-to-peak or baseline-to-peak),
area, and duration of the potential. The sec- Continuous Waves
ond set describes the waveform configuration
and includes the rate of change of the compo- Much of the electric activity that is generated
nents of the potential, the number and timing by neural tissue occurs as continuously varying
of changes in the direction of the current flow, potentials that may persist over long periods.
and the components of the potential. The com- These potentials usually have a sinusoidal con-
ponents include the phases and turns of the figuration. Recurrent, single events recorded
potential. The third set describes the pattern at a considerable distance from the genera-
and frequency of occurrence of the potential. tor may also appear as continuously varying
A spike might occur at a regular, low rate waves. Continuous waves are characterized
(e.g., a voluntary motor unit potential) or as by variables similar to—but different from—
high-frequency, short bursts (e.g., a myokymic the variables that characterize single events
discharge). The fourth set describes the dis- (Table 7–2).
tribution or field of occurrence. For example, Variables that are used to measure the size
an epileptic spike might occur in the frontal of continuous waves include amplitude (peak-
lobe or in the temporal lobe. A waveform to-peak or base-to-peak), root mean square
may have different variables in different parts (square root of mean amplitude over time), and
of its field. A potential may be described by power (square of the amplitude). In most situ-
its relationship to other events, such as the ations, the major variable to measure is the fre-
latency of a response. Disease can alter the quency, or the number of cycles of the wave per
variables of existing waveforms, eliminate a second. Frequency can be measured simply as
normal waveform, or initiate a new waveform. baseline or zero crossings per second. More
Table 7–1 Measurable Variables of Single Potentials

Size Amplitude, area, duration
Configuration Rate of change, direction, number, and timing of
reversal of direction
Recurrence Rate, pattern, timing
Distribution Field, area, location
Relationship to other waveforms or events Time-locked, latency, order of interpotential interval
Stability Pattern and type of change with recurrence over time
Table 7–2 Measurable Variables of Continuous Waves

Size Peak-to-peak amplitude, root mean square, power
Frequency Cycles per second, zero crossing
Appearance Usually sinusoidal, frequency bands
Distribution Field or area, symmetry
Relationship to other waves Phase relation, synchrony
complex automated analyses of frequency phase relation. Waves may be in phase or out of
spectra are the fast Fourier transforms and phase. Measurement of the timing, frequency,
autoregressive modeling. Continuous waves and spatial distribution of the waves can pro-
may be simple, with a single frequency, or vide valuable information about the presence
they may be complex, with more than one and the stage of disease.
frequency contributing to the waveform. The
addition of multiple frequencies changes the Key Points
appearance of the wave from a simple sinu-
soidal pattern to a more complex, varying one. • Continuous waves are described by their
Continuous waves can be analyzed with regard variation from normal in
to their frequency components and the power ◦ Size (amplitude, root mean square, and
of each component. Polarity or direction is sel- power)
dom described because the waves are continu- ◦ Frequency (cycles per second, fre-
ous. Frequency analysis can provide a precise quency analysis, Fourier transform, and
measurement of the waveform, but it requires autoregression)
defining the amount of each of the compo- ◦ Distribution of the potential, phase
nent frequencies. This is sometimes done with relations among areas, and differences
frequency bands (Table 7–3). with location.
The distribution of continuous waves is
another important variable to measure. It is
usually described as broad areas, and compar- Signal Display
isons are made between homologous areas of
the body for symmetry. The relationship of The single potentials and continuous waves
continuous waves to other waves in the same or generated by neural and muscle tissues can
other areas is another important variable that be recorded as analog or digital signals. Mod-
is measured to identify alterations produced by ern equipment uses a digital format that allows
disease. Waveforms may occur in synchrony for the signals to be readily stored for subsequent
defined periods or may not be in synchrony review and analysis. This capability makes it
but still have a definable time relationship, the possible to analyze signals without displaying
Table 7–3 Voltages, Display Times, and Frequency of Common Signals

in Clinical Neurophysiology
Voltage (μV) Time (ms) Frequency (Hz)
Electromyography 50–1000 20–1000 32–16,000

Nerve conduction studies 1–20,000 10–500 1–8000
Electroencephalography 1–2000 5000–200,000 0.1–1000
Brain stem auditory evoked potentials 0.1–2 5–20 –
Somatosensory evoked potentials 0.1–20 50–200 20–3000
Visual evoked potentials 1.0–200 100–200 20–3000
Skin potentials 100–5000 1000–10,000 0.1–100
Electrocardiography 1000–5000 10,000–50,000 0.5–100
Respiratory movements 50–2000 5000–200,000 32–10,000
Electronystagmography 1000–5000 5000–100,000 1–2000
Electroretinography 1000–5000 500–2000 1–500
Vascular reflexes 1000–5000 1000–5000 0.1–100
the raw data, showing only the processed data. Physiologic Artifacts
Although this can improve the recording effi-
ciency, it has the risk of recording and ana- Physiologic artifacts are unwanted noise that
lyzing unwanted signals, such as the artifacts are the signals of interest in other settings.
described in the following section. Thus, it is These include (1) the electrocardiogram—a
preferable to display the raw, unprocessed sig- relatively high-amplitude, widely distributed
nal for review before proceeding to analyze potential generated by heart muscle that can
the information. The human eye and ear are interfere with any clinical neurophysiology
better than automated systems for recognizing recording; (2) EMG signals that accompany
artifact. For example, the raw signal recorded muscle contractions during EEG and evoked
during evoked potential testing should be dis- potential recordings; (3) potentials that occur
played along with the averaged potential dur- with the movement of electrically charged stru-
ing data collection. ctures (e.g., tongue movement, eye movement,
Unprocessed signals are best displayed as a or blink); and (4) autonomic nervous system
horizontal trace in which the horizontal axis potentials, such as those arising from changes
(sweep) is time and the vertical axis is volt- in skin impedance with perspiration.12–14 Com-
age change. The sweep speed and amplification mon artifacts are listed in Table 7–4. Each of
vary widely with the many different forms of these signals and other waveforms that may be
signals (Table 7–3). Multiple signals from dif- recorded to study a particular structure in one
ferent areas are often recorded simultaneously setting may be an artifact that interferes with
as vertically separated lines. the recording of a different signal in another
There are many formats for displaying pro- setting (Fig. 7–7). Although physiologic arti-
cessed data. The most common one is a line facts are phenomena that cannot be dissoci-
format, as used for averaged signals. Results ated from normal function, they must be cir-
may also be shown as histograms, bar graphs, cumvented as much as possible. For example,
numerical tables, topographic maps, or fre- decreasing the level of muscle activity can help
quency plots (e.g., compressed spectral arrays). circumvent EMG artifact. Another method is
Statistical analysis of the data is used with to filter out unwanted frequencies. In some
many of these displays. The assumptions of any cases, physiologic artifacts need only to be rec-
statistical analysis performed must be under- ognized and mentally discounted or subtracted
stood and appropriate for the problem to be electronically, for example, eye movement arti-
solved. fact on EEG records.
Key Points
Key Points • Physiologic artifacts must be recognized
• Clinical neurophysiologic recordings are and minimized in any recording, including
virtually all digital rather than analog. ◦ Muscle contraction EMG during EEG
• Primary raw data needs to be reviewed ◦ High-voltage EEG during evoked poten-
along with processed data to assure its tial recording
validity and recognize artifact. ◦ Electrocardiogram with any clinical
• Waveforms are best displayed as linear neurophysiology measurement
time data in addition to any other modified ◦ Skin impedance changes.
formats.
Nonphysiologic Artifacts
ARTIFACTUAL WAVEFORMS Nonphysiologic artifacts are from technical
sources, for example, the recording electrodes,
Artifacts are unwanted signals generated by the electric amplification and display system,
sources other than those of interest. They are electric stimulation, and the external elec-
not of clinical value. Artifacts can be classi- tric devices or wiring.15 The most common
fied as signals from living tissue, physiologic source of such artifacts is movement of the
artifact, or as signals from other sources, non- wires that connect the electrodes to the equip-
physiologic artifact. ment or movement of the electrodes on the
Table 7–4 Common Forms of Artifact and

Interference
Source Appearance
Movement of charged structures

Eye movement Slow positive, lateralized
Eye blink V-shaped positive
Tongue movement Slow positive
Eye flutter Rapid, rhythmic, alternating
Normal activation
Muscle potentials Rapid, recurrent spikes
Perspiration Very slow oscillation
Electrocardiogram Sharp and slow, regular
Dental fillings touching Short spikes
Transcutaneous stimulator 70–150 Hz spikes
Cardiac pacemaker 1-Hz spike
Paging and radio signals Intermittent, recognizable sound
Recording System
Electrode movement Irregular, rapid spike
Wire movement Irregular, slower waves
Poor electrode contact Mixture of rapid spikes and 60 Hz
Rubbing materials, static Sharp spikes
Display terminal 300 Hz, regular
White thermal noise Random, high-frequency spikes
Electromagnetic, external
Equipment 60 cycle Regular, 60 Hz
Switch artifact Rapid spikes
Diathermy Complex, 120 Hz
Cautery Dense, high-frequency spikes
Muscle difference in the very low impedance of the

metal and that of the skin result in the for-
ECG
mation of ionic layers with capacitance and
a junctional potential. Movement of the elec-
trodes causes a change in the junctional poten-
Pulse tial and capacitance that exist at the inter-
face of the electrode and the skin, resulting
in an artifact due to a local voltage. Simi-
Eye lar changes occur at the ground electrode. If
the ground is not well applied, it can gen-
erate artifacts, or more commonly a 60-Hz
Electrode
artifact is recorded. A 60-cycle artifact always
Figure 7–7. Rhythmic physiologic artifacts in elec- requires initial electrode removal for reclean-
troencephalographic (EEG) recordings. ECG, electro- ing of the skin and application of electrode
cardiogram.
gel. In general, the impedance of metal or
saline-moistened electrodes is lower than stick-
skin. The interface between the metal elec- on electrodes. Changing stick-on electrodes for
trode and the skin is complex; the epider- metal will often eliminate 60-Hz artifact. If
mal skin layers are relatively nonconducting. one of the electrodes is not well applied, the
Cleansing and abrading the skin and the appli- ground electrode becomes the reference with
cation of electrode paste with sodium and distortion of the evoked potential.
chloride concentrations that readily carry ionic Wire leads also frequently produce artifacts,
current between the skin act to reduce the especially if near 60 cycle generators or near
skin impedance. Nonetheless the remaining the stimulating electrode. Alteration in static
field or electromagnetic induction around the location of the equipment or its relationship
wire leads by movement can produce large, to the power source. At times, averaging can
slow wave artifacts. Artifacts also can arise from reduce activity if it is not time-locked to the
the opening and closing of switches on equip- stimulus. Differential amplification that is used
ment; from poor connections of the recording in all modern recording equipment markedly
electrodes, with high resistance of the elec- reduces external artifacts. Appropriate ground-
trodes; and from the use of dissimilar metals. ing can also help.17
Spurious signals generated within the record- Continuously occurring artifacts are some-
ing apparatus are usually a 60- or 300-Hz times referred to as noise and compared with
signal. the signal as a signal-to-noise ratio. This ratio
A shock artifact is virtually always seen if the determines the likelihood of eliminating the
recording electrodes are near the stimulating artifact by averaging the signal.
electrode, or if the leads to either of them are
near or touching each other. In some cases, Key Points
slow, gradual reorientation of the two stimu- • Nonphysiologic artifacts require recogni-
lating electrodes to each other will reduce the tion and knowledge of methods to elimi-
artifact by changing the current flow paths on nate or reduce them.
the skin. • Nonphysiologic artifacts arise from innu-
Several external power sources generate merable sources:
specific artifacts. Examples include the 60- ◦ Wires and electrode contact with skin
cycle signal caused by electromagnetic radia- ◦ Equipment malfunction
tion from power lines; the modified 60-cycle ◦ Surrounding equipment.
signal of fluorescent lights; the high-frequency,
complex discharges from cautery and diather-
mic equipment; and the irregular waveforms
from radio sources, and magnetic resonance
SUMMARY
imaging power16 (Fig. 7–8).
Artifacts can alter all of the variables used to
Artifacts sometimes are referred to as inter-
describe the continuous and discrete wave-
ference because they interfere with record- forms recorded in clinical neurophysiology.
ing the activity of interest. By recognizing the Changes in amplitude, frequency, and dis-
nature and source of an artifact, clinical neu-
tribution of waveforms occur in continuous
rophysiologists can often reduce or eliminate it waveforms. Frequency change may include the
by changing the electrodes or by changing the addition of new, abnormal frequencies, the loss
of normal frequencies, and either an increase
or a decrease in amplitude. Discrete events
themselves may be abnormal. The configura-
tion, distribution, size, and pattern of normally
occurring discrete events may be changed by
disease.
The waveforms recorded in clinical neuro-
physiology are divided into continuous wave-
forms and discrete waveforms, or events. Con-
tinuous waveforms are described by their fre-
quency components, amplitudes, and distri-
butions. Discrete waveforms are described
by their individual amplitudes, durations, and
configurations as well as by their patterns of
occurrence and distribution.
REFERENCES
1. Samar, V. J., A. Bopardikar, R. Rao, and K. Swartz.
Figure 7–8. Nonphysiologic artifacts recorded during 1999. Wavelet analysis of neuroelectric waveforms: A
surgical monitoring of muscle activity. conceptual tutorial. Brain and Language 66:7–60.
2. Paramanathan, P., and R. Uthayakumar. 2008. Appli- 9. Rasheed, S., D. Stashuk, and M. Kamel. 2006. Adap-
cation of fractal theory in analysis of human elec- tive fuzzy k-NN classifier for EMG signal decomposi-
troencephalographic signals. Computers in Biology & tion. Medical Engineering & Physics 28(7):694–709.
Medicine 38(3):372–8. 10. Dumitru, D. 2000. Physiologic basis of potentials
3. Abel, E. W., H. Meng, A. Forster, and D. Holder. recorded in electromyography. Muscle & Nerve
2006. Singularity characteristics of needle EMG IP 23:1667–85.
signals. IEEE Transactions on Biomedical Engineering 11. Dumitru, D., J. C. King, and D. F. Stegeman. 1998.
53(2):219–25. Normal needle electromyographic insertional activity
4. Arikidis, N. S., E. W. Abel, and A. Forster. 2002. Inter- morphology: A clinical and simulation study. Muscle &
scale wavelet maximum—A fine to coarse algorithm Nerve 21:910–20.
for wavelet analysis of the EMG interference pat- 12. Croft, R. J., and R. J. Barry. Removal of ocular artifact
tern. IEEE Transactions on Biomedical Engineering from the EEG: A review. Neurophysiologie Clinique
49(4):337–44. 30:5–19.
5. Kobayashi, K., C. J. James, T. Nakahori, T. Akiyama, 13. Picton, T. W., P. van Roon, M. L. Armilio, P. Berg, N.
and J. Gotman. 1999. Isolation of epileptiform Ille, and M. Scherg. 2000. The correction of ocular
discharges from unaveraged EEG by indepen- artifacts: A topographic perspective. Clinical Neuro-
dent component analysis. Clinical Neurophysiology physiology 111:53–65.
110:1755–63. 14. Jung, T. P., S. Makeig, C. Humphries, et al. 2000.
6. De Lucia, M., J. Fritschy, P. Dayan, and D. S. Removing electroencephalographic artifacts by blind
Holder. 2008. A novel method for automated source separation. Psychophysiology 37:163–78.
classification of epileptiform activity in the human 15. Wichmann, T. 2000. A digital averaging method
electroencephalogram-based on independent compo- for removal of stimulus artifacts in neurophysiologic
nent analysis. Medical & Biological Engineering & experiments. Journal of Neuroscience Methods 98:
Computing 46(3):263–72. 57–62.
7. Sugi, T., M. Nakamura, A. Ikeda, and H. Shibasaki. 16. Allen, P. J., O. Josephs, and R. Turner. 2000. A
2002. Adaptive EEG spike detection: Determination method for removing imaging artifact from continuous
of threshold values based on conditional probabil- EEG recorded during functional MRI. Neuroimage
ity. Frontiers of Medical & Biological Engineering 12:230–9.
11(4):261–77. 17. Tenke, C. E., and J. Kayser. 2001. A conve-
8. Rasheed, S., D. Stashuk, and M. Kamel. 2006. Adap- nient method for detecting electrolyte bridges in
tive certainty-based classification for decomposition of multichannel electroencephalogram and event-related
EMG signals. Medical & Biological Engineering & potential recordings. Clinical Neurophysiology 112:
Computing 44(4):298–310. 545–50.
SECTION 2
ELECTROPHYSIOLOGIC
ASSESSMENT OF NEURAL
FUNCTION
PART A
Assessment of Cortical Function
In clinical neurophysiology, neural function is episodes and to define their nature, charac-
assessed by measuring the electric potentials ter, and spread. Some abnormalities may not
generated by neural tissue and the changes in be detected with a standard 30- to 60-minute
these potentials and waveforms produced by EEG recording. For abnormal electric activ-
disease. The characteristics of the waveforms ity that occurs only in an outpatient setting
and their alteration with disease are a function or under specific circumstances, ambulatory
of the neural generators producing the wave- recordings from a few scalp electrodes are
form. A particular modality of recording in made continuously during activities at home
clinical neurophysiology reflects only the alter- or work (Chapter 12) to obtain a full picture
ation in the area of the nervous system gen- of their nature. To help define the nature and
erating the activity. For example, routine elec- origin of frequent seizures, a patient under-
troencephalography (EEG) records the wave- goes prolonged EEG recording with multi-
forms arising from cerebral cortex using elec- ple electrodes left in place for several days
trodes applied to the scalp. The EEG record- (Chapter 13). Electroencephalograms can also
ings described in Chapters 8–11 reflect the be recorded in other specialized situations,
normal and abnormal EEG findings and the such as in the intensive care unit or oper-
disease processes that directly involve the cere- ating room, or with computerized quantita-
bral cortex. The EEG records the ongoing tion, as described in Chapter 14. This chapter
spontaneous electric activities of the cerebral also reviews additional information that can
cortex and the cortical response to external now be obtained from magneto-EEG. Patients
stimuli. These patterns of responses can pro- being considered for epilepsy surgery require
vide important clues to the underlying disease highly specialized recordings, including new
process. correlations with magnetic resonance imaging
Variations of standard EEG recordings that (Chapter 15). Cortical function can also be
provide unique information for specific situ- assessed with potentials that occur before a
ations have been developed. Longer record- planned movement or in response to external
ings are needed to document infrequent stimulation (Chapter 16).
117
Chapter 8
Electroencephalography: Adult,
Normal, and Benign Variants
INTRODUCTION NORMAL EEG ACTIVITY OF

RECORDING THE ADULTS
ELECTROENCEPHALOGRAM Awake State
DISPLAY OF EEG ACTIVITY The EEG in Older Adults
ACTIVATION PROCEDURES Drowsiness
Sleep State
Hyperventilation
Photic Stimulation BENIGN VARIANTS
Sleep Variants During Wakefulness
Other Activating Procedures Benign Variants During Drowsiness
and Sleep
ARTIFACTS
SUMMARY
INTRODUCTION cortical neurons. The EEG is performed by

attaching small disc electrodes on the scalp
Electroencephalography is a dynamic electro- of the head or utilizing an electrode cap. The
physiologic procedure that evaluates the func- electrodes are connected to the EEG instru-
tion or disturbance of function of the brain by ment which amplifies the brain wave potentials
recording the electrical activity of the brain. and activity by a million times and then dis-
Purpose and Role of the EEG in the Normal plays it on a moving strip of paper or on a
Adult computer reader station if a digital system is
used.
• To evaluate the function of the brain.
• To describe normal EEG and benign vari-
ants.
DISPLAY OF EEG ACTIVITY
The EEG activity is displayed using a variety
RECORDING THE of combinations of electrode pairs or mon-
ELECTROENCEPHALOGRAM tages. The American Clinical Neurophysiology
Society has recommended that all EEG labora-
The electroencephalogram (EEG) records the tories include a minimum number of standard
electrical activity and waveforms generated by montages for routine EEG recording.1 The
119
minimum recommendations are longitudinal • A combination of different montages may

bipolar (Fig. 8–1A), transverse bipolar (Fig. 8– be needed to best display the finding or
1B), and referential displays (Fig. 8–1C). Since abnormality in the EEG.
one type of montage may be insufficient for • Digital EEG recording has been helpful in
the problem, it is advisable to use a combina- selecting various montages.
tion of montages and to emphasize whatever
array is most useful for the problem1–3 (Fig. 8–
1D). One should be able to select the most ACTIVATION PROCEDURES
advantageous montage to best display the type
of activity or abnormality in the EEG. The Activation procedures are commonly used to
advent of digital EEG recording has proved help bring out abnormal activity such as epilep-
to be very helpful for the electroencephalog- tiform discharges when the resting record
rapher to reformat and select any of several is noncontributory. The activating procedures
different montages for the most appropriate consist of hyperventilation, photic stimulation,
display of a particular segment of the EEG2 and sleep recordings.2, 3
(see Chapter 4).
Key Points
Hyperventilation
• Various combinations of electrode pairs
constitute montages. Hyperventilation is performed for 3–5 min-
• The minimum number of standard mon- utes. In adults, this usually produces little
tages for an EEG recording should change in the EEG. If there is a change,
include longitudinal, transverse, and ref- this usually consists of generalized slowing
erential montages. or bursts of slow waves, which is sometimes
A 1
Age: 4 2 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
200 μV
1 sec
Figure 8–1. EEG from a 4½-year-old girl showing the appearance of focal right frontocentral (F4 ,C4 ) spikes in, A, a lon-
gitudinal (anteroposterior) bipolar montage, B, a transverse bipolar montage, C, a referential montage, and, D, a montage
combining bipolar and referential recording.
B 1
Age 4 2 yrs
FP1-FP2
F7-F3
F3-FZ
FZ-F4
F4-F8
A1-T3
T3-C3
C3-CZ
CZ-C4
C4-T4
T4-A2
T5-P3
P3-PZ
PZ-P4
P4-T6
O1-O2
200 μV
1 sec
C 1
Age 4 2 yrs
FP1-A1
FP2-A2
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
P4-A2
O1-A1
O2-A2
F7-A1
F8-A2
T3-A1
T4-A2
T5-A1
T6-A2
200 μV
1 sec
Figure 8–1. (Continued).
121
D 1
Age: 4 2 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
FZ-CZ
CZ-PZ
FP2-F4
F4-C4
C4-P4
P4-O2
FZ-A1
F4–
CZ –
C4–
PZ –
P4–
200 μV
1 sec
referred to as a buildup. Slowing is usually • The slowing may be more prominent in

more prominent in children when compared children.
to adults (Fig. 8–2). The degree of the slow- • The degree of slowing depends on
ing depends on age, the vigor of hyperventila- age, vigor of hyperventilation, and hypo-
tion, whether the patient is hypoglycemic, and glycemia.
posture. The response may be potentiated by • Hypoglycemia can potentiate the degree
hypoglycemia. Abnormal responses to hyper- of slowing with hyperventilation.
ventilation include epileptiform discharges, the • Hyperventilation can activate epilepti-
most common being 3-Hz generalized spike- form activity, with the most common
and-wave discharge that is usually seen with being 3-Hz generalized spike-and-wave
typical absence seizures. Focal or lateralized discharges.
slowing or asymmetry of activity that is brought
on by hyperventilation is also an abnormal
finding.
Photic Stimulation
Key Points Intermittent photic stimulation is another acti-

vating procedure. Stimulation usually includes
• Hyperventilation may produce general- frequencies ranging from 1 to 30 Hz and is
ized slowing. usually done with the eyes open and then
Electroencephalography: Adult, Normal, and Benign Variants 123
M Age: 6
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
At rest 50 μV After hyperventilation 2½ min 300 μV
1 sec 1 sec
Figure 8–2. EEG from a 6-year-old boy awake and at rest (segment on left) and normal response to hyperventilation
(segment on right).
closed. The normal response consists of flash- Sleep

evoked responses over the posterior head
regions, termed photic driving. An abnormal Sleep recordings can be an important way
response is a consistent asymmetric photic of bringing out focal or generalized epilep-
driving response or activation of epilepti- tiform activity. In particular, discharges aris-
form discharges. One characteristic response ing from the temporal region are frequently
is termed the photoparoxysmal response and activated during drowsiness and sleep. Par-
is an abnormal response, which consists of tial sleep deprivation the night prior to the
generalized and bilaterally synchronous spike- recording may increase the chance of detecting
and-wave discharges evoked by the photic epileptiform activity.
stimulus.2–4 This is usually seen in patients
with a generalized seizure disorder. The other Key Points
response is the photomyogenic response, which
consists of myogenic twitching of the fore- • Sleep recordings are an important way to
head or eyelids in association with the photic activate epileptiform discharges.
stimulus.2–4 • Discharges arising from the temporal
region are frequently activated during
drowsiness and sleep.
• Partial sleep deprivation may increase the
Key Points chance of detecting epileptiform activity.
• Photic stimulation evokes flash-evoked
responses over the posterior head regions,
termed photic driving. Other Activating Procedures
• An abnormal response consists of activa-
tion of epileptiform activity or an asym- Other activating procedures can be used to
metric driving response. evoke the patient’s symptoms if the patient is
• A photoparoxysmal response, which con- sensitive to a certain stimulus such as light,
sists of generalized spike-and-wave dis- sound, music, or somatosensory stimulation.3, 5
charges, is an abnormal response that is
seen in patients with a generalized seizure
disorder. ARTIFACTS
• A photomyogenic response consists of
myogenic twitching of the forehead or An artifact is a potential or waveform that is
eyelids. not due to cerebral activity. Artifacts are often
FP1-F3
Age: 47 yrs
F3-C3
C3-P3
FP1 FP2
F3 F4
P3-O1
C3 C4
P3 P4
FP2-F4
O1 O2
F4-C4
50 μV
1 sec C4-P4
P4-O2
Figure 8–3. Normal EEG from a 47-year-old man showing symmetric alpha rhythm predominantly in the occipital
regions (O1 and O2 ).
seen in the EEG. Artifacts can take a variety head regions when the person is awake and
of forms and shapes, including those resem- relaxed and has the eyes closed (Fig. 8–3); it
bling cerebral EEG activity. It is important is attenuated by eye opening, alerting stimuli,
to recognize the artifacts and not misinterpret or attention.7, 8 The usual alpha amplitude in
them as cerebral activity.2, 3, 6 At times muscle an adult is 15–50 μV. The maximal amplitude
and movement artifacts can obscure the EEG occurs over the occipital region, with variable
recording. spread to the parietal, temporal, and, at times,
central leads. The alpha activity may be of
Key Points higher voltage and wider distribution over the
right hemisphere.
• Artifacts are potentials that are not due to The alpha rhythm should attenuate bilater-
cerebral activity. ally with eye opening, alerting stimuli, or men-
• Artifacts can take any shape or form. tal concentration. Failure of the alpha rhythm
to attenuate on one side with either eye open-
ing or mental alerting indicates an abnormality
NORMAL EEG ACTIVITY OF on the side that fails to attenuate.9
ADULTS
Key Points
Awake State • Alpha activity is any activity between 8 and
13 Hz.
The activity seen in the EEGs of awake adults • The alpha rhythm is present over the pos-
consists of frequencies in the alpha and beta terior head regions.
range, with the alpha rhythm constituting the • The alpha rhythm occurs in the awake
predominant background activity. state with the eyes closed.
• The alpha rhythm is attenuated by eye
opening, alerting stimuli, or attention.
ALPHA RHYTHM
Alpha activity refers to any activity in the BETA ACTIVITY
range between 8 and 13 Hz, whereas the
alpha rhythm is a specific rhythm consisting Beta activity has a frequency greater than
of alpha activity occurring over the posterior 13 Hz.2 The average voltage is between 10 and
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
Age: 14 yr
50 μV
1 sec
Figure 8–4. EEG containing beta activity which is maximal over the anterior head regions in a 14-year-old girl receiving
diazepam.
20 μV7, 8 (Fig. 8–4). The three main types of can be seen in adults during drowsiness, occur-
beta activity, based on distribution, are the fol- ring in a generalized distribution or over the
lowing: (1) the precentral type occurs predom- posterior head regions. In older patients, theta
inantly over the frontal and central regions, components can occur as single transients or
which is increased with drowsiness, and may as part of a mixed alpha–theta burst over the
attenuate with bodily movement; (2) posterior temporal regions.7
dominant beta activity can be seen in chil-
dren up to 1–2 years old; it is also enhanced Key Points
by drowsiness; and (3) generalized beta activ- • Theta activity is in the range of 4–7 Hz.
ity which is induced or enhanced by certain • It is part of the background activity in
drugs, such as benzodiazepines and barbitu-
young children.
rates. Focal accentuation of beta activity can • It is seen in adults during drowsiness.
result from a lesion or defect in the skull. • Theta activity can be seen over the tempo-
ral regions in older adults.
Key Points
• Beta activity is greater than 13 Hz. DELTA ACTIVITY
• There are three main types of beta
Delta activity (<4 Hz) is the predominant
activity:
◦ Precentral beta background frequency seen in infants.3, 7 Delta
◦ Posterior beta in children slowing can be seen in adults as a normal
◦ Generalized beta which is seen with finding in deeper levels of sleep.7
drugs. Key Points
• Delta activity is less than 4 Hz.
THETA ACTIVITY • Delta activity is seen in infants.
Theta activity (4–7 Hz) constitutes part of the • Delta activity can be seen in adults during
background activity in children.7 Theta activity deeper levels of sleep.
MU RHYTHM most prominent component usually surface-

positive (Fig. 8–6). The amplitude is usually
The Mu rhythm consists of arciform waveforms between 20 and 50 μV, and the duration is
with a frequency of 7–11 Hz, which occurs approximately 100–250 ms. Lambda waves are
independently over the central head regions.7, 8 bilateral and synchronous but may be asym-
Mu activity is functionally related to the sen- metrical.
sorimotor cortex and is attenuated by touch,
active or passive movement of the extremi-
ties, or thought of movement (Fig. 8–5). It can Key Points
occur in an asymmetrical fashion or predomi- • Lambda waves resemble the Greek
nate over one hemisphere. Mu rhythm can be letter λ.
quite prominent if there is an overlying skull • They are present over the occipital regions
defect (breach rhythm). when the person is scanning a picture.
• The lambda waves represent an evoked
Key Points response to visual stimuli in association
• Mu activity has a frequency range of 7– with eye movements.
11 Hz.
• Mu activity is attenuated by touch, move-
ment of the extremities, or thought of The EEG in Older Adults
movement.
• Mu activity can be quite prominent with In older adults, there may be a shift to slower
an overlying skull defect. background frequencies. However, it has been
shown that the alpha frequency can continue
LAMBDA WAVES to be greater than 8 Hz in normal elderly sub-
jects.7, 10, 11 There is a tendency, however, for the
The lambda wave has a configuration resem- alpha rhythm to be of lower voltage in older
bling the Greek letter λ and occurs over the subjects and perhaps to show less reactivity.10
occipital regions when the subject actively Benign temporal slow wave transients con-
scans a picture.7 Lambda waves appear to rep- sist of sporadic delta and theta slow waves
resent an evoked cerebral response to visual that occur over the temporal regions in older
stimuli produced by scanning movements of adults, usually after the age of 60 years.10 The
the eyes when looking at a picture. The wave- temporal slow waves usually have a left-sided
forms are monophasic or diphasic, with the preponderance and appear to be related to
M Age: 75
Fp1-F3
Fp2-F4
F3-C3
F4-C4
C3-P3
C4-P4
P3-O1
P4-O2
Eye opening Wiggle thumbs
50 μV
1 sec
Figure 8–5. Normal EEG showing bilateral mu rhythm in the central regions; the rhythm persists when the eyes are
opened and attenuates with movement of the thumbs. This is in contrast to the alpha rhythm (O1 and O2 ), which is
attenuated by eye opening.
Age: 43 yrs
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Eyes opened Eyes closed 50 μV

1 sec
Figure 8–6. EEG showing normal lambda waves maximal in O1 and O2 when the patient’s eyes are open and looking
around the room.
a normal aging process.2, 10, 11 Hyperventilation occur. Mu activity may also be seen during
and drowsiness facilitate the appearance of drowsiness and may persist after the alpha
temporal slow waves. rhythm disappears.
Key Points Key Points

• In older adults, the EEG may shift to a • Drowsiness is associated with slowing, dis-
slower background frequency. appearance of alpha activity, and enhance-
• The EEG may show lower voltage activity.
ment of theta and beta activity.
• Temporal slow wave transients may be • Wicket waves may be present during
present in older adults. drowsiness.
• Hyperventilation and drowsiness can facil- • Mu activity can persist during drowsiness.
itate temporal slow wave transients.
Drowsiness Sleep State
In adults, drowsiness is typically associated Sleep activity consists of slow waves, spindles,
with slow eye movements, slowing of the back- V waves, K complexes, and positive occipital
ground frequency, a disappearance of alpha sharp transients of sleep (POSTS).
activity, and enhancement of beta activity. At
times, rhythmic 5–7 Hz theta activity can be
SLEEP SPINDLES
present, occurring in a generalized fashion or
over the anterior or posterior head regions. In adults, the sleep spindles of NREM sleep
In older subjects, there may be an enhance- usually have a frequency of 14 Hz and occur
ment of theta and delta waves over the tem- in a symmetrical and synchronous fashion
poral regions. Sharply contoured waveforms, over the two hemispheres in the frontocen-
called wicket spikes or wicket waves, may also tral regions (Fig. 8–7). In deeper levels of
be present over the temporal regions, maxi- sleep, the spindle frequency decreases to
mal over the left temporal region.5, 12 The beta approximately 12 Hz and maximal amplitude
activity over the frontocentral regions often is located more anteriorly. More continuous
increases in prominence during drowsiness. spindle activity may be seen in some patients
This usually has a frequency of 16–20 Hz, but who are receiving drug therapy, particularly
occasional bursts of faster frequencies may benzodiazepines.
Age: 20 yrs
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
P4-A2
50 μV
1 sec
Figure 8–7. EEG from a 20-year-old woman during sleep showing normal 14-Hz sleep spindles.
Key Points K-COMPLEX

• Sleep spindles have a frequency of 12– The K-complex is a diphasic or polyphasic wave
14 Hz. that is maximal at the vertex and is usually
• Spindles occur maximally over the fronto- longer than 500 ms. It is frequently associated
central regions. with spindle activity. The K-complex repre-
• More continuous spindle activity can be sents a nonspecific response to afferent stim-
seen with certain drugs. ulation and is generally linked to the arousal
mechanism.7
VERTEX SHARP TRANSIENTS
Vertex sharp transients (V waves) are sharp- Key Points
contoured transients that occur maximally • K-complexes are broad diphasic wave-
over the central vertex region during sleep forms that are maximal at the vertex
(Fig. 8–8). In children and young adults, V region.
waves may have a sharp or spiky appearance • K-complexes are usually longer than
and attain high voltages. They are typically 500 ms.
symmetrical in the central leads but may show • K-complexes represent a nonspecific arousal
transient asymmetries at the time of sleep response.
onset. F waves, or frontally dominant V waves,
are often broader than the centrally dominant
V waves and may extend asymmetrically into POSTS
the lateral frontal regions. Positive Occipital Sharp Transients of Sleep
Key Points (POSTS) are sharp-contoured, surface-positive
transients that occur singly or in clusters over
• Vertex waves are sharp-contoured tran- the occipital regions (Fig. 8–9). They are usu-
sients that occur maximally over the cen- ally bilaterally synchronous but may be some-
tral vertex region. what asymmetrical. They are predominantly
• F waves are frontally dominant V waves. seen during light-to-moderate levels of sleep
F Age: 34
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
P4-A2
O1-A1
O2-A2
50 μV
1 sec
Figure 8–8. EEG from a 34-year-old woman during sleep showing normal V waves maximal in C3 and C4 and P3 and P4 .
Age: 18 yrs
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
50 μV
1 sec
Figure 8–9. EEG during sleep showing prominent normal POSTS maximal in O1 and O2 .
and should not be mistaken for abnormal sharp OLDER ADULTS

waves.
Changes during drowsiness and sleep occur
in normal elderly subjects.10, 11 During drowsi-
Key Points
ness, rhythmic trains of high-amplitude delta
• POSTS are sharp-contoured, surface- activity may occur in older adults and are
positive waves that are seen over the often maximal over the anterior head regions.
occipital regions during sleep. This drowsy pattern of older subjects needs
to be carefully distinguished from abnor- • The fast alpha variant has a frequency
mal frontal intermittent rhythmic delta activ- twice the rest alpha frequency.
ity.2, 3, 10, 11 During NREM sleep, the V waves
and K-complexes are lower in amplitude and PHOTIC RESPONSES
less sharp in appearance than in young adults,
and sleep spindles are less prominent. Delta Complex waveforms may be induced by photic
activity is lower in voltage and less abundant stimulation when harmonics or subharmonic
in older subjects than in young adults during components are admixed with the fundamental
deeper stages of NREM sleep. frequency of the driving response. Occasion-
ally, the resultant mixture of frequencies can
RAPID EYE MOVEMENT SLEEP produce waveforms that simulate epileptiform
spikes or spike-wave complexes.
During rapid eye movement (REM) sleep, the
EEG shows a low-voltage pattern that has Key Points
some similarities to an awake pattern when the
• Complex waveforms may be induced
eyes are open. The EEG also shows intermit-
tent rapid eye movements. In addition, rhyth- by photic stimulation and can produce
mic groups of saw-toothed waves may occur waveforms that simulate spike-wave com-
intermittently over the frontal and central leads plexes.
and may precede or occur in association with
the rapid eye movements. SUBCLINICAL RHYTHMIC
ELECTROGRAPHIC DISCHARGE
Key Points OF ADULTS
• REM sleep is associated with a low- Subclinical rhythmic electrographic discharge
voltage pattern. of adults (SREDA) is an uncommon phe-
• Saw-toothed waves may be seen in associ- nomenon that occurs mainly in the older
ation with rapid eye movements. patient.13, 14 The pattern consists of a mixture
of theta and delta frequencies, but most often
predominating in the theta frequency range.
BENIGN VARIANTS It resembles an epileptiform seizure discharge
but is not accompanied by any clinical symp-
toms and has no significance for the diagnosis
Variants During Wakefulness of epileptic seizures (Fig. 8–10). The character-
istics of SREDA are listed in Table 8–1.
ALPHA VARIANTS
The alpha-variant patterns consist of activity BREACH RHYTHM
over the posterior head regions.13 This activ-
ity has a harmonic relationship to the alpha Various normal rhythms are more prominent
rhythm and shows reactivity and a distribu- when recorded over a skull defect. The term
tion similar to those of the alpha rhythm. breach rhythm has been used to refer to a focal
The slow alpha variant appears as dicrotic or increase in the amplitude of sharp-contoured
notched waveforms that result from a subhar- EEG activity over or near the area of a skull
monic component of the alpha rhythm, usually defect.15, 16 When mu and beta rhythms are
in the range of 4–5 Hz. The fast alpha variant present, the activity can resemble epileptiform
contains a frequency twice that of the resting spikes (Fig. 8–11).
alpha activity, usually between 18 and 20 Hz.
Key Points
Key Points
• Breach rhythm refers to a focal increase
• Alpha-variant patterns have a harmonic in the amplitude of EEG activity over the
relation to the alpha rhythm. area of a skull defect.
• The slow alpha variant is a subharmonic • Benign variants seen in the awake EEG
of the alpha rhythm with a frequency of consist of
4–5 Hz. ◦ Slow and fast alpha variants
F Age: 57 50 μV 75 μV
Fp1-A1
Fp2-A2
T3-A1
T4-A2
C3-A1
C4-A2
O1-A1
O2-A2
During hyperventilation How do you feel? 1 sec
“OK”
Figure 8–10. Onset of SREDA in a 57-year-old woman.
Table 8–1 Characteristics of Subclinical

Rhythmic Electrographic Discharge of Adults
(SREDA)
Feature Characteristic
Onset Segmented or abrupt

Repetition rate Theta
Duration Average 1 minute
Distribution Maximal parietal–posterior temporal
Laterality Symmetrical or asymmetrical
State Awake at rest or during hyperventilation
Background activity Often visible during the discharge
Delta aftermath None
Clinical accompaniment None
Patient age Mainly older adults
Age: 21 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
Skull defect Rt. central region 50 μV

1 sec
Figure 8–11. EEG from a 21-year-old man showing breach rhythm (C4) over the area of a skull defect.
131
◦ Spike-like photic responses 14 AND 6 HZ POSITIVE BURSTS

◦ Subclinical rhythmic electrographic dis-
The 14 and 6 Hz positive bursts (previously
charge of adults (SREDA)
◦ Breach rhythm where there is an known as 14 and 6 per second positive spikes)
are displayed best on long interelectrode
enhancement of activity over the area of
distance referential montages and are most
a skull defect.
prominent over the posterior temporal region
during light sleep. As the name implies, the
Benign Variants During bursts occur at a rate of 14 Hz (Fig. 8–13A)
Drowsiness and Sleep or between 6 and 7 Hz (Fig. 8–13B) and are
from 0.5 to 1 second in duration. The 14 and
There are several benign variants that occur 6 Hz positive bursts usually occur indepen-
mainly during drowsiness or light sleep.2, 13 dently over the two hemispheres and vary from
Some of these have an appearance suggestive side to side in occurrence. They are most fre-
of epileptiform abnormality, but are of little or quently seen in subjects between 12 and 20
no clinical significance and have no importance years old.2, 13
for the diagnosis of seizures or cerebral lesions.
Key Points
RHYTHMIC TEMPORAL THETA • 14 and 6 Hz positive bursts are most
BURSTS OF DROWSINESS prominent over the posterior head regions.
• 14 and 6 Hz positive bursts are seen most
The rhythmic temporal theta bursts of drowsi-
frequently in teenagers.
ness (previously known as the psychomotor-
variant pattern) frequently have a flat-topped
or notched appearance because of the harmon-
ics of the fundamental theta frequency. The BENIGN SPORADIC SLEEP SPIKES
bursts may occur bilaterally or independently Benign sporadic sleep spikes (BSSS), also
over the two temporal regions, with a shifting known as small sharp spikes (SSS) or benign
emphasis from side to side (Fig. 8–12). This epileptiform transients of sleep (BETS), occur
pattern differs from a true seizure discharge in mainly in adults during drowsiness and light
that it does not evolve into other frequencies or levels of sleep2, 13 (Fig. 8–14). They are usu-
waveforms.13 It occurs predominantly in young ally low-voltage, short-duration diphasic spikes
or middle-aged adults. with a steep descending limb. The BSSS occur
typically as single spikes, rarely as doublets,
Key Points
and never in repetitive trains. The BSSS may
• Rhythm temporal theta bursts of drowsi- have a single low-voltage aftercoming slow-
ness have a notched appearance and occur wave component, but they do not distort the
independently or bilaterally over the tem- background and are not associated with slow-
poral regions. wave activity, as temporal sharp waves are.
F: Age: 43 yrs
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
50 μV
1 sec
Figure 8–12. EEG from a 43-year-old woman showing rhythmic temporal theta activity during drowsiness.
A
Age: 12 yrs
F7-A1
F8-A2
T3-A1
T4-A2
T5-A1
T6-A2
50 μV
1 sec
B
Age: 12 yrs
F7-A1
F8-A2
T3-A1
T4-A2
T5-A1
T6-A2
50 μV
1 sec
Figure 8–13. EEG from a 12-year-old boy during sleep. A, A 14-Hz positive burst (maximal in T5 and T6 ) and, B, 6-Hz
positive bursts (maximal in T5 and T6 ).
They are best seen with long interelectrode dis- because the BSSS have no significance for the
tance derivations, including the temporal and diagnosis of epileptic seizures.2, 13
ear leads. Provided a long enough recording is
obtained, they almost always have a bilateral
representation, occurring either independently 6-HZ SPIKE-AND-WAVE
or synchronously over the two hemispheres. The 6-Hz spike-and-wave pattern has also been
Their characteristics are summarized in called the fast spike-and-wave, because of its
Table 8–2. They need to be carefully distin- repetition rate, or the phantom spike-and-wave,
guished from more important types of spikes because of its usual low-amplitude spike
Age: 66
F7-A1
F8-A2
T3-A1
T4-A2
T5-A1
T6-A2
Asleep 50 μV
1 sec
Figure 8–14. A typical BSSS in a 66-year-old patient.
component. It occurs mainly during drowsi- 6–11 Hz and an amplitude ranging from 60 to
ness and disappears during deeper levels of 200 μV and are seen mainly in adults. They
sleep (Fig. 8–15). It has no associated clinical occur during drowsiness and light sleep and
manifestations and has no correlation with clin- become apparent when the alpha and other
ical seizures or other symptoms.2, 13 Its typical awake patterns drop out. Wicket spikes are
characteristics are listed in Table 8–3. present over the temporal regions, occurring
bilaterally or independently over the two tem-
WICKET SPIKES poral regions, and they may occur more fre-
quently on one side, usually the left. When
Wicket spikes12, 13 consist of single spike-like wicket spikes occur as a single waveform, they
waveforms and appear as a monophasic frag- may be mistaken for a temporal spike dis-
ment of a mu-like rhythm (Fig. 8–16). Wicket charge; however, they are not accompanied
spikes, or wicket waves, have a frequency of by aftercoming slow waves or a distortion or
Table 8–2 Characteristics of Benign
Sporadic Sleep Spikes
Amplitude Low
Duration Short
Morphology Sharp, diphasic (steep descent)
Associated slow wave None or minimal
Background activity No disruption
Distribution Widespread
Laterality
Single Maximal unilateral
Multiple Bilateral
Occurrence Mainly adults
Event Sporadic
State Drowsiness or light sleep
Patient age Adult
Figure 8–15. A 6-Hz spike-and-wave burst during drowsiness in a 14-year-old girl with headaches but no seizures.
Table 8–3 Characteristics of 6-Hz

Spike-and-Wave
Frequency 6±1 Hz
Repetition rate Regular
Burst duration Brief, 1–2 seconds
Spike duration Brief
Amplitude Low
Distribution Diffuse, maximal anterior or posterior
Laterality Bisynchronous, symmetrical
Clinical state Drowsiness
Patient age Young adult
Age: 44 yrs
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
50 μV
1 sec
Figure 8–16. Wicket spikes in the left temporal region of a 44-year-old man.
135
slowing of the background that occurs with a which render clinical EEG an art as much as
true epileptogenic temporal spike. a science.”5
Key Points REFERENCES

• Wicket spikes have a spike or mu-like con- 1. American Electroencephalography Society. 2006.
figuraton and can occur as a single wave- Guideline 6: A proposal for standard montages to be
form or in a train with a frequency of used in clinical EEG. Journal of Clinical Neurophysi-
ology 23:111–17.
6–11 Hz. 2. Fisch, B. J. 1999. Spehlmann’s EEG primer. Princi-
• Wicket spikes occur over the temporal ples of digital and analog EEG. 3rd ed. Amsterdam:
regions during drowsiness or sleep. Elsevier Science Publishers.
• Wicket spikes should not be mistaken for 3. Ebersole, J. S., and T. A. Pedley, eds. 2003. Current
practice of clinical electroencephalography. 3rd ed.
a true epileptogenic spike discharge. Philadelphia: Lippincott Williams & Wilkins.
4. Kasteleijn-Nolst Trenite, D. G., R. Guerrini, C.
D. Binnie, and P. Genton. 2001. Visual sensitivity
MITTEN PATTERNS and epilepsy: A proposed terminology and classifica-
Mitten patterns are seen during sleep and con- tion for clinical and EEG phenomenology. Epilepsia
42:692–701.
sist of fast-wave and slow-wave components 5. Klass, D. W., and B. F. Westmoreland. 2001. Elec-
that resemble a mitten, with the thumb of the troencephalography: General principles and adult
mitten formed by the last wave of a spindle EEGs. In Clinical neurophysiology, ed. J. R. Daube,
and the hand portion by the slower wave com- 2nd ed., 77–107. London: Oxford University Press.
6. Klass, D. W. 1995. The continuing challenge of arti-
ponent. Mittens are a variant of a V wave or facts in the EEG. The American Journal of EEG
K-complex and should not be mistaken for a Technology 35:239–69.
spike-and-wave discharge. 7. Kellaway, P. 2003. Orderly approach to visual
analysis: Characteristics of the normal EEG of
adults and children. In Current practice of clinical
Key Point electroencephalography, ed. J. S. Ebersole, and T.
A. Pedley, 2nd ed., 100–59. Philadelphia: Lippincott
• Mitten V-wave variants should not be mis- Williams & Wilkins.
taken for spike-and-wave discharges 8. Tatum IV, W. O., A. M. Husain, S. R. Benbadis, and
P. W. Kaplan. 2006. Normal adult EEG and pat-
terns of uncertain significance. Journal of Clinical
BENIGN VARIANTS DURING Neurophysiology 23:194–207.
DROWSINESS AND SLEEP 9. Westmoreland, B. F., and D. W. Klass. 1998. Defective
alpha reactivity with mental concentration. Journal of
Summary Key Points Clinical Neurophysiology 15:424–28.
10. Klass, D. W., and R. P. Brenner. 1995. Electroen-
• Rhythmic temporal theta bursts of drowsi- cephalography of the elderly. Journal of Clinical Neu-
ness rophysiology 12:116–31.
11. Van Cott, A. C. 2002. Epilepsy and EEG in the elderly.
• 14 and 6 Hz positive bursts Epilepsia 43(Suppl 3):94–102.
• Benign sporadic sleep spikes 12. Reiher, J., and M. Lebel. 1977. Wicket spikes: Clin-
• 6-Hz spike and wave ical correlates of a previously undescribed EEG pat-
• Wicket spikes tern. The Canadian Journal of Neurological Sciences
• Mitten V-wave variants 4:39–47.
13. Westmoreland, B. F. 2003. Benign electroencephalo-
graphic variants and patterns of uncertain clinical
significance. In Currrent practice of clinical electroen-
cephalography, ed. J. S. Ebersole, and T. A. Pedley,
3rd ed., 235–45. Philadelphia: Lippincott Williams &
SUMMARY Wilkins.
14. Westmoreland, B. F., and D. W. Klass. 1981. A
In conclusion, this chapter provides an overview distinctive rhythmic EEG discharge of adults. Elec-
of the different types of normal EEG activity troencephalography and Clinical Neurophysiology
51:186–91.
and benign variants that are seen in the EEG. 15. Cobb, W. A., R. J. Guiloff, and J. Cast. 1979.
One needs to be aware of the normal variability Breach rhythm. The EEG related to skull defects.
at different ages and different states of wake- Electroencephalography and Clinical Neurophysiol-
fulness, drowsiness, and sleep. Dr. Klass has ogy 47:251–71.
16. Radhakrishnan, K., D. Chandy, G. Menon, and S.
stated that the “detection and interpretation Sarma. 1999. Clinical and electroencephalographic
of the EEG data derived from visual analysis correlates of breach activity. American Journal of Elec-
involve matters of judgment and experience, troneurodiagnostic Technology 39:138–47.
Chapter 9
Epileptiform Activity
INTRODUCTION AND Atypical Spike and Wave

OVERVIEW OF EPILEPTIFORM Paroxysmal Rhythmic Fast Activity
ACTIVITY Hypsarrhythmia
SPECIFIC FOCAL INTERICTAL Photoparoxsysmal Response
DISCHARGES Temporal Intermittent Rhythmic
Anterior Temporal and Frontal Spikes Delta Activity
Occipital Spikes
ICTAL DISCHARGES
Central-Temporal Spikes
EPILEPTIFORM ACTIVITY WITH A
Periodic Lateralized Epileptiform
POTENTIAL SEIZURE ASSOCIATION
Discharges
Stimulus-Induced Rhythmic,
GENERALIZED EPILEPTIFORM Periodic, or Ictal Discharges (SIRPIDS)
PATTERNS Burst Suppression
3-Hz Spike and Wave
SUMMARY
Slow Spike and Wave Discharges
INTRODUCTION AND OVERVIEW waveform, which has a pointed peak or sharp

OF EPILEPTIFORM ACTIVITY morphology but is less sharp than a spike
discharge and has a duration of 70–200 ms
The concept of defining specific epileptiform (Fig. 9–1). A spike and slow wave complex con-
abnormalities on routine electroencephalo- sists of a spike followed by an after-coming
gram (EEG) began decades ago. Epileptiform slow wave.
discharges are paroxysmal waveforms with dis- The routine EEG, performed with the usual
tinctive morphology that stand out from the activation procedures of sleep, hyperventila-
ongoing background activity, and typically are tion, sleep deprivation, and photic stimulation,
recorded from patients with epileptic seizures. is used to assess people with spells or potential
Epileptiform activity refers to spike discharges seizures, where epilepsy is considered in the
and sharp waves with after-coming slow waves. differential diagnosis. In patients with epilepsy,
A spike discharge is a waveform that has a interictal epileptiform discharges may be focal,
duration of less than 70 ms and pointed or more diffuse, or generalized. While not all
sharp morphology with steep ascending and abnormal epileptiform discharges have specific
descending limbs. A sharp wave is a broader clinical correlations, some provide important
137
T3–Fp1 Fp1–A1
Fp1–F7
T4–Fp2 Fp2–A2
F7–T3
T3–T5
Fp1–F3 F3–A1
T5–O1
Fp2–F4 F4–A2
Sharp wave Spike Spike and wave

(>70 ms) (<70 ms)
100 μV
1 sec
Figure 9–1. Sharp waves, spike with after-coming slow waves, and generalized spikes and waves.
clues about specific epilepsy or seizure syn- • A sharp wave has a pointed peak or sharp
dromes. However, not all epileptiform activity morphology and has a duration of 70–
is necessarily associated with the presence 200 ms.
of seizures; and routine EEGs containing • Epileptiform activity can occur without
unequivocal sharp waves or spikes with after- a seizure disorder, and not all patients
coming slow waves are present in a small with seizures have epileptiform activity
number of people with no history of seizures between seizures.
or epilepsy.1 Many waveforms have these • Epileptiform discharges are important
specific epileptiform features but do not pre- clues to specific epilepsy or seizure syn-
dict seizures or are even considered abnor- dromes.
mal.1, 2 The converse situation, where the rou-
tine EEG is normal in people with suspected
or definite epilepsy despite performing mul- SPECIFIC FOCAL INTERICTAL
tiple recordings, is more common.3 There- DISCHARGES
fore, a normal routine EEG does not rule
out the possible presence of epilepsy. Since
EEG waveforms having epileptiform charac- Anterior Temporal and Frontal
teristics may be normal or abnormal, the Spikes
electroencephalographer should be aware of
both normal variants with epileptiform fea- Temporal and frontal spike and sharp wave
tures and abnormal epileptiform waveforms. discharges represent an important finding if
Electroencephalographers should be proficient found in a patient being evaluated for seizures
in distinguishing between normal and abnor- and are highly correlated with seizures. Over
mal recordings and avoid “over interpretation” 90% of patients with anterior temporal spikes
of normal variants or other epileptiform-like and over 70% of patients with frontal spikes
activity as abnormal.4 have seizures (Fig. 9–2). During the routine
EEG, temporal spikes are activated by slow
Key Points wave sleep, with a 40%–60% increase in spike
occurrence during sleep.5
• Epileptiform discharges are paroxysmal
waveforms with distinctive morphology Key Points
that stand out from the ongoing back-
ground activity. • Temporal and frontal spike and sharp
• A spike discharge is a waveform with a wave discharges are highly correlated with
duration of less than 70 ms and pointed seizures.
or sharp morphology with steep ascending • Temporal spikes are activated during slow
and descending limbs. wave sleep.
Epileptiform Activity 139
M Age: 35 4–5–89
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
50 μV
1 sec
Figure 9–2. Left temporal spike.
Occipital Spikes compared to other epileptiform discharges,

as central-temporal spikes have a horizon-
Occipital spikes are most commonly seen in tal dipole in which both ends of the dipole
children. The presence of occipital spikes is are recorded with standard surface leads. The
associated with seizures in approximately 30%– surface positive end of the dipole is ori-
50% of children between 3 and 6 years of ented toward the frontal lobe and the neg-
age. The presence of occipital spikes may be ative end of the dipole is mostly oriented
associated with a history of visual deprivation, toward the central (rolandic) or temporal areas
and many children with this finding expe- (Fig. 9–3). They are usually of high ampli-
rience visual abnormalities. If this entity is tude and diphasic, with a high amplitude after-
seen in older children it may be associated coming slow wave. The spikes may be unilat-
with benign occipital epilepsy where the spikes eral or bilateral often shifting from one side
attenuate with eye opening.6, 7 As with other to another.9 Their occurrence is often fre-
spikes, occipital spikes may be associated with quent, appearing in trains or short clusters
underlying structural lesions. that shift from one hemisphere to the other.
The background activity of the EEG is oth-
Key Points erwise normal. Like anterior-temporal spikes,
• In 30%–50% of children between 3 and 6 central-temporal spikes may be activated by
sleep.
years of age, occipital spikes are associated
The presence of these central-temporal
with seizures.
• In older children, occipital spikes are asso- spikes is associated with the clinical condi-
tion of benign rolandic epilepsy of childhood
ciated with benign occipital epilepsy and
(BREC). This disorder is primarily seen in chil-
attenuate with eye opening.
dren aged 4–12 years and is characterized by
frequent nocturnal seizures, although daytime
Central-Temporal Spikes seizures may also occur. The typical seizure is
of focal onset with symptoms referable to the
Central-temporal spikes represent important Sylvian area, and includes facial or hand twitch-
epileptiform discharges, typically occurring in ing, motor speech arrest, or sensory symp-
children aged 4–12 years, which are associ- toms involving the upper extremity or face.
ated with seizures in up to 80% of them.8 These discharges are usually not associated
The morphology of these discharges is unique with structural lesions. The syndrome, and
Age: 7 yrs
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Nocturnal seizure 50 μV
1 sec
Figure 9–3. Central-temporal (rolandic) spikes.
EEG abnormalities, often resolves during the PLEDs often occur in the absence of seizures,
second decade of life. occasionally the presence of PLED may be
seen with ongoing partial seizures. Correla-
Key Points tive motor activity or an evolving ictal pattern
• Central-temporal spikes typically occur in on the EEG along with the PLED is sugges-
tive of seizure activity. Distinguishing the pres-
children aged 4–12 years, which are asso-
ence of seizures is important for therapeutic
ciated with seizures in up to 80% of them.
• Central-temporal spikes have a horizontal intervention.11
PLEDs are usually indicative of a unilat-
dipole in which both ends of the dipole are
eral hemispheric lesion. The most common
recorded with standard surface leads.
• The surface positive end of the dipole and classical etiology is acute ischemic stroke.
However, infectious etiologies such as herpes
is oriented toward the frontal lobe and
simplex encephalitis can also be an etiology.
the negative end of the dipole is mostly
The term PLED implies laterality, but on occa-
oriented toward the central (rolandic) or
sion independent bilateral PLEDs (BiPLEDs)
temporal areas.
• The presence of central-temporal spikes may be seen in similar patients. The presence
of BiPLEDs is associated with infections and
is associated with the clinical condition of
epileptic seizures.12
benign rolandic epilepsy of childhood.
Periodic Lateralized Epileptiform

Key Points
Discharges
• PLEDs are recurring sharp waves that
Periodic lateralized epileptiform discharges recur with a regular periodicity.
(PLEDs) are recurring sharp waves that recur • The rate of occurrence of the discharges is
with a regular periodicity (Fig. 9–4). The rate typically about one every second.
of occurrence of the discharges is typically • PLEDs are usually indicative of a uni-
about one every second, but they may be less lateral hemispheric lesion. The most
frequent.10–12 PLEDs are usually continuous common and classical etiology is acute
and may be acute or subacute in nature. While ischemic stroke.
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp1-F3
F3-C3
C3-P3
P3-O1
Fpz-Fz
Fz-Cz
Cz-Pz
Fp2-F4
F4-C4
C4-P4
P4-O2
Fp2-F8
F8-T4
T4-T6
T6-O2
EKG–EKG+
Figure 9–4. Left periodic lateralized epileptiform discharges (15 μV/mm, low-frequency filter 1.0 Hz, high-frequency
filter 70 Hz).
GENERALIZED EPILEPTIFORM activity is usually normal, no matter the age of

PATTERNS the patient.
The 3-Hz spike and wave discharges are
In contrast to focal epileptiform discharges, associated with absence seizures. The duration
some epileptiform activity occurs in a gener- of these discharges on the routine EEG varies
alized distribution. These generalized epilepti- from fractions of a second to many seconds.
form patterns are highly associated with spe- During brief discharges, clinical accompani-
cific epilepsy syndromes. ments may not be witnessed. However, when
the activity lasts for more than a few seconds,
consciousness may be impaired. With longer
discharges, other minor clinical signs may be
3-Hz Spike and Wave seen and noted by the technologist performing
the EEG, including eye blinks or minor motor
The 3-Hz spike and wave discharge is seen activity of the face or limbs that may correlate
most commonly in childhood, between the with the 3-Hz spike and wave activity. Ambula-
ages of 4 and 14, but may persist later in tory EEG may help quantify the frequency of
life. This pattern is characterized by gener- the discharges.
alized, paroxysmal, high amplitude spike and
slow wave complexes recurring at a frequency Key Points
of 3 Hz (Fig. 9–5). The maximum amplitude
of the pattern is seen in the frontal regions. • The 3-Hz spike and wave discharge is seen
The 3-Hz spike and wave pattern is classi- most commonly in childhood, between
cally enhanced by hyperventilation and is more the ages of 4 and 14 years
prominent during hypoglycemia. The mor- • The 3-Hz spike and wave discharge is
phology of the discharges changes with sleep, characterized by generalized, paroxysmal,
when they become less well formed and more high amplitude spike and slow wave com-
fragmented with shorter bursts during early plexes recurring at a frequency of 3 Hz.
sleep, with increased occurrence with deeper • The maximum amplitude of the pattern is
stages of NREM sleep.13 Interictal background seen in the frontal regions.
Age: 10 yrs
FP1-A1
FP2-A2
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
P4-A2
Stops hyperventilating
is unresponsive
has blinking of the eyelids 200 μV
1 sec
Figure 9–5. Generalized 3-Hz spike and wave pattern with frontal predominance.
• The 3-Hz spike and wave pattern is occurs in patients with a variety of underly-
enhanced by hyperventilation and hypo- ing pathologies. In young children it is asso-
glycemia. ciated with the Lennox–Gastaut syndrome—a
• The morphology of the discharges changes syndrome of intractable seizures, developmen-
with sleep when they become less well tal delay with motor dysfunction, and poor
formed and more fragmented with shorter response to anti-seizure medications.15, 17
bursts during early sleep.
• The 3-Hz spike and wave discharges are Key Points
associated with absence seizures.
• The slow spike and wave pattern has
characteristic recurring sharp and slow
Slow Spike and Wave Discharges wave discharges firing repetitively at a fre-
quency of 1.5–2.5 Hz.
• The background activity of slow spike and
The slow spike and wave pattern has important
clinical correlations. With this pattern, recur- wave discharges is usually abnormal with
ring sharp and slow wave discharges fire repet- diffuse slowing.
• Multiple generalized seizure types are
itively at a frequency of 1.5–2.5 Hz, which is
distinctly slower than the more classic 3-Hz associated with this discharge, including
spike and slow wave activity. The spike has tonic–clonic, akinetic, and myoclonic.
• In young children it is associated with the
a slightly longer duration than typical 3-Hz
spike and wave complexes and has more of a Lennox–Gastaut syndrome.
sharp wave morphology14 (Fig. 9–6). In con-
trast to the normal inter-burst background
activity with 3-Hz spike and wave discharges, Atypical Spike and Wave
the background activity of slow spike and wave
discharges is usually abnormal with diffuse The so-called atypical spike and wave dis-
slowing. charges are another example of a generalized
Slow spike and wave pattern is seen mostly epileptiform discharge. The spike and wave
in children between the ages of 2 and 6, but discharges occur in a more irregular nonrhyth-
may persist into adolescence.14–16 Multiple gen- mic interval as compared to the classic 3-Hz
eralized seizure types are associated with this spike and wave activity. Atypical spike and wave
discharge, including tonic–clonic, akinetic, and bursts may consist of both spike or polyspike
myoclonic. This epileptiform pattern typically and slow waves with frequencies in the 2–5 Hz
F Age: 10
FP1-A1
FP2-A2
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
P4-A2
O1-A1
O2-A2
Frequent generalized seizures 150μV
1 sec
Figure 9–6. Slow spike and wave pattern.
range. These discharges may occur in relatively Key Points

short bursts (1–3 seconds) when they occur • Paroxysmal rhythmic fast activity is a
between seizures or may be a part of an ictal
pattern associated with various types of gen- 12–20 Hz repetitive spike discharge that
eralized seizures, especially tonic–clonic and occurs in the parasagittal region, usually
during sleep that is associated with tonic
myoclonic seizures. Like most forms of gener-
alized spike and wave discharges, atypical spike or tonic–clonic seizures.
and wave activity may be activated by sleep or
photic stimulation. The pattern may be seen in
both children and adults. Hypsarrhythmia
Hypsarrhythmia is an epileptiform pattern with
Key Points features of high amplitude, virtually continu-
• Atypical spike and wave discharges have ous mix of arrhythmic slowing, and multifocal
an irregular interval and vary from 2 to sharp waves and spikes (Fig. 9–7). This activ-
5 Hz. ity is usually seen in epilepsy patients aged
• The atypical spike and wave pattern is between 4 months and 4 years, who often have
associated with various forms of general- infantile spasms.19 If the EEG is performed
ized seizures. during the infantile spasms, the clinical seizure
is associated with an electrodecremental pat-
tern in which there is a sudden suppression of
virtually all activity on the EEG.20 The combi-
Paroxysmal Rhythmic Fast Activity nation of infantile spasm, developmental delay,
and hypsarrhythmia is West syndrome. No one
This unique activity consists of repetitive spike specific disease causes the syndrome and the
discharges ranging from 12 to 20 Hz and occurs pattern is felt to be the result of significant
in the parasagittal regions. It may appear syn- cerebral insult that usually occurred before the
chronous but may be asymmetrical as well age of one year. Tuberous sclerosis is one of
and is often seen during sleep. Toward the the more common causes of the disorder and
end of the discharge, which may be brief is often associated with very difficult to control
in duration, one or more slow waves may seizures. Other metabolic or infectious etiolo-
be seen. The occurrence of these waveforms gies account for one third of the cases, and in
has been associated with tonic or tonic–clonic approximately one third of the cases a cause is
seizures.18 never found.
Age: 9 months (6–10–82)

FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
Infantile spasms 200μV

1 sec
Figure 9–7. Hypsarrhythmia with high amplitude discharges.
Key Points precipitated. These discharges should not be

confused with the photomyogenic response in
• Hypsarrhythmia is an epileptiform pattern
which repetitive contractions of facial or other
with features of high amplitude, virtually muscles are time-locked to the frequency of
continuous mix of arrhythmic slowing, and the strobe light.
multifocal sharp waves and spikes. Photoparoxysmal discharges are often asso-
• Hypsarrhythmia is seen in patients aged
ciated with clinical seizures, most commonly
between 4 months and 4 years, who often the generalized epilepsies, including juve-
have infantile spasms. nile myoclonic epilepsy, childhood absence
• If the EEG is performed during the infan-
epilepsy, and the West and Lennox–Gastaut
tile spasms, the clinical seizure is associ- syndromes. There is a slight female predomi-
ated with an electrodecremental pattern nance for the photoparoxysmal abnormality.22
on EEG.
• The combination of infantile spasm, Although poorly understood, there appears
to be a hereditary photosensitivity that may
developmental delay, and hypsarrhythmia
appear in asymptomatic siblings of affected
is West syndrome.
individuals, especially females.22–24
Key Points
Photoparoxsysmal Response
• Photoparoxysmal activity is abnormal epilep-
Abnormal epileptiform activity in response to tiform activity that occurs in response to
strobe light stimulation is a well-known entity. strobe light stimulation.
The light is typically flashed at frequencies of • The activity is usually seen as a bilateral
1–30 Hz, with abnormal activity mostly seen generalized synchronous spike (frontal
at stimulus rates between 15 and 25 Hz. The predominant) or polyspike (posterior pre-
activity is usually seen as a bilateral general- dominant) and slow wave discharges or
ized synchronous spike (frontal predominant) rarely as a focal occipital discharge.
or polyspike (posterior predominant) and slow • Photoparoxysmal activity is mostly seen at
wave discharges or rarely as a focal occipital stimulus rates between 15 and 25 Hz.
discharge (Fig. 9–8).21 The discharges may be • The discharges may be self-limited and
self-limited and end with discontinuation of end with discontinuation of the stimulus
the stimulus or may persist after the strobe is or may persist after the strobe is turned
turned off (photoconvulsive response). If the off (photoconvulsive response).
strobe light is not turned off when the epilep- • Photoparoxysmal discharges are often
tiform activity is present, a seizure may be associated with clinical seizures.
F Age:14
FP1-A1
FP2-A2
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
P4-A2
O1-A1
O2-A2
Photic
12 Hz
500μV
1 sec
Figure 9–8. Photoparoxysmal response with discharges that persist beyond stimulus termination.
Temporal Intermittent Rhythmic Key Points

Delta Activity • TIRDA is a pattern that occurs during
drowsiness or light sleep and likely reflects
Temporal intermittent rhythmic delta activity a deep focus in the temporal lobe.
(TIRDA) is an uncommon finding that is seen
in approximately 0.3% of all EEGs. TIRDA
usually occurs during drowsiness and light
sleep. The presence of this activity is often seen ICTAL DISCHARGES
in patients being evaluated for temporal lobe
resection and has a high predictive value for It is uncommon to record ictal activity dur-
temporal lobe epilepsy.25, 26 The lack of spike ing routine EEGs utilizing surface electrodes
or sharp activity is likely a reflection of a deep in the EEG lab or at the bedside. Seizures
focus in the temporal lobe that is seen slow- recorded on an EEG are more typically cap-
ing in a routine EEG using scalp electrodes tured with extended recordings, such as ambu-
(Fig. 9–9). latory devices or in an epilepsy monitoring unit.
M Age: 75
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
50 μV
1 sec
Figure 9–9. Temporal intermittent rhythmic delta activity.

Ictal discharges have many frequencies, loca- of activity using scalp electrodes during a
tions, and morphologies. Ictal discharges may partial seizure is particularly common when
occur at any EEG waveform frequency (beta, seizure foci are located in the midline or deep
alpha, theta, or delta) and the morphology within the frontal lobe.
can range from sharp waves to spike and slow Status epilepticus is defined as continuous
wave or any combination of morphologies. Ictal or frequently recurring seizures that may be
abnormalities may also be focal or generalized. partial or generalized and convulsive or non-
The ictal activity may begin as a certain pat- convulsive. Figure 9–11 shows a continuous,
tern and remain constant or more likely evolve generalized spike discharge in a patient with
as the seizure progresses. The changing or nonconvulsive status. The discharges would be
evolving pattern may be manifest by changing the same with an ongoing clinical seizure.
amplitudes (increased or decreased), frequen-
cies, or both. Partial seizures begin focally and
Key Points
either stay localized or spread to become gen-
eralized. Post-ictal slowing and suppression is • Seizures recorded on an EEG are more
common and may be focal in a partial seizure typically captured with extended record-
or generalized in a primary or secondary gen- ings, such as ambulatory devices or in an
eralized seizure. Focal post-ictal slowing may epilepsy monitoring unit.
have localizing value. Figure 9–10 depicts a • Ictal discharges have many frequencies,
partial seizure that evolves and spreads to locations, and morphologies.
other regions. Although the EEG is very good • Ictal abnormalities may also be focal or
in recording epileptiform activity during a generalized.
clinical seizure, 5% or less of all complex • The ictal activity may begin as a cer-
partial seizures show no ictal activity on EEGs tain pattern and remain constant, or more
utilizing scalp recording electrodes.27 This lack likely evolve as the seizure progresses.
A
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
Fp8-T4
T4-T6
T6-O2
Figure 9–10. Evolving left central seizure. A. Initiation at T5–01 (left posterior temporal). B. Spread of rhythmic, ictal
activity to entire left temporal region with head turning.
B
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Figure 9–11. Continuous spike and wave activity in generalized status epilepticus.
147
EPILEPTIFORM ACTIVITY WITH Key Points

A POTENTIAL SEIZURE • SIRPIDS are discharges that are typically
ASSOCIATION seen during long-term trending in the
intensive care unit setting.
Stimulus-Induced Rhythmic, • This pattern is comprised of rhythmic,
Periodic, or Ictal Discharges periodic, or ictal appearing discharges that
are consistently induced by some type of
(SIRPIDS) alerting stimuli and is associated with sta-
tus epilepticus.
Stimulus-induced rhythmic, periodic, or ictal
discharges (SIRPIDS) are discharges that
are typically seen during long-term trend- Burst Suppression
ing in the intensive care unit setting. This
pattern is comprised of rhythmic, periodic, The burst suppression pattern is characterized
or ictal appearing discharges that are con- by intermittent bursts of paroxysmal sharp and
sistently induced by some type of alerting slow activity or an admixture of various fre-
stimuli (Fig. 9–12). The pattern has been asso- quencies separated by periods of absent or rel-
ciated with status epilepticus. The observa- atively quiescent EEG activity (Fig. 9–13). The
tion of these waveforms is largely a product bursts generally last from less than a second to
of the advances in equipment and record- several seconds with relative quiescent periods
ing capabilities that have permitted long-term lasting variable lengths of time, usually 4–10
recordings. The study of this type of epilepti- seconds in length. Etiologies that produce this
form activity and its potential clinical correla- pattern include ongoing end-stage seizures,
tion is ongoing.28 hypothermia, medication effect (particularly
Figure 9–12. SIRPIDS activated by a noxious sternal rub. (From Hirsch, L. J., J. Claassen, S. A. Mayer, and R. G. Emer-
son. 2004. Stimulus-induced rhythmic, periodic, or ictal discharges (SIRPIDs): A common EEG phenomenon in the
critically ill. Epilepsia 45:109–23. By permission of Blackwell Publishing.)
F Age: 66
T3–Fp1
T4–Fp2
Fp1–F3
Fp2–F4
C3–P3
C4–P4
P3–O1
P4–O2
Several hours cardiac arrest 50 μV
1 sec
Died 12 hours later
Figure 9–13. Burst suppression pattern 7 μV/mm; LFF, low-frequency filter 1.0 Hz; HFF, high-frequency filter 70 Hz,
10 mm/sec.
during anesthesia or drug overdose), and post In Clinical electroencephalography of children, ed.
anoxic/hypoxic insult.29–33 The usefulness of P. Kelaway, and I. Petersén, 167–88. New York: Grune
using ongoing monitoring of the burst sup- & Stratton.
2. Eeg-Olofsson, O., I. Petersén, and U. Selldén. 1971.
pression pattern during surgery for depth of The development of the electroencephalogram in nor-
anesthesia has recently been questioned.34 mal children from the age of one through fifteen years.
Paroxysmal activity. Neuropädiatrie 4:375–404.
Key Points 3. Drazkowski, J., and D. Blum. 1998. Electroen-
cephalography and evoked potential studies. In Adult
• Burst suppression pattern is character- neurology, ed. J. corey-Bloom, 2nd ed., 15–33. Mosby,
ized by intermittent bursts of paroxysmal St. Louis, MO: Blackwell.
4. Benbadis, S. R., and W. O. Tatum. 2003. Overinterpre-
sharp and slow activity or an admixture tation of EEGs and misdiagnosis of epilepsy. Journal of
of various frequencies separated by peri- Clinical Neurophysiology 20:42–4.
ods of absent or relatively quiescent EEG 5. Klass, D. W. 1975. Electroencephalographic mani-
activity. festations of complex partial seizures. Advances in
• Etiologies that produce burst suppres- Neurology 11:113–40.
6. Kivity, S., T. Ephraim, R. Weitz, and A. Tamir. 2000.
sion pattern include ongoing end-stage Childhood epilepsy with occipital paroxysms: Clinical
seizures, hypothermia, anesthesia, or drug variants in 134 patients. Epilepsia 41:1522–33.
overdose, and post anoxic/hypoxic insult. 7. Niedermeyer, E., and F. Lopes da Silva, eds. 1999.
Electroencephalography: Basic principles, clinical
applications, and related fields, 4th ed. Baltimore:
Williams & Wilkins.
SUMMARY 8. Lombroso, C. T. 1967. Sylvian seizures and midtem-
poral spike foci in children. Archives of Neurology
EEG continues to be an important test to 17:52–9.
functionally evaluate people with suspected 9. Kellaway, P. 2000. The electroencephalographic fea-
seizures. Although few specific EEG patterns tures of benign centrotemporal (rolandic) epilepsy of
childhood. Epilepsia 41:1503–6.
exist for specific diseases, the presence of 10. Chatrian, G. E., C. M. Shaw, and C. N. Luttrell. 1965.
epileptiform discharges and patterns on the Focal electroencephalographic seizure discharges in
EEG may help identify certain syndromes. acute cerebral infarction. Electrographic, clinical, and
Clinicians using EEG in the evaluation and pathological observations. Neurology 15:123–31.
11. Chatrian, G. E., C. M. Shaw, and F. Plum. 1964. Focal
management of people with suspected epilepsy periodic slow transients in epilepsia partialis continua:
should be familiar with the different epilep- Clinical and pathological correlations in two cases.
tiform discharges and their associated clinical Electroencephalography and Clinical Neurophysiol-
significance. ogy 16:387–93.
12. Chatrian, G. E., C. M. Shaw, and H. Leffman. 1964.
The significance of periodic lateralized epileptiform
discharges in EEG: An electrographic, clinical and
REFERENCES pathological study. Electroencephalography and Clini-
cal Neurophysiology 17:177–93.
1. Petersén, I., O. Eeg-Olofsson, and U. Selldén. 1968. 13. Dalby, M. A. 1969. Epilepsy and 3 per second spike
Paroxysmal activity in EEG of normal children. and wave rhythms. A clinical, electroencephalographic
and prognostic analysis of 346 patients. Acta Neurolog- J. K. Penry, and C. R. Sing, 113–21. New York: Raven
ica Scandinavica 40(Suppl):3. Press.
14. Blume, W. T., R. B. David, and M. R. Gomez. 1973. 24. Newmark, M. E., and J. K. Penry. 1979. Photosen-
Generalized sharp and slow wave complexes. Associ- sitivity and epilepsy: A review. New York: Raven
ated clinical features and long-term follow-up. Brain Press.
96:289–306. 25. Geyer, J. D., E. Bilir, R. E. Faught, R. Kuzniecky,
15. Markand, O. N. 1977. Slow spike-wave activity in and F. Gilliam. 1999. Significance of interictal tem-
EEG and associated clinical features: Often called poral lobe delta activity for localization of the primary
“Lennox” or “Lennox–Gastaut” syndrome. Neurology epileptogenic region. Neurology 52:202–5.
27:746–57. 26. Normand, M. M., Z. K. Wszolek, and D.W. Klass.
16. Daly, D. D. 1979. Use of the EEG for diagnosis 1995. Temporal intermittent rhythmic delta activity
and evaluation of epileptic seizures and nonepilep- in electroencephalograms. Journal of Clinical Neuro-
tic episodic disorders. In Current practices of clinical physiology 12:280–4.
electroencephalography, ed. D. W. Klass, and D. D. 27. Gastaut, H., and R. Broughton. 1972. Epileptic
Daly, 221–68. New York: Raven Press. seizures. Springfield, Illinois: Thomas.
17. Gastraut, H., J. Roger, R. Soulayrol, et al. 1966. Child- 28. Hirsch, L. J., J. Claassen, S. A. Mayer, and R.G.
hood epileptic encephalopathy with diffuse slow spike- Emerson. 2004. Stimulus-induced rhythmic, peri-
waves (otherwise known as “petit mal variant”) or odic, or ictal discharges (SIRPIDs): A common
Lennox syndrome. Epilepsia 7:139–79. EEG phenomenon in the critically ill. Epilepsia 45:
18. Westmoreland, B. F., and D. W. Klass. 1990. Unusual 109–23.
EEG patterns. Journal of Clinical Neurophysiology 29. Treiman, D. M. 1995. Electroclinical features of sta-
7:209–28. tus epilepticus. Journal of Clinical Neurophysiology
19. Jeavons, P. M., and B. D. Bower. 1974. Infantile 12:343–62.
spasms. In Handbook of clinical neurology: The epilep- 30. Bird, T. D., and F. Plum. 1968. Recovery from bar-
sies, ed. P. J. Vinken, and G. W. Bruyn, Vol. 15, 219–34. biturate overdose coma with a prolonged isoelectric
Amsterdam: North-Holland Publishing Company. electroencephalogram. Neurology 18:456–60.
20. Bickford, R. G., and D. Klass. 1969. Scalp and depth 31. Haider, I., H. Matthew, and I. Oswald. 1971. Elec-
electrographic studies of electro-decremental seizures troencephalographic changes in acute drug poisoning.
(abstract). Electroencephalography and Clinical Neu- Electroencephalography and Clinical Neurophysiol-
rophysiology 12:638. ogy 30:23–31.
21. Sunku, A. J., M. R. Gomez, and D. W. Klass. 32. Silverman, D. 1975. The electroencephalogram in
1998. Epileptic seizures, EEG abnormalities, and neu- anoxic coma. In Altered states of consciousness, coma
ronal heterotopia in the Dubowitz syndrome. Amer- and cerebral death. Handbook of electroencephalog-
ican Journal of Electroneurodiagnostic Technology raphy, ed. A. Rémond, Vol. 12, 81–94. Amsterdam:
38:156–63. Elsevier.
22. Wolf, P., and R. Goosses. 1986. Relation of photosen- 33. Martin, J. T., A. Faulconer, Jr., and R. G. Bick-
sitivity to epileptic syndromes. Journal of Neurology, ford. 1959. Electroencephalography in anesthesiology.
Neurosurgery, and Psychiatry 49:1386–91. Anesthesiology 20:359–76.
23. Doose, H. 1982. Photosensitivity: Genetics and signifi- 34. Fabregas, N., and C. Gomar. 2001. Monitoring in neu-
cance in the pathogenesis of epilepsy. In Genetic basis roanaesthesia: Update of clinical usefulness. European
of the epilepsies, ed. V. E. Anderson, W. A. Hauser, Journal of Anaesthesiology 18(7):423–39.
Chapter 10
Adult EEG: Abnormal

Nonepileptiform Activity
INTRODUCTION Subdural Hematoma

TYPES OF EEG ABNORMALITIES Inflammatory Disorders
Slow-Wave Abnormalities Herpes Simplex Encephalitis
Asymmetry Other Focal Cerebral Lesions
Asymmetric or Altered Reactivity Transient Disorders
Suppression
Epileptiform Activity ELECTROENCEPHALOGRAPHIC
Periodic Patterns MANIFESTATIONS OF DIFFUSE
DISORDERS
FOCAL INTRACRANIAL Slow-Wave Abnormalities
PROCESSES CAUSING EEG Specific Patterns
ABNORMALITIES Generalized Periodic Patterns
The EEG in Focal Intracranial Lesions Coma Patterns
Cerebral Infarction
Intracerebral Hemorrhage EVALUATION FOR SUSPECTED
Arteriovenous Malformations BRAIN DEATH
Tumors CONCLUSION
Head Trauma
INTRODUCTION lesion or disease process, and the degree

of the disturbance of function. In a patient
Electroencephalography (EEG) is a clinical with altered consciousness, the EEG can con-
neurophysiology test that assesses the func- firm an organic disturbance of function, help
tion of the brain. Neuroimaging studies look make the diagnosis, and help determine the
at the structure of the brain, but the EEG prognosis.
can give evidence of normal or altered func- The most common referrals for EEGs are
tion of the brain and evolution of a dis- seizure disorders (discussed in Chapter 9),
ease process that may not be apparent on focal lesions, encephalopathies, and alteration
neuroimaging studies. The EEG is useful of consciousness.
for evaluation of cerebral function and can This chapter will review the types of EEG
help determine whether there is a focal vs. abnormalities in focal lesions, diffuse disorders,
diffuse process, the site and extent of the and brain death.
151
TYPES OF EEG ABNORMALITIES involving the subcortical white matter usually

cause the delta slowing seen on the EEG.1–4
The EEG is one measure or parameter of Another type of slowing consists of inter-
brain function or dysfunction, and anything mittent rhythmic and bilaterally synchronous
causing a disturbance of cerebral function can slow-wave activity which shows reactivity or
cause changes in the EEG. There are certain is attenuated to eye opening or alerting stim-
common ways in which lesions or disease pro- uli. When present or maximal over the frontal
cesses can affect the EEG, and the main types head regions, this is referred to as FIRDA
of abnormalities that are seen with cerebral (Frontal Intermittent Rhythmic Delta Activ-
lesions are slowing, asymmetry, altered reac- ity). When present over the posterior head
tivity, suppression, epileptiform activity, and regions, this is referred to as OIRDA (Occipital
periodic patterns. Intermittent Rhythmic Delta Activity). This
type of slowing is often seen with diffuse
encephalopathies, particularly in metabolic or
Slow-Wave Abnormalities toxic encephalopathies.5, 6 FIRDA or OIRDA
can also be seen as a disturbance of func-
The slower the frequency, the more severe tion arising at a distance from more deep-
the abnormality. A slowing of the alpha fre- seated structures such as the diencephalon and
quency reflects a mild abnormality, theta slow- periventricular structures. Since this is pro-
ing reflects a moderate abnormality, and delta jected to the surface, this has been referred to
slowing a more severe abnormality. as a projected rhythm. The projected rhythm
Polymorphic and arrhythmic delta slowing, can occur as a result of a primary lesion
which persists throughout the tracing and affecting midline structures, a shift or distor-
shows little or no reactivity to afferent stimu- tion of midline structures from a mass lesion,
lation, usually indicates a disturbance of cere- increased intraventricular pressure, or obstruc-
bral function (Fig. 10–1). Lesions primarily tion of ventricular flow.1–3, 5
Figure 10–1. Electroencephalogram from a 20-year-old man with a left frontal astrocytoma showing persistent polymor-
phic delta activity predominantly in the left frontal region (Fp1).
Abnormal Nonepileptiform Activity 153
Key Points abnormal side following cessation of hyperven-

tilation.3
• Slower frequencies are seen with more
severe abnormalities. Key Points
• Polymorphic delta slowing reflects a dis-
turbance of cerebral function. • Asymmetric reactivity can occur with fail-
• Intermittent rhythmic delta slowing is ure of alpha activity to attenuate on one
seen with encephalopathies, more deep- side.
seated lesions, and increased intraventric- • An asymmetric driving response is seen
ular pressure. with occipital lesions.
• Hyperventilation may accentuate or elicit
focal slowing or an asymmetry on one side.
Asymmetry
Asymmetry refers to an amplitude difference Suppression
in similar activity which may be increased or
decreased on the side of the lesion as com- Suppression refers to both a decrease in ampli-
pared to homologous regions on the opposite tude and amount, or an absence of activity.
side. An additional abnormality such as slow- While often generalized, it may occur in a
ing, in addition to the asymmetry, is helpful in focal or lateralized distribution. Suppression is
lateralizing the side of the abnormality. A skull graded mild, moderate or severe with the latter
defect may be associated with a focal increase showing essentially no activity.
in the amplitude of activity in the region of the Key Point
skull defect. This has been termed a breach
rhythm.7, 8 An amplitude difference between • Suppression refers to a decrease in ampli-
sides due to suppression is not referred to as tude and amount of activity.
an asymmetry.
Key Points Epileptiform Activity

• Asymmetry is an increase or decrease in
the amplitude of activity on one side. Epileptiform activity can be present with
• A breach rhythm is a focal increase in the spikes, sharp waves, periodic lateralized epilep-
amplitude of activity over a skull defect. tiform discharges (PLEDs), or seizure dis-
charges, which can occur focally in the vicinity
of the lesion or in a more generalized distri-
Asymmetric or Altered Reactivity bution.3–5 The most epileptogenic areas of the
brain are the temporal and frontal lobes, and
There may be a focal or lateralized alteration lesions involving these regions are more likely
in the reactivity of EEG activity on one side. to be accompanied by epileptiform abnormali-
This can be seen with the alpha rhythm, photic ties and seizures.
driving, and hyperventilation. If there is a fail- Key Points
ure of alpha activity or rhythm to attenuate
on one side, the side that fails to attenuate is • Epileptiform activity may be focal or gen-
the abnormal side. This may occur with eye eralized.
opening or mental alerting procedures.9 An • The most epileptogenic areas of the brain
altered or asymmetric photic driving response are the temporal and frontal regions.
is most commonly seen with lesions affect-
ing the occipital lobe and in which the photic
driving is usually decreased or absent on the Periodic Patterns
side of the lesion. On rare occasions, it may
be increased on the side of the lesion. Hyper- Generalized periodic patterns or waveforms
ventilation may elicit or accentuate focal or consisting of triphasic waves, sharp waves,
lateralized slowing or an asymmetry of back- slow-wave complexes, and epileptiform dis-
ground activity. The changes may persist on the charges may be seen with various disorders
of cerebral function.3–5, 10 More focal periodic Key Points

waveforms can occur as with PLEDs.3, 10, 11
• The most common focal cerebral lesions
are vascular lesions and tumors.
Key Points • The degree of the EEG abnormality
depends on various factors including loca-
• Generalized periodic patterns may be
tion, type, and evolution of lesion, the
seen with disorders of cerebral dysfunc- presence of clinical seizures, the presence
tion of skull defect, the state and age of the
• Periodic patterns include triphasic waves,
patient, and other concomitant systemic
sharp waves, slow-wave complexes, and processes.
epileptiform discharges
Cerebral Infarction
FOCAL INTRACRANIAL
PROCESSES CAUSING EEG During the acute phase, the EEG often shows
ABNORMALITIES focal delta slowing over the involved area and
a decrease in background activity.3–5 During
the first few days, the EEG abnormalities
The EEG in Focal Intracranial may become worse secondary to the effects
Lesions of local edema and vasospasm. At times more
widespread slowing can also be seen in addi-
The EEG is one measure or parameter of brain tion to the more focal abnormalities. After a
function or dysfunction, and lesions causing few days, the EEG usually improves with a
a disturbance of cerebral function can cause return of faster frequencies and more normal
changes in the EEG. The degree and severity background components. In some patients, the
of the EEG abnormalities depend on a num- EEG will become normal after an infarct even
ber of factors, which include location, size, though there may be a persistence of a neu-
and type of lesion. Also, the stage of evolu- rologic deficit. In other cases, there may be
tion and rapidity of evolution of the lesion some persistence of slow-wave abnormalities
determine the severity of the abnormal find- or asymmetry of activity. If the infarct involves
ings in the EEG. Other factors that play a role more deep-seated structures such as the inter-
in the degree of EEG findings include med- nal capsule, the EEG shows minimal or little
ications, seizures, postictal state, skull defect, change even though the patient may have a
and other systemic processes. The state of the severe neurologic deficit.
patient, whether awake or asleep, level of con- On occasion, an acute epileptic focus devel-
sciousness, and the age (child or adult) are ops within a few hours or days after an
also factors determining the degree of EEG acute cerebral infarct. This usually consists of
abnormalities.12 PLEDs on the EEG, which are frequently
associated with seizures, a neurologic deficit,
and obtundation of the patient. The clinical
Purpose and Role of EEG in Focal manifestations and PLEDs usually improve or
Intracranial Processes resolve after a few days to a couple of weeks
• Indicating the presence of a focal lesion. after the acute onset.3, 11, 13
• Indicating the presence of focal epilepti-
form activity. Key Points
• Following the patient’s clinical course. • During an acute infarction, the EEG
• Determining whether the lesion is resolv- shows focal slowing and a decrease in the
ing, static, or progressing. background activity.
• The EEG usually improves after a few
The most common cerebral lesions are vascular days.
lesions and tumors. Other lesions include head • There may be little change in the EEG
trauma, infectious lesions, and other types of with more deep-seated infarcts.
focal pathology involving the cerebral hemi- • On occasions, PLEDs and seizures may be
spheres.3–5 seen following an acute vascular insult.
Intracerebral Hemorrhage polymorphic delta activity (Fig. 10–1). Usually

the slowest frequency components and the
Marked focal slowing and attenuation of back- most persistent slowing occur over the site of
ground activity occur with an intracerebral the tumor.3, 4, 12 Slow-wave activity may also be
hemorrhage. If there is a shift across the mid- reflected to the areas adjacent to the tumor.
line or compromise of the midline structures, Sometimes the more focal slow-wave abnor-
intermittent rhythmic delta activity may also be mality may be masked by other EEG activity,
present. muscle, or eye movement artifact (Fig. 10–2).12
At times more widespread bursts of intermit-
Key Point
tent rhythmic slow waves are present.
• Intracerebral hemorrhage is associated Asymmetry of background activity may also
with more severe slow-wave abnormali- be present. Usually the background activity is
ties. decreased on the side of the tumor. On occa-
sion the background activity may be increased,
particularly if a skull defect is present.7, 8
Arteriovenous Malformations Slower growing tumors are more likely
to produce epileptiform abnormalities than
The most frequent finding in arteriovenous
faster growing tumors. Thus, the oligo-
malformations is focal epileptiform activity.
dendroglioma, a slow growing tumor, has
Other types of findings that may be present are
the highest incidence of associated epilep-
focal slowing or asymmetry of activity.3, 5
tiform abnormalities followed by astrocy-
Key Point tomas, meningiomas, metastatic carcinomas,
and glioblastomas. Certain areas of the brain
• Arteriovenous malformations are usually
are more epileptogenic or have a lower thresh-
associated with focal epileptiform abnor- old for developing epileptiform abnormalities
malities. than other areas of the brain, and tumors
involving the temporal lobe, the rolandic
Tumors strip, and the frontal lobe are more likely
to produce epileptiform abnormalities than
The most common finding overlying the site tumors affecting the parietal and occipital
of the tumor is the presence of persistent, lobes.
Figure 10–2. Focal polymorphic delta activity in the right frontal region (Fp2) that is clearly evident when the patient’s
eyes are open but masked by eye movement when the eyes are closed.
The type and degree of EEG abnormality Complications, such as vascular compromise,
depend on various factors. Focal abnormal- infarct, hemorrhage, edema, pressure effects,
ities are more likely to occur with a lesion and increased intraventricular pressure, can
near the surface of the brain as compared to also contribute to the degree of EEG abnor-
a more deeply situated lesion.3, 12 The extent malities.
and degree of slowing are related to the size
of the tumor. The more rapidly growing and
malignant the tumor, the more severe the slow- Key Points
wave abnormalities. The stage of evolution at
which the EEG was recorded determines the • The EEG in patients with cerebral tumors
degree of abnormality. If the EEG is recorded may show focal slowing, asymmetry, or
at an early stage, there may be little abnormal- epileptiform activity.
ity reflected in the EEG as compared to a later • The slowest frequency and most persis-
stage (Fig. 10–3).12 The state of the patient and tent slowing are seen over the site of the
level of consciousness can affect the degree of tumor.
the EEG abnormalities. Another factor is the • An asymmetry of activity may be seen with
age of the patient. Children tend to show more a decrease on the side of the tumor.
severe abnormalities than adults. Medications • An increased amplitude of activity can be
can cause diffuse slowing of the EEG and/or seen with a skull defect.
accentuate more focal slow-wave abnormali- • Epileptiform activity is more likely to
ties. The presence of a skull defect can enhance occur with slow growing tumors.
slowing or spike or sharp wave-like activity. • Tumors involving the temporal and frontal
Seizures and a postictal effect can also affect regions are more likely to be associated
the EEG, particularly if the EEG is recorded with epileptiform activity.
shortly after a seizure. Interictal activity, such • The location, size, state of evolution, med-
as temporal intermittent rhythmic delta activity ication and seizures all play a role in
(TIRDA), can also affect the EEG (Fig. 10–4). determining the degree of EEG findings.
Figure 10–3. Electroencephalograms from a 67-year-old man showing focal persistent polymorphic delta activity in the
right temporal region (F8 and T4) (segment on left), and the addition 2½ months later (segment on right) of intermittent
rhythmic delta activity caused by a rapidly progressive brain tumor.
Figure 10–4. TIRDA over right tem-

poral region in a 75-year-old man with
complex partial seizures.
Head Trauma Epileptiform abnormalities are usually

uncommon during the acute period follow-
Head trauma may be associated with a variety ing head trauma, but if present, usually indi-
of EEG abnormalities.3–5 Mild head trauma cate a significant brain injury. The patients
may be associated with a transient asymme- may have associated focal seizures in asso-
try or mild slowing on the side of the head ciation with focal epileptiform discharges on
injury. Severe head injuries may be associ- their EEG (Fig. 10–5). Generalized spike-
ated with significant slowing, an asymmetry and-wave discharges are uncommon following
or suppression, and/or epileptiform activity. head injury and usually reflect a preexisting
The degree of slowing often parallels the seizure disorder. Patients who are most likely
degree of head injury and level of conscious- to have epileptiform abnormalities on their
ness. An asymmetry or suppression of activity EEG and seizures are those with penetrating
may also be present over an area or side of head injuries, depressed skull fractures, pro-
the brain that has received more significant longed periods of altered consciousness, and
trauma. long periods of retrograde amnesia.3, 5, 12
Figure 10–5. Electroencephalogram from a 71-year-old man with recent head trauma showing left-sided PLEDs.
Key Points in evaluation of patients with inflammatory

disorders.
• Mild head trauma is associated with tran-
Brain abscesses may occur as a result of
sient slowing and/or asymmetry meningitis, septicemia, septic emboli, or as an
• Severe head injuries are associated with
extension of an infectious process involving the
slowing, asymmetry, and suppression of ears, mastoids, and sinuses. Focal polymorphic
activity delta slowing can develop overlying the site of
• Epileptiform abnormalities may be present
the abscess, particularly if the lesion is located
in patients with severe head trauma close to the surface of the brain. On infre-
• Epileptiform abnormalities are more likely
quent occasions, focal or lateralized periodic
to be seen in patients with penetrating sharp- or slow-wave complexes and PLEDs
head injuries, skull fractures, prolonged may be present over the involved area of the
periods of unconsciousness, and amnesia. brain.3, 11, 13, 14 Infratentorial abscesses produce
less severe slow-wave abnormalities, and at
times there may be little or no change on
Subdural Hematoma the EEG.
After treatment, the slow-wave abnormali-
A subdural hematoma is one of the com- ties improve. There is a decrease in the degree
mon complications of head injury. The most of slow-wave abnormalities within the first
common findings produced by a subdural few days after the surgery; however, some
hematoma are asymmetry—with decreased slowing and asymmetry of activity can con-
amplitude on the side of the subdural—and tinue to be present over the involved area.3
focal or lateralized slowing. Intermittent rhyth- Epileptiform abnormalities are not very com-
mic delta slowing may also be present.3, 4 It mon in the acute stages of the abscess; how-
is often difficult to distinguish on a single ever, patients with cerebral abscess can sub-
recording the effects produced by a subdural sequently develop seizures and epileptiform
hematoma versus that of a cerebral contu- activity in their EEG.
sion during the first 1 or 2 weeks following
a severe head injury. A subdural hematoma Key Points
should be considered, however, if a later EEG • Brain abscess is associated with focal
shows increased focal or generalized slowing or
slowing.
the combination of a focal abnormality, either • PLEDs and epileptiform discharges may
slowing or asymmetry, in association with inter-
be seen with an acute abscess.
mittent rhythmic delta activity.12 • Epileptiform abnormalities may occur
In summary, the EEG can be helpful in fol-
later following treatment of an abscess.
lowing the patients with head trauma in deter-
mining or confirming the presence of seizures,
seeing if patients develop focal complications
such as intracerebral or subdural hematoma, or
Herpes Simplex Encephalitis
diffuse disorders, including metabolic, toxic, or
The EEG often shows a characteristic pattern
inflammatory encephalopathies.
of PLEDs, which can be of great value in
Key Point the diagnosis of herpes simplex encephalitis.
The periodic pattern is usually seen between
• Subdural hematomas are associated with 2 and 5 days after the onset of the illness,
asymmetry of activity and slowing. but on occasion has been observed several
days later. Usually the complexes occur in a
unilateral fashion; however, if there is bilat-
Inflammatory Disorders eral involvement of the brain, bilateral PLEDs
can occur synchronously or independently over
Focal abscess or cerebritis can be associated the two hemispheres.3, 13, 14 Focal or lateral-
with focal EEG abnormalities. This can con- ized electrographic seizure discharges may be
sist of focal polymorphic delta activity and/or present over the involved area or hemisphere.
associated epileptiform activity.3, 4, 14 The EEG The discharges may or may not be associated
can be helpful in determining whether there with clinical seizures. In nonfatal herpes sim-
is a focal, lateralized, or more diffuse process plex encephalitis, the PLEDs disappear as the
disease process resolves. The EEG, however, postictal changes may be present from several
frequently continues to show subsequent slow- minutes to hours after the seizure.
wave abnormalities, loss of background activity,
or focal epileptiform activity over the involved Key Points
area. • Transient disorders such as migraine
headaches, transient ischemic attacks,
Key Points and postictal states may cause transient
• The characteristic finding in herpes sim- changes in the EEG.
• The EEG during or following a migraine
plex encephalitis is PLEDs.
• PLEDs may be unilateral or bilateral. headache may show slowing and an asym-
• In nonfatal herpes simplex encephalitis, metry of activity.
• The EEG during a transient ischemic
the EEG improves but continues to show
episode shows slowing and/or decreased
slowing, loss of background activity, and
amplitude of activity over the involved
epileptiform activity.
area.
• Focal or lateralized slowing or asymmetry
can be seen following a focal or lateralized
Other Focal Cerebral Lesions seizure discharge.
Other types of focal cerebral lesions such as a In summary, different types of focal cerebral
cyst or congenital defects can also cause focal lesions can cause similar types of EEG abnor-
slowing, asymmetry, suppression, or epilepti- malities and although the EEG findings are not
form abnormalities on the EEG.3–5 specific for a single lesion, they do give evi-
dence of a focal disturbance of cerebral func-
tion. Sequential recordings may be necessary
to distinguish a resolving or static lesion from a
Transient Disorders more progressive lesion. Although neuroimag-
ing studies are now used to evaluate patients
Transient disorders that can cause tran-
with suspected focal lesions, the EEG is still
sient changes in the EEG include migraine
very useful as it reflects function or dysfunction
headaches, transient ischemic episodes, and
of the brain and can indicate the presence of
postictal effects.
focal or lateralized pathophysiologic processes
Migraine headaches can be accompanied
that may not be detectable on neuroimaging
by focal or lateralized slowing and/or asym-
studies.12
metry on the side of the migraine headache.
These changes can occur during the migrain-
ous episode and persist for 1 to 2 days in adults.
Hemiplegic migraine is associated with hemi- ELECTROENCEPHALOGRAPHIC
spheric slowing and loss of background activity MANIFESTATIONS OF DIFFUSE
on the side of the migrainous episode and may DISORDERS
persist for several days before resolving.3
If a recording is performed during a tran- The EEG is helpful in the evaluation of dif-
sient ischemic attack, the EEG can show focal fuse disorders of cerebral function and serves
delta slowing and/or decreased amplitude of as a measurement of the severity of the dis-
activity over the involved area.3, 4 Usually the turbance.3, 5, 15 Diffuse encephalopathies can
EEG abnormality resolves in association with be caused by various conditions, including
resolution of the patient’s symptoms. If the metabolic, toxic, inflammatory, posttraumatic,
abnormality persists beyond 24 hours, then this hypoxic, and degenerative disorders.16
would indicate that an infarct has occurred. The type of diffuse disorder and whether
Focal or lateralized slowing and/or an asym- it involves white or gray matter influence
metry of activity may occur as a postictal effect the EEG pattern.1, 2 Processes that predomi-
following a seizure discharge. This can be help- nantly affect superficial white matter usually
ful in indicating or confirming the focal or cause polymorphic delta slowing in the EEG,
lateralized onset of a seizure discharge. The whereas diffuse processes that involve both
cortical and subcortical gray matter are more rhythmic slow waves can occur in a general-
likely to cause intermittent bilaterally syn- ized fashion or have a maximal expression over
chronous paroxysmal slow-wave activity.1, 2, 5, 6 the anterior or posterior head regions. Usually
Epileptiform abnormalities are seen more the degree of slowing parallels the degree of
commonly in gray matter disease than in white disturbance of function or alteration in level
matter disease. Other factors that influence of consciousness (or both). These findings can
the degree and type of EEG abnormalities be caused by various diffuse disorders and,
include the state of the patient, the type and therefore, are considered nonspecific changes
stage of the disease process, contributing fac- in that they are not diagnostic of any single
tors such as infectious processes, metabolic condition.
derangements, drug effects, and/or other sys-
temic processes. Key Points
• The most common EEG finding in diffuse
Key Points
disorders is slowing.
• Processes affecting white matter usually • The slowing may consist of slowing of the
cause delta slowing. background, theta and/or delta slowing, or
• Diffuse processes involving cortical and intermittent rhythmic bisynchronous slow
subcortical structures often cause inter- waves.
mittent slow-wave activity. • The degree of slowing reflects the degree
• Epileptiform abnormalities are seen with of disturbance of function and altered
gray matter disease. consciousness.
• Factors influencing the degree and type
of EEG abnormality include state of the
patient, type of disease process, and sys- Specific Patterns
temic disorders.
At times, however, the EEG may show a
more specific pattern, such as periodic pat-
terns or the various distinctive coma patterns.
Slow-Wave Abnormalities The generalized periodic patterns include
those associated with Creutzfeldt–Jakob dis-
The most common type of EEG finding in dif- ease, subacute sclerosing panencephalitis, and
fuse disorders, or encephalopathies, consists hepatic coma.3, 4, 10 The distinctive coma pat-
of slowing of varying degrees (Fig. 10–6).3–5, 16 terns include alpha, beta, spindle coma, and
This may involve slowing of the background burst-suppression patterns.
activity, slowing in the theta frequency range,
or generalized polymorphic delta slowing.16
Intermittent bursts of bilaterally synchronous Generalized Periodic Patterns
Creutzfeldt–Jakob disease is a diffuse, suba-
cute, and progressive disorder of the central
nervous system that occurs predominantly in
middle-aged patients. It is characterized by
dementia, motor dysfunction, myoclonus, and,
when the disease is fully developed, a char-
acteristic periodic EEG pattern consisting of
generalized, bisynchronous, and periodic sharp
waves recurring at intervals of 0.5–1 second,
with a duration of 200–400 ms (Fig. 10–7).3, 10, 14
Myoclonic jerks are often associated with
the periodic sharp waves; however, there is
not always a constant relationship between
Figure 10–6. Severe diffuse slow-wave (delta) abnor- the two. Although occasionally other degen-
mality in a 9-year-old boy with encephalitis. erative or toxic disorders may be associated
Figure 10–7. Diffuse periodic sharp waves in a 71-year-old man with Creutzfeldt–Jakob disease.
with a quasiperiodic sharp-wave pattern, the present during the intermediate stages of the
presence of periodic sharp waves, progressive disease. In a single recording from a single
dementia, and myoclonus is strongly suggestive patient, the morphology of the complexes is
of Creutzfeldt–Jakob disease. stereotyped; however, the shape of the com-
Subacute sclerosing panencephalitis (SSPE) plexes can vary in different patients and change
is a degenerative disorder that occurs in chil- from time to time in the same patient at dif-
dren and adolescents and is believed to be ferent stages of the disease. The complexes are
caused by the measles virus. This degenerative usually generalized and bisynchronous, but at
disorder is characterized by abnormal move- times they may be asymmetrical or more later-
ments, intellectual deterioration, and a diag- alized. Stereotyped motor jerks or spasms are
nostic, periodic EEG pattern. This consists of often associated with the periodic complexes.
repetitive stereotyped high-voltage sharp- and In hepatic coma, the EEG often shows a
slow-wave complexes recurring every 4–15 sec- triphasic wave pattern consisting of medium-
onds (Fig. 10–8).3, 10, 14 This pattern is usually to high-voltage broad triphasic waves that
Figure 10–8. Diffuse periodic complexes in a 12-year-old girl with subacute sclerosing panencephalitis.
occur rhythmically or in periodic serial trains Key Points

at a rate of 1–2 Hz in a bilaterally syn- • Creutzfeldt–Jakob disease is associated
chronous and symmetrical fashion over the two
hemispheres with a fronto-occipital or occip- with generalized periodic sharp waves.
• SSPE is associated with generalized peri-
itofrontal time lag. The triphasic waves usually
have a frontal predominance and consist of odic slow-wave complexes.
• Hepatic coma is associated with triphasic
a short-duration, low-voltage surface-negative
component followed by a prominent positive waves.
sharp-contoured wave and then a longer dura-
tion surface-negative slow wave (Fig. 10–9).17
Although triphasic waves are often associ- Coma Patterns
ated with liver dysfunction, atypical triphasic
waves can be seen in other conditions, includ- Comatose patients often show generalized slow
ing metabolic derangements, electrolyte dis- waves or abnormalities; however, more specific
turbances, toxic states, degenerative processes, coma patterns may be present in the EEG.
or after a hypoxic episode. The triphasic wave The specific coma patterns include alpha- and
pattern needs to be carefully distinguished beta-frequency coma, spindle coma, and burst-
from epileptiform activity.3, 17 suppression patterns.
Figure 10–9. Triphasic waves in a 77-year-old woman in hepatic coma.

The alpha-frequency coma pattern consists pattern usually indicates a poor prognosis.3, 15
of diffusely distributed invariant alpha activity However, the burst-suppression pattern can
that shows little or no reactivity or variability. also be seen with potentially reversible condi-
This type of pattern has been seen after car- tions, such as anesthesia, drug intoxication, and
diac arrest or hypoxic insult to the brain and hypothermia.
with significant brain stem lesions.3, 18 When With regard to the EEG in coma and
the alpha-frequency coma pattern is seen in the prognosis, favorable findings in EEG include
context of a hypoxic insult, it usually indicates variability, reactivity, variable wake-and-sleep
a poor prognosis. A reversible alpha-frequency patterns, and a progressive increase in back-
coma pattern, on the other hand, can be seen ground frequencies. Patterns that indicate a
with medications, anesthetic agents, or over- poor prognosis for return of useful neurologic
dose of drugs.18 function include an invariant monorhythmic
The beta-frequency coma pattern consists pattern with little or no reactivity, a burst-
of generalized beta activity superimposed on suppression pattern, and generalized suppres-
underlying delta slowing. This pattern is usu- sion of activity.
ally associated with drug toxicity or anesthesia.
The spindle coma pattern resembles a sleep
EEG and consists predominantly of spindle
Key Points
activity with some V waves, but it shows lit-
tle or no reactivity.3–5 This type of pattern • Patients in coma often show generalized
can result from various causes, including head slow-wave abnormalities.
trauma, hypoxic insults, or brain stem lesions. • The alpha-frequency coma pattern is seen
Depending on the type of underlying cause following cardiac arrest or with a brain
and severity of damage to the central nervous stem lesion.
system, the pattern indicates that the poten- • A reversible alpha-frequency coma pat-
tial for improvement exists. In many types of tern can be seen with overdose of drugs
coma, spontaneous variability of EEG activ- or anesthesia.
ity, including the sleep-like pattern, indicates • A beta-frequency coma pattern is seen
a better prognosis than a prolonged invariant with drug toxicity or anesthesia.
pattern. • The spindle coma pattern can be seen
The burst-suppression pattern consists of with head trauma, hypoxic insults, or brain
periodic or episodic bursts of activity, usu- stem lesions.
ally irregular mixtures of sharp waves or • The burst-suppression pattern can be seen
spikes, alternating with intervals of suppression following anoxic insult to the brain and
(Fig. 10–10). The bursts may be accompanied usually indicates a poor prognosis.
by myoclonic jerks. This pattern is often seen • Burst suppression can also be seen with
after a severe insult to the brain, such as a potentially reversible conditions such as
hypoxic or anoxic insult, in which case the drug overdose, anesthesia, or hypothermia.
Figure 10–10. Diffuse burst-suppression pattern after cardiac arrest.

• EEG variability, reactivity, variable wake recording in patients with suspected cerebral
and sleep patterns and a progressive death. These criteria include the following:19
increase in background frequencies indi-
cate a favorable prognosis in coma.
• EEG patterns that indicate a poor 1. A minimum of eight scalp electrodes
prognosis in coma include an invari- should be used.
ant monorhythmic pattern with little or 2. Interelectrode impedances should be
no reactivity, a burst-suppression pat- less than 10,000 V but more than 100 V.
tern, and a generalized suppression of 3. The integrity of the entire recording sys-
activity. tem must be verified.
4. Interelectrode distances should be at
least 10 cm.
In summary, in patients with diffuse disorders, 5. The sensitivity should be at least
the EEG is useful in documenting a distur- 2 μV/mm for at least 30 minutes of
bance of cerebral function, in determining the recording.
degree of disturbance, in monitoring changes 6. Appropriate filter settings should be
and trends in the course of the disease process, used.
and in helping to establish the diagnosis in cer- 7. Additional monitoring techniques should
tain conditions in which a characteristic EEG be used when necessary.
pattern is present. Also, the EEG sometimes 8. There should be no EEG reactivity to
helps to detect the presence of an additional, afferent stimulation.
more focal cerebral process. 9. The recording should be made by a qual-
ified technologist.
10. A repeat EEG should be performed if
EVALUATION FOR SUSPECTED there is doubt about the presence of
electrocerebral silence.
BRAIN DEATH
The EEG can provide confirmatory evidence Because temporary and reversible ECI can
of brain death, which is manifested by an be caused by drug overdose and hypothermia,
absence of spontaneous or induced electric these conditions should be excluded before
activity of cerebral origin (Fig. 10–11). Elec- reaching a conclusion of brain death.3 In young
trocerebral inactivity (ECI) is defined as “no infants, because of uncertainties about the sig-
EEG activity over 2 microvolts/mm.”19 There nificance of ECI, one should exercise cau-
are important minimal technical criteria for tion in the interpretation of this finding.3 The
Figure 10–11. Electrocerebral inactivity.

Guidelines of the Task Force for the Deter- 3. Monitor changes in the course of a dis-
mination of Brain Death in Children20 recom- ease process and help determine whether
mend that for infants between 7 days and 2 the patient is improving or deteriorating.
months old, two EEGs demonstrating ECI be 4. Rule out the possibility of a more focal
performed at least 48 hours apart, and that for cerebral process such as an expanding
infants between 2 months and 1 year old, two mass lesion.
records showing ECI be done at least 24 hours 5. Make the diagnosis in certain conditions
apart. when a characteristic EEG pattern is
present.
Key Points 6. Help in the evaluation of suspected brain
death.
• Electrocerebral inactivity is defined as no
EEG activity over 2 μV/mm.
• EEG in suspected brain death must meet In conclusion, this chapter provides an overview
specific performance criteria. of nonepileptiform abnormalities in the EEG.
• Drug overdose and hypothermia should The EEG findings can be seen with a number
be excluded as potentially reversible of different focal lesions or diffuse disorders.
causes of cerebral inactivity. As Dr. Donald Klass has stated “the useful-
ness of the EEG depends on an enlightened
interpretation of the findings with regard to the
CONCLUSION specific clinical problem that the EEG is being
used to solve.”12
Since the EEG reflects altered function or
pathophysiology of the brain, the EEG can be
very helpful in evaluating patients with focal
or diffuse disorders, altered consciousness, or REFERENCES
who are comatose. In evaluation of patients,
the EEG should also be interpreted in asso- 1. Gloor, P., O. Kalabay, and N. Giard. 1968. The elec-
ciation with other factors including state of troencephalogram in diffuse encephalopathies. Elec-
troencephalographic correlates of grey and white
consciousness, reactivity, neurologic function matter lesion. Brain 91:779–802.
or dysfunction, drugs or medications, under- 2. Gloor, P., G. Ball, and N. Schaul. 1977. Brain lesions
lying metabolic, cardiovascular, pulmonary, or that produce delta waves in the EEG. Neurology
other systemic medical problems, and age of 27:326–33.
3. Ebersole, J. S, and T. A. Pedley, eds. 2003. Current
the patient. The EEG can be particularly help- practice of clinical electroencephalography, 3rd ed.
ful in patients who are comatose, on respi- Philadelphia: Lippincott Williams & Wilkins.
rators, are paralyzed, or when the neurologic 4. Fisch, B. J. 1999. Spehlmann’s EEG primer,principles
status cannot be evaluated. of digital and analog EEG, 3rd ed. Amsterdam: Else-
The occurrence of certain patterns and vier Science Publishers.
5. Niedermeyer, E., and F. Lopes da Silva, eds.
the presence of variability and reactivity of 2005. Electroencephalography: Basic principles, clin-
the EEG can help ascertain the progno- ical applications, and related fields, 5th ed. Philadel-
sis or potential for improvement. Sequential phia: Lippincott Williams & Wilkins.
recordings are very helpful in determining 6. Watemberg, N., F. Alehan, R. Dabby, T. Lerman-
Sagie, P. Pavot, and A. Towne. 2002. Clinical and
whether the patient is improving or deteriorat- radiological correlates of frontal intermittent rhyth-
ing or developing other complications such as mic delta activity. Journal of Clinical Neurophysiology
seizures, metabolic encephalopathies, or toxic 19:535–9.
or medication effect. Specific EEG patterns, 7. Cobb, W. A., R. J.Guiloff, and J. Cast. 1979. Breach
when present, can indicate or give a clue as to Rhythm. The EEG related to skull defects. Elec-
troencephalography and Clinical Neurophysiology
what the underlying process is. 47:251–71.
In summary, the EEG can: 8. Radhakrishnan, K., D. Chandy, G. Menon, and
S. Sarma. 1999. Clinical and electroencephalographic
1. Document a disturbance of cerebral correlates of breach activity. American Journal of Elec-
troneurodiagnostic and Technology 39:138–47.
functioning. 9. Westmoreland, B. F., and D. W. Klass. 1998. Defective
2. Determine the degree of disturbance of alpha reactivity with mental concentration. Journal of
cerebral functioning. Clinical Neurophysiology 15:424–28.
10. Brenner, R. P., and N. Schaul. 1990. Periodic 15. Young, G. B. 2000. The EEG in coma. Journal of
EEG patterns: Classification, clinical correlations, and Clinical Neurophysiology 17:473–85.
pathophysiology. Journal of Clinical Neurophysiology 16. Kaplan, P. W. 2004. The EEG in metabolic
7:249–67. encephalopathy and coma. Journal of Clinical Neuro-
11. Chatrian, G. E., C.-M. Shaw, and H. Leffman. 1964. physiology 21:307–18.
The significance of periodic lateralized epileptiform 17. Fisch, B. J, and D. W. Klass. 1988. The diag-
discharges in EEG: An electrographic, clinical and nostic specificity of triphasic wave patterns. Elec-
pathological study. Electroencephalography and Clini- troencephalography and Clinical Neurophysiology
cal Neurophysiology 17:177–93. 70:1–8.
12. Klass, D. W., and B. F. Westmoreland. 2001. Elec- 18. Kaplan, P. W, D. Geroud, T. W. Ho, and P. Jallon.
troencephalography: General principles and adult 1999. Etiology, neurologic correlations, and prog-
EEGs. In Clinical neurophysiology, ed. J. R. Daube, nosis in alpha coma. Clinical Neurophysiology 110:
2nd ed., 77–107. Oxford University Press: New York. 205–13.
13. Pohlman-Eden, B., D. B. Hoch, J. I. Cochius, and 19. American Electroencephalography Society. 2006.
K. H. Chiappa.1996. Periodic lateralized epileptiform Guideline 3: Minimal technical requirements for
discharges—A critical review. Journal of Clinical Neu- recording in suspected cerebral death. Journal of Clin-
rophysiology 13:519–30. ical Neurophysiology 23:97–104.
14. Westmoreland, B. F. 2005. EEG in cerebral inflam- 20. Task force for the determination of brain death
matory processes. In Electroencephalography, ed. in children. Guidelines for the determination of
E. Niedermeyer, and F. Lopes da Silva, 5th ed., brain death in children. 1987. Neurology 37:
323–37. Philadelphia: Lippincott Williams & Wilkins. 1077–78.
Chapter 11
Electroencephalography:
Electroencephalograms of Infants
and Children
INTRODUCTION Posterior Slow Waves Associated

NEONATAL EEG PATTERNS with Eye Blinks or Slow Lambdas
Normal EEGs of Youth
Abnormal EEGs O Waves or Cone-Shaped Waves
Posterior Slow Waves of Youth
DEVELOPMENTAL CHANGES 14 and 6 Hz Positive Spike Bursts
DURING INFANCY,
CHILDHOOD, AND ADOLESCENCE ABNORMALITIES
Hyperventilation Epileptiform Abnormalities
Photic Responses Slow-Wave Abnormalities
Drowsiness Conditions Giving Rise to
Sleep Abnormal EEG Patterns in Children
Transient Abnormalities
BENIGN VARIANTS IN CHILDREN SUMMARY
INTRODUCTION disorder. The EEG is helpful in the evaluation

of nonepileptic spells and transient CNS symp-
Electroencephalography (EEG) is an impor- toms by indicating whether there is an asso-
tant part of the evaluation of many disorders ciated EEG change and, if so, the degree,
in infants and children, including seizures, location, and type of change. Ambulatory and
spells, transient central nervous system (CNS) prolonged video EEG monitoring can be of
symptoms, behavioral disorders, altered states additional help in the evaluation of seizures,
of consciousness, and lesions or conditions spells, and transient symptom. Another com-
resulting in a disturbance of cerebral func- mon referral is for evaluation of children who
tion. The most common referral for EEG is have altered consciousness or are comatose.
for evaluation of seizures, because the EEG The EEG can help determine the degree
can show the presence of epileptiform abnor- and extent of the disturbance and, if spe-
malities and, often, indicate the type of seizure cific patterns are present (as described in
167
Chapter 10), help indicate the diagnosis or eye and body movements, increased muscle
prognosis (or both). tone, regular respirations, and a regular ECG.
The EEG is also helpful in evaluating con- During this state, the EEG shows a discontin-
ditions or lesions causing a disturbance of uous pattern, with bursts of mixed sharp- and
cerebral function, in determining whether the slow-wave activity alternating with periods of
process is focal or generalized, and in identi- flattening of the background. This is referred
fying the extent of the disturbance. The EEG to as the tracé alternant pattern.
reflects the degree and extent of the distur-
bance and, if certain diagnostic EEG patterns Key Points
are present, helps to make the diagnosis.
Children with behavioral disturbances, atten- • Active sleep is similar to REM sleep in
tion deficits, or learning disorders are also adults.
referred for EEGs to rule out an underlying • Active sleep in infants consists of eye
organic process. movement, body twitches, reduction in
muscle tone, irregular respirations, and a
Purpose of EEG Studies in Infants continuous low amplitude EEG pattern.
and Children • Quiet sleep is similar to non-REM sleep
in adults.
• Evaluate seizures, transient spells, and • Quiet sleep consists of reduced eye and
altered states of consciousness. body movements, increased muscle tone,
• Evaluate conditions and lesions causing a regular respirations and ECG, and a dis-
disturbance of cerebral function. continuous pattern called the tracé alter-
• Determine whether the condition or nant pattern.
lesion is focal or generalized.
• Help make the diagnosis if certain diag- Other types of activity that are present at this
nostic EEG patterns are present. age include2–6
1. Occipital dominant slow waves, which are

NEONATAL EEG PATTERNS broad, high-amplitude slow waves over
the occipital head regions, shifting from
Normal EEGs side to side in prominence.
2. The spindle delta brush pattern, which
Recordings in newborn infants usually require consists of moderate-to-high-amplitude
monitoring of physiologic variables such as delta waves with superimposed 8- to
the electrocardiogram (ECG), respirations, eye 20-Hz activity; it is seen over the
movements, and chin myogram in addition to rolandic, temporal, and occipital head
EEG recording to help determine the state of regions (Fig. 11–1).
the infant.1 3. Random multifocal sharp transients or
In infants of less than 32 weeks’ concep- focal sharp waveforms occurring in mul-
tional age, which is the age since the first day of tiple locations but most frequently over
the mother’s last menses, the EEG consists of the frontal head regions (frontal sharp
an intermittent or discontinuous pattern, with transients) (Fig. 11–2).
bursts of activity alternating with long quies- 4. Anterior slow waves, which consist of
cent periods. There is little distinction between rhythmic delta slow waves over the ante-
the EEG of the awake and sleep states.2–6 rior head regions; they are usually associ-
Between 32 and 37 weeks’ conceptional age, ated with frontal sharp transients.
there is an EEG distinction between the awake 5. Bursts or trains of sharp-contoured theta
state and the two types of sleep states, active waves over the temporal and central ver-
and quiet sleep.2–6 Active sleep, which is similar tex regions. In the younger, premature
to rapid eye movement (REM) sleep in adults, infant, these patterns are seen in both the
manifests eye movements, body twitches, gri- awake and sleep states.
maces, reduction in muscle tone, and irreg-
ular respirations. During this state, the EEG As the child matures, these patterns are seen
shows a more continuous pattern, similar to primarily during sleep and then disappear after
that of the awake state. Quiet sleep has reduced about 44 weeks’ conceptional age.
Electroencephalograms of Infants and Children 169
Age: 32 wks C.A.

FP1-C3
C3-O1
FP1-T3
T3-O1
FP2-C4
C4-O2
FP2-T4
T4-O2
Asleep 70 μV
1 sec
Figure 11–1. EEG from a normal premature infant at 32 weeks’ conceptional age (C. A.) during sleep, showing the delta
brush pattern.
Infants 38–42 weeks old show four basic ◦ Spindle delta brush pattern
patterns:2–6 ◦ Anterior slow waves
◦ Sharp-contoured theta waves in the
1. A low-voltage irregular pattern that is pre- temporal and central regions
sent during wakefulness and active sleep.
2. A high-voltage slow-wave pattern that is
seen during quiet sleep.
3. A tracé alternant pattern that is also seen Abnormal EEGs
during quiet sleep.
4. A mixed pattern of theta and delta waves The EEG in premature and newborn infants
seen during drowsiness and active sleep should be interpreted with care. The age of the
and as a transitional pattern between the infant is important because the EEG activity
various states. changes with maturation, and what might be
normal at one age may be abnormal for a more
Key Points mature infant. The background rhythms are
the most significant EEG finding.2–7 If these
• Specific patterns and waveforms in pre- are appropriate for the infant’s age, the infant
mature and neonatal infants consist of usually has a fairly good prognosis; if abnormal,
◦ Tracé alternant pattern then the degree of abnormality usually reflects
◦ Occipital dominant slow waves the degree of disturbance and the ultimate
Age: 34 wks C.A.

FP1-C3
C3-O1
FP1-T3
T3-O1
FP2-C4
C4-O2
FP2-T4
T4-O2
Asleep 50 μV
1 sec
Figure 11–2. EEG from a normal premature infant at 34 weeks’ conceptional age (C.A.) during sleep, showing a frontal
sharp transient (Fp2). Less prominent sharp transients also occur in T3 and T4.
outcome of the infant. Abnormalities may be disturbance of function or, in a more focal fash-
divided into mild and significant types. ion, in association with focal lesions such as
porencephaly, subdural collection of fluids, or
congenital abnormalities.
MILD ABNORMALITIES
Mild abnormalities, such as excessive multifocal
Epileptiform Activity
sharp transients or immature and dysmature
patterns, can be seen in stressed premature Epileptiform activity is one of the most fre-
or term infants. These findings are nonspe- quent types of abnormalities seen in EEGs
cific and rarely suggest a specific diagnosis.2, 4 of neonates and consists of focal or multifocal
They often are transient and usually disappear interictal and ictal discharges.2–9 The interic-
within a few days. Mild and transient focal tal discharges usually take the form of spikes,
abnormalities in the EEG usually are not asso- sharp waves, and broad slow waves. The ictal
ciated with any obvious focal pathologic condi- discharges consist of rhythmic activity that may
tion. However, persistent focal EEG abnormal- take the form of spikes, sharp waves, slow
ities are often associated with structural lesions waves, or rhythmic activity in the alpha, beta,
such as intracranial hemorrhage or congenital theta, or delta range and may evolve and persist
defects.2–7 for relatively long periods. Ictal electrographic
discharges often occur in association with clin-
ical seizures but may be present without any
SIGNIFICANT ABNORMALITIES clinical accompaniment (Fig. 11–3). If associ-
Significant abnormalities in the EEG are usu- ated with seizures, the seizures usually take the
ally associated with an important disturbance form of clonic or tonic movements, but there
of brain function and often indicate a poor may be diverse and subtle manifestations that
prognosis or poor neurologic outcome. The may not be easily recognizable as epileptic.2–9
more abnormal the pattern, the more severe
the underlying encephalopathy or disturbance
of brain function.2–7 The following patterns are Positive Rolandic Sharp Waves
significantly abnormal. Positive rolandic sharp waves occur unilater-
ally or bilaterally and are most common in
the rolandic and midline areas.2–6 They were
Isoelectric EEG described initially in infants with intraventric-
An isoelectric EEG is a flat record that ular hemorrhage; however, they also occur in
meets the criteria for electrocerebral inactivity. patients who have periventricular leukomalacia
Infants with a single flat EEG may survive and deep white matter lesions.2–6
the neonatal period but usually suffer severe
long-term neurologic sequelae.2–7
Asymmetry
An excessive and persistent asymmetry of the
Burst-Suppression Pattern
activity during both the wake and sleep states,
A burst-suppression pattern consists of dif- occurring focally or lateralized to one hemi-
fuse bursts of abnormal activity superimposed sphere, is a significant abnormal finding in an
on an isoelectric or very low-amplitude back- infant’s EEG.2–7 This can occur with congenital
ground. This is an invariant pattern that does lesions, porencephalic cysts, vascular insults, or
not change with state of sleep-wakefulness or subdural collections of fluid.
in response to stimuli. It, too, is associated
with severe encephalopathy and poor long-
term prognosis.2–7 Periodic Discharges
Periodic lateralized epileptiform discharges
(PLEDs) can occur with an acute or subacute
Persistent Low Voltage
process, most often caused by an ischemic,
Persistent low voltage can occur in a gen- hypoxic, or vascular insult, or neonatal herpes
eralized fashion in association with a diffuse simplex encephalitis.10, 11
Age: 7 days
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
Cretin, Hypocalcemia,
Hypoglycemia, Kernicterus 150 μV
1 sec
Figure 11–3. Focal subclinical EEG-seizure discharge arising from the left frontal region (F3) in a 7-day-old girl with
neonatal seizures.
Persistent Slowing Metabolic Disorders

Persistent focal or generalized slowing in the The most common metabolic disorders that
delta and theta range with decreased reac- produce abnormal EEG patterns are hypo-
tivity can occur with a focal lesion or a diffuse glycemia and hypocalcemia, which previously
process.2–6 were the most frequent causes of neonatal
seizures. Metabolic disorders can be associated
Key Points with focal or multifocal epileptiform abnormal-
ities in the EEG; the EEG usually improves
• Mild abnormalities consist of excessive after the metabolic disorder has been cor-
multifocal sharp transients or immature rected.2–9
and dysmature patterns.
• Significant abnormalities consist of Drugs and Drug Withdrawal
◦ Isoelectric EEG
◦ Burst-suppression pattern Drug withdrawal is becoming a more fre-
◦ Persistent low-voltage pattern quent cause of seizures in newborn infants.
◦ Epileptiform activity The EEG in these infants often shows evidence
◦ Positive Rolandic sharp waves of cortical irritability, as manifested by focal
◦ Excessive asymmetry or multifocal epileptiform activity and seizure
◦ Periodic discharges discharges. Drugs such as anticonvulsants can
◦ Persistent slowing also cause changes in an infant’s EEG, includ-
ing a burst-suppression pattern or generalized
suppression of activity.2, 5–7
CONDITIONS GIVING RISE TO
ABNORMAL EEG PATTERNS IN Infectious Diseases
INFANTS
Infectious diseases involving the CNS are fre-
Hypoxic-Ischemic Insult quently associated with abnormalities in the
This gives rise to severe EEG abnormalities EEG. The most characteristic finding is seen
and is the most common cause of neonatal in neonatal herpes simplex encephalitis and
seizures.2–9 consists of PLEDs.10 Other findings associated
with infectious processes include significant
asymmetries, interhemispheric asynchronies,
Intraventricular Hemorrhage multifocal sharp waves, and seizure discharges.
Intraventricular hemorrhage, which can occur Inborn Errors of Metabolism
in premature infants who have a hypoxic-
ischemic insult, is associated with rolandic Inborn errors of metabolism, biochemical dis-
positive sharp waves in the EEG.2–7 orders, and aminoacidurias can be associated
Figure 11–4. Focal spikes in an 8-month-old boy with focal right-sided motor seizures caused by tuberous sclerosis
complex.
with abnormal EEG patterns. Phenylketonuria location of the tubers2, 11, 12 (Fig. 11–4). Sturge–
previously was a common cause, but other Weber syndrome is associated with an asym-
types of aminoacidurias and inborn errors of metry of background activity and epileptiform
metabolism can present with neonatal seizures activity on the side of the facial nevus.2, 11, 13, 14
and epileptiform discharges.2–9
Cortical Malformations
Dysgenetic Disorders or Cortical malformations, particularly cortical

Neurocutaneous Disorders dysplasia, may be highly epileptogenic and
can result in focal or multifocal epileptiform
Tuberous sclerosis complex presenting in abnormalities with frequent interictal and ictal
infancy may be associated with a hypsarrhyth- discharges. At times almost continuous trains
mic pattern or focal or multifocal EEG abnor- of spike discharges or rhythmic epileptiform
malities that may or may not be related to the discharges may be present14–17 (Fig. 11–5).
Figure 11–5. Repetitive spike discharges over the left hemisphere in a 2-year-old boy with left hemisphere pachygyria.
Congenital Abnormalities During the first decade of life, there is consid-

erable variability among children of the same
These may be associated with various types age with regard to the amount of alpha, theta,
and combinations of abnormal patterns reflect- and delta activity present.2, 5, 18, 19
ing the abnormality of patients with congenital
abnormalities.2–8, 14
Key Points
Key Points • Developmental changes in infants and
• Many different conditions may be asso- children consist of evolution of rhyth-
ciated with abnormal EEG patterns mic activity over the central and posterior
in infants, including hypoxic insults, head regions.
• The central rhythm evolves to a mu
metabolic disorders, infectious diseases,
inborn errors of metabolism, dysgenesis, rhythm.
• The posterior rhythm evolves to alpha
and congenital anomalies.
rhythm.
• As the infant and child matures, the cen-
tral and posterior rhythms increase in fre-
DEVELOPMENTAL CHANGES quency with the adult frequency range
DURING INFANCY, being reached by 10–12 years of age.
CHILDHOOD, AND
ADOLESCENCE
Hyperventilation
In the first 3 months of life, a transition occurs
from the neonatal to the infant EEG pat- In younger children, the slowing produced by
tern.2, 6 The background consists of irregular hyperventilation is often maximal over the pos-
low-amplitude delta activity. terior head regions (Fig. 11–6), whereas in
Rhythmic activity in the range of 5–6 Hz is older children, the buildup response is usually
present in the central regions and is probably maximal over the anterior head regions.18, 19
a precursor of the mu rhythm. At 3 months of
age, rhythmic occipital activity in the range of
3–4 Hz is present. This activity can be atten-
uated with eye opening and represents a pre- Photic Responses
cursor of the alpha rhythm. Between 4 and 6
months of age, the central rhythm becomes Young children show responses in the occipital
better developed and shows a frequency of regions predominantly at slower flash frequen-
5–8 Hz; a better defined occipital rhythm in the cies. In older children, driving can be seen at
range of 5–6 Hz is present when the eyes are faster flash frequencies.
closed. Between 6 months and 2 years of age,
the central rhythm is well developed at a fre-
quency of 5–8 Hz. After 6 months, the occipital Drowsiness
rhythms become more prominent, and there is
a gradual shift to higher amplitude and faster Drowsiness in young children is character-
frequency activity, ranging from 6 to 8 Hz. In ized by high-amplitude sinusoidal 4–5 Hz theta
patients between 2 and 5 years of age, the activity that is maximal over the frontal, cen-
central and occipital rhythms are further differ- tral, and parietal regions. These slow waves
entiated. By 3 years of age, the occipital alpha initially occur in prolonged rhythmic trains
rhythms range from 6 to 8 Hz, and the ampli- and, in children between 1 and 9 years old, in
tude of this activity gradually increases.2, 5, 18, 19 bursts2, 5, 18, 19 (Fig. 11–7). Sometimes the slow
Between 6 and 16 years of age, there is a pro- waves may have a notched or sharp appearance
gressive increase in the alpha frequencies, and because of superimposed faster frequencies.19
the typical adult frequency range of 9–10 Hz is During adolescence, characteristic trains of
usually reached by 10–12 years of age. Some monorhythmic sinusoidal theta activity occur
interspersed theta activity may still be present, over the frontal regions and may precede the
predominantly over the anterior head regions. disappearance of the alpha rhythm.
Age: 6 yrs (3-12-75)

T3-FP1
T4-FP2
Fp1-F3
Fp2-F4
C3-P3
C4-P4
P3-O1
P4-O2
A) At rest B) After 1 min hyperventilation 100 μV
1 sec
Figure 11–6. Normal EEG from a 6-year-old boy, A, during resting wakefulness and, B, during hyperventilation.
Key Points slow-wave activity during quiet sleep, and the

percentage of time spent in non-rapid eye
• Drowsiness in children may be associated
movement (NREM) sleep is increased. Spin-
with a high amplitude 4–5 Hz rhythm over dles become apparent by 1–3 months of age
the frontal, central, and parietal regions. and are well developed and bisynchronous by
• In adolescents, the drowsy rhythm may
1–2 years of age. The spindles in the patient’s
consist of theta activity over the frontal first year of life may have a characteristic arci-
regions. form or comb-like appearance and occur in
prolonged and asynchronous trains. V waves
are apparent by 3–5 months of age. In chil-
Sleep dren from 2 to 5 years old, V waves have a
high amplitude and a sharp or spiky appear-
During the first few months of life, the tracé ance and occur in groups. O waves are large,
alternant pattern of newborns (see the section broad, bioccipital delta waves that are present
on Normal EEGs) is replaced by generalized over the occipital regions during drowsiness
1
Age: 3 2 yrs
FP1-F3
FP2-F4
F3-C3
F4-C4
C3-P3
C4-P4
C3-O1
P4-O2
100 μV
1 sec
Figure 11–7. Normal burst of rhythmic slow waves during drowsiness in a 3½-year-old boy.
and sleep.18 Occasionally they may have a O Waves or Cone-Shaped Waves

somewhat sharply contoured monophasic or
diphasic waveform.19 Occipital slow waves are In young children, high-voltage slow-wave
most prominent at 1–2 years of age but may transients, O waves, varying from a cone-
persist to at least 6 years of age. shaped appearance to diphasic slow-wave tran-
sients may be present over the occipital head
regions and interspersed with occipital domi-
Key Points nant slow delta waves of sleep.19 This activity
• During the first few months of life, the should not be mistaken for abnormal sharp-
tracé alternant pattern is replaced by a wave or slow-wave activity.
generalized slow-wave pattern.
• Spindles become apparent by 1–3 months
of age.
• Spindles may occur in prolonged and Posterior Slow Waves of Youth
asynchronous trains.
• V waves become apparent by 3–5 months Single delta frequency waves, referred to as
of age. posterior slow waves of youth or slow fused
• Occipital slow waves (O waves) are most transients, are common over the posterior head
prominent at 1–2 years of age but can regions in children and adolescents.18, 19 The
persist up to 5–6 years of age. posterior slow waves of youth occur sporadi-
cally rather than in consecutive trains, do not
protrude much above the average amplitude of
the alpha rhythm, and attenuate together with
BENIGN VARIANTS the alpha rhythm when the eyes are opened.
IN CHILDREN
Several benign variants can be seen in the
EEGs of children and adolescents. These 14 and 6 Hz Positive Spike Bursts
include the slow lambdas of youth, O waves,
posterior slow waves of youth, and 14 and 6 Hz The 14 and 6 Hz positive spike bursts con-
positive bursts. sist of brief bursts or trains of positive spikes
that occur during drowsiness at a rate of 14
and 6 Hz, mainly over the posterior tempo-
ral regions. They can be seen in children but
Posterior Slow Waves Associated are seen most often in adolescents5, 21 (see
with Eye Blinks or Slow Fig. 8–13).
Lambdas of Youth
A phenomenon similar to that of the lambda Key Points
waves is the posterior slow waves associated • Benign variants in children consist of
with eye blinks in some children (slow lambdas ◦ Posterior slow waves associated with
of childhood or shut-eye waves). They are sin- eye blinks
gle, broad, and monophasic or diphasic wave- ◦ O waves
forms that occur bilaterally over the occipital ◦ Posterior slow waves of youth
head regions after eye blinks or eye move- ◦ 14 and 6 Hz positive spike bursts
ments.20 The amplitude is often 100–200 μV,
and the duration is approximately 200–400 ms.
The predominant polarity is surface-negative.
The waveform may have a sharp-contoured ABNORMALITIES
appearance, but it should not be misinter-
preted as abnormal epileptiform activity. These Electroencephalograms in children can show
waveforms can be seen in children who are a wide variety of abnormalities. The most
between 6 months and 10 years old and are common are epileptiform abnormalities and
most prominent in those 2–3 years old. slowing.
Figure 11–8. Hypsarrhythmia in

a 9-month-old boy with infantile
spasms.
Epileptiform Abnormalities 2. 3-Hz generalized spike-and-wave—a pat-

tern usually seen in children between 3
Almost any type of epileptiform abnormality and 15 years old. It is associated with
can be seen in children. Some types of epilepti- absence seizures11, 21, 24 (Fig. 11–9).
form abnormalities and associated seizure dis- 3. Generalized slow spike-and-wave—a pat-
orders are unique or seen more commonly tern consisting of generalized sharp- and
in children.13, 14, 18, 22 The types of epilepti- slow-wave discharges seen in young chil-
form abnormalities most common in children dren with frequent seizures and men-
include the following: tal retardation and which constitutes
the Lennox–Gastaut syndrome11, 18, 21, 25
(Fig. 11–10).
1. Hypsarrhythmia—a pattern seen in chil- 4. Central-temporal spikes—a pattern seen
dren aged 4 months to 4 years. It in children 4–12 years old. The spikes
consists of high-amplitude multifocal have a blunt spike-and-wave appearance.
spikes, sharp waves, and slow waves. This pattern occurs in children with
This type of epileptiform activity is benign rolandic epilepsy of childhood
often seen in association with infantile (BREC)18, 22, 26 (Fig. 11–11). The above
spasms11, 14, 21, 23 (Fig. 11–8). The combina- epileptiform patterns are described more
tion of the hypsarrhythmia pattern and fully in Chapter 9.
infantile spasms is referred to as the West 5. Occipital spikes and seizures—a pattern
Syndrome. seen in children with occipital lesions
Age: 10 years
FP1-A1
FP2-A2
F3-A1
F4-A2
C3-A1
C4-A2
P3-A1
Figure 11–9. Absence seizure P4-A2

accompanied by typical paroxysm of Stops hyperventilating
3-Hz spike and slow-wave complexes is unresponsive
during hyperventilation in a 200 μV
has blinking of the eyelids 1 sec
10-year-old girl.
Age: 5 yr
T3-FP1
T4-FP2
FP1-F3
FP2-F4
C3-P3
C4-P4
Figure 11–10. Slow spike-and-wave P3-O1

pattern (sharp- and slow-wave com-
plexes) in a 5-year-old boy with seizures P4-O2
and mental retardation (Lennox–Gastaut Generalized seizures 100 μV
syndrome). mental and motor retardation 1 sec
because of birth injury, vascular lesions, have nocturnal seizures with head and
congenital malformations, cortical dysge- eye deviation, nausea, and vomiting.
nesis, Sturge–Weber syndrome, tuberous The seizures sometimes are followed by
sclerosis, tumors, or trauma.11, 13, 14, 18, 22, 27, 28 migraine headaches. The interictal EEG
Occipital spikes are also seen in the shows spike-and-wave discharges over the
benign epilepsies of children with occip- occipital head regions. These discharges
ital paroxysms27, 28 (Fig. 11–12). This occur in a unilateral, bilaterally inde-
is a seizure disorder associated with pendent, or bilaterally synchronous man-
elementary visual phenomena; it may ner, are attenuated with eye opening,
progress to secondarily generalized tonic– and reoccur with eye closure. Ictal dis-
clonic seizures. The child may also charges consist of low-voltage fast activity
Age: 7 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
F4-T6
F6-O2
Nocturnal seizure 50 μV
1 sec
Figure 11–11. Central-temporal spikes (maximal in C4 and T4) in a 7-year-old boy with a history of a single nocturnal
seizure.
Figure 11–12. Occipital spike discharges attenuated with eye opening in a 10-year-old boy with benign occipital seizures
of childhood.
over the occipital head regions, but it Slow-Wave Abnormalities

can spread more widely. The child often
outgrows the seizures and the spike dis- Slow-wave abnormalities may be focal or dif-
charges.21, 22, 27, 28 fuse and are often more prominent and take
longer to resolve in children than in adults.
Slow-wave abnormalities in children also tend
The Panayiotopoulos variant or syndrome to have maximal expression over the posterior
consists of benign focal seizures of childhood head region.11, 21
associated with vomiting and other autonomic Occipital intermittent rhythmic delta activ-
symptoms and multifocal spike discharges ity (OIRDA) is also seen in children. This
which are often maximal over the occipital or consists of intermittent rhythmic delta activ-
posterior head regions.27 ity over the posterior head region which can
Focal occipital spike discharges can also be be seen with a variety of conditions includ-
seen in young children with amblyopia, without ing epilepsy; metabolic, toxic, degenerative
associated seizures.21, 27, 28 disorders; encephalopathies; posterior fossa
tumors; and following head trauma.29

• Hypsarrhythmia occurs in infants with • Slow-wave abnormalities may consist of
infantile spasms (West Syndrome). focal or generalized slowing or OIRDA.
• 3-Hz generalized spike-and-wave is seen • The slowing is often more prominent than
in children with absence seizures. seen in adults.
• Generalized slow spike-and-wave is seen • The slowing is often maximal over the
in children with frequent seizures and posterior head regions.
mental retardation (Lennox–Gastaut syn-
drome).
• Central-temporal spikes is seen in chil-
dren with benign rolandic epilepsy of Conditions Giving Rise to
childhood. Abnormal EEG Patterns
• Occipital spikes can be seen in infants in Children
and children with occipital lesions, with
benign epilepsy of childhood, or young Electroencephalographic abnormalities can be
children with amblyopia. seen in many disorders, including degenerative
disorders, inflammatory diseases, tumors, vas- in the EEG of patients following the onset of
cular insults, head trauma, and various types seizures.2, 11, 13, 14, 21
of encephalopathies. Transient abnormalities
also occur with migraine headaches and pos- White Matter Disease
tictal states. White matter disease, including the various
leukodystrophies, is associated with the loss
of background activity and moderate- to high-
DEGENERATIVE DISORDERS
amplitude delta slowing, which is often max-
Degenerative disorders, various aminoacidurias, imal over the posterior head regions2, 11, 18, 21
and inborn errors of metabolism may be asso- (Fig. 11–13).
ciated with slowing and multifocal epileptiform
Key Points
abnormalities in the EEG, particularly if the
child has seizures.2–11, 13 • Gray matter disorders are often associated
The type of degenerative process and with epileptiform activity.
whether it involves the white or gray matter • White matter disorders are associated
influences the EEG pattern. Processes that with delta slowing.
affect predominantly the white matter usually
cause polymorphic delta slowing in the EEG.
Epileptiform abnormalities are more common INFLAMMATORY DISORDERS
in gray matter disease but also can occur in Meningitis
white matter disease. Other factors influenc-
ing the degree and type of EEG abnormalities This is associated with differing degrees of
include the age of onset of the disease pro- slowing in the EEG depending on the type
cess, the age and state of the patient at the of meningitis and the degree of involvement
time of the EEG, the stage of the disease proof the CNS. Purulent meningitis is often
cess, and other complicating factors, including associated with moderate-to-severe general-
infectious, metabolic, and drug effects. ized slow-wave abnormalities, and epileptiform
discharges may be present in patients who have
seizures.18, 30
Gray Matter Disease
Encephalitis
Gray matter disease such as the progressive
myoclonic epilepsies is associated with gener- The EEG abnormalities in encephalitis often
alized epileptiform abnormalities and slowing are more severe than those in meningitis, with
Age: 3 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
PP2-F4
P4-C4
C4-P4
P4-O2
150 μV
1 sec
Figure 11–13. Delta slowing over the posterior head regions in a 3-year-old boy with metachromatic leukodystrophy.
the EEG showing high-voltage arrhythmic or Brain Abscess

rhythmic delta slowing (see Fig. 10–6). Focal
Focal polymorphic delta slowing is often
or multifocal epileptiform abnormalities may
present over the site of the abscess if the lesion
occur in patients who have seizures.30
is located close to the surface of the brain
Herpes simplex encephalitis can occur in
(Fig. 11–14). Although epileptiform abnormal-
infants and children. As in adults, PLEDs are
ities are not usually seen in acute stages, they
a prominent feature in the EEG.10 After res-
may develop in the later stages of resolution of
olution of the infection, the EEG may show
the abscess.30
localized areas of attenuation overlying cystic
areas of the brain or multifocal epileptiform
abnormalities or both. Key Points
Rasmussen encephalitis—a syndrome of • Meningitis and encephalitis are associated
chronic smoldering encephalitis, is character- with generalized slowing in the EEG.
ized by progressive neurologic and intellec- • Focal epileptiform abnormalities may be
tual deterioration and recurrent seizures.31, 32 present in meningitis and encephalitis.
The patients have varying types of seizures, • Herpes simplex encephalitis is associated
which may progress to epilepsia partialis with PLEDs.
continua. The EEG shows various types of • Rasmussen encephalitis is associated with
epileptiform and slow-wave abnormalities that lateralized epileptiform abnormalities and
occur in different locations during the dif- slowing.
ferent stages of the disease process. Often, • SSPE is associated with periodic sharp-
Rasmussen encephalitis initially involves one and slow-wave complexes.
hemisphere and then may spread more • Brain abscesses are often associated with
widely. focal delta slowing, and epileptiform
Subacute sclerosing panencephalitis abnormalities may develop at a later stage.
(SSPE)—a slow virus disorder believed to
be caused by the measles virus that occurs
SUBDURAL EFFUSION AND
in children and adolescents, is associated
EMPYEMA
with repetitive stereotyped high-voltage sharp-
and slow-wave complexes that recur every Subdural effusions, hygromas, and empyemas
4 or 5 seconds and are associated with act like subdural hematomas and cause an
stereotyped motor jerks or spasms2, 11, 21, 30 (see attenuation or decreased amplitude of activity
Fig. 10–8). and slowing over the involved hemisphere.30
Age: 8 yrs
FP1-F7
F7-T3
T3-T5
T5-O1
FP2-F8
F8-T4
T4-T6
T6-O2
Congenital cardiac valve defect
100 μV
1 sec
Figure 11–14. Focal delta slowing over the right frontal region in an 8-year-old boy with a right frontal abscess.
HEMICONVULSIONS, VASCULAR LESIONS

HEMIPLEGIA, AND EPILEPSY
Vascular insults, including infarcts and hemor-
In the hemiconvulsions, hemiplegia, and epi- rhage, are less common in children than adults,
lepsy syndrome (HHE), the infant or child but when present, they are associated with
has a series of seizures or hemiconvulsive focal slowing and loss of background activity.
status epilepticus during an acute febrile ill- PLEDs may be seen in the acute stage and
ness. Following the acute episode of seizures focal epileptiform abnormalities in the chronic
the child has hemiparesis, and later, chronic stage.11, 21
epilepsy develops. In the acute stage of the syn-
drome, the EEG shows frequent or continuous
spike discharges or spike-and-wave discharges Key Points
over the involved side. After the acute stage, • The EEG in vascular lesions may show
there is persistent attenuation of the EEG
slowing, loss of background activity, and
activity and slowing over the affected side.
epileptiform discharges.
The child usually develops epilepsy after the
acute event with focal, unilateral, or multifocal
epileptiform discharges being present over the
affected side. HEAD TRAUMA
The EEG may show focal, lateralized, or gener-
alized slow-wave abnormalities or an asymme-
Key Points
try of activity (or both slow-wave abnormalities
• In HHE, the child has hemiconvulsive and asymmetry) depending on the extent of the
status during a febrile illness. head injury (Fig. 11–15). The slow-wave abnor-
• The EEG shows frequent or continu- malities often are maximal over the posterior
ous epileptiform activity over the involved head regions. A moderate degree of slowing
hemisphere during the acute seizures. is not uncommon after relatively minor head
• Later epileptiform activity, attenuation of injury, and slowing may be out of proportion to
activity, and slowing are seen over the the degree of head injury.11, 21
affected side.
Key Points
TUMORS • The EEG in head trauma can show
The EEGs of children with tumors show a focal, lateralized, or generalized slow-
greater predominance of slow-wave abnormal- wave abnormalities or asymmetry of
ities over the posterior head regions than those activity.
of adults with tumors. This may partly reflect • The slowing may be out of proportion to
two things: (1) children have a greater inci- the degree of head trauma.
dence of posterior fossa tumors than adults and
(2) the predominance of slow-wave abnormali-
ties over the posterior head region in children HYDROCEPHALUS
is an age-related phenomenon. In supratento-
rial tumors, the EEG shows focal or lateral- The EEG abnormalities in patients with
ized slow-wave abnormalities, asymmetry, or hydrocephalus may consist of focal or gen-
epileptiform activity over the involved area.11 eralized slow-wave abnormalities, epileptiform
abnormalities, asynchronous sleep activity, or
asymmetry of the background activity. There
Key Points may be an increase in the slow-wave and
epileptiform abnormalities with obstructive
• Tumors are associated with focal slow- hydrocephalus because of malfunction of the
wave abnormalities. shunt (Fig. 11–16). The incidence of EEG
• Because of the frequency of posterior abnormalities is higher in children who have
fossa tumors, the slowing may predomi- had a ventricular shunt, and focal abnormalities
nate over the posterior head regions. are often present in the area of the shunt.
Age: 10 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
Fell from a horse 5 days ago 150 μV

Stuporous 1 sec
Figure 11–15. Moderate diffuse slow-wave abnormalities after head trauma in a 10-year-old girl.
Key Points frequent spikes or sharp-wave discharges

occurring focally or bilaterally and which are
• Hydrocephalus may be associated with
often maximal over the temporal or parietal
focal or generalized slow-wave abnormal- occipital head regions (Fig. 11–17). There is a
ities, asymmetry, and epileptiform dis- marked increase in the spike discharges dur-
charges. ing sleep. The discharges can become more
• Focal EEG abnormalities may be present
widespread and continuous during slow-wave
in the area of the shunt. sleep.2, 5, 13, 14, 33
LANDAU–KLEFFNER SYNDROME
Key Points
This is a disorder seen in children who develop
acquired aphasia with loss of spontaneous • Landau–Kleffner syndrome consists of
speech, seizures, and epileptiform abnormal- acquired aphasia and frequent epilepti-
ities on the EEG. The EEG typically shows form abnormalities on the EEG.
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
50 μV
1 sec
Figure 11–16. Intermittent rhythmic slow-wave abnormalities in a 12-year-old child with obstructive hydrocephalus with
a blocked shunt.
Figure 11–17. Left temporal spike discharges in a 5-year-old child with Landau–Kleffner syndrome.
Transient Abnormalities POSTICTAL SLOWING

Postictal slowing and attenuation of the back-
MIGRAINE HEADACHE
ground may be more prominent and persist
In between migrainous attacks, the EEG may longer in children than in adults.
show nonspecific abnormalities. During and
after a migraine episode, asymmetry and focal, Key Points
lateralized, or generalized slowing may be • The EEG in transient disorders such as
present. Slow-wave abnormalities are usually
migraine headache and in postictal states
more prominent and widespread in children
can show focal, lateralized, or generalized
than in adults11, 21 (Fig. 11–18). The abnor-
slowing and asymmetry of activity.
malities may persist longer in children than • Postictal slowing may persist longer and
in adults and may take up to a week to
be more prominent in children than in
resolve.
adults.
Age: 10 yrs
FP1-F3
F3-C3
C3-P3
P3-O1
FP2-F4
F4-C4
C4-P4
P4-O2
Headache
Left Hemiplegia 50 μV
1 sec
Figure 11–18. Prominent delta slowing over the right hemisphere during a hemiplegic migrainous episode in a 10-year-
old girl.
SUMMARY ed. M. R. Gómez, 3rd ed., 63–74. New York: Oxford

University Press.
This chapter discusses the normal and abnor- 13. Pellock, J. M., W. E. Dodson, and B. F. D. Bourgeois,
eds. 2001. Pediatric epilepsy. Diagnosis and therapy,
mal EEG patterns in neonates, infants, and 2nd ed. New York: Demos.
children. Because the EEG is a neurophysio- 14. Wyllie, E., ed. 2006. The treatment of epilepsy: Prin-
logic test that reflects a disturbance of cerebral ciples & practice, 4th ed. Lippincott Williams &
function, it is an important tool in evaluating Wilkins.
15. Kuzniecky, R. I. 2006. Malformations of cortical devel-
infants and children. The EEG helps deter- opment and epilepsy, part 1: Diagnosis and classi-
mine whether there is an underlying organic fication scheme. Reviews in Neurological Diseases
disorder, indicates the degree and extent of 3(4):151–62.
cerebral dysfunction, confirms or helps make 16. Palmini, A. N. I., G. Avanzini, T. Babb, et al. 2004. Ter-
the diagnosis of a specific disorder such as minology and classification of the cortical dysplasias.
Neurology 62:S2–8.
seizures, indicates whether there is a focal or 17. Raymond, A. A., and D. R. Fish. 1996. EEG
generalized process, and, in certain instances, features of focal malformations of cortical devel-
aids in determining the prognosis. opment. Journal of Clinical Neurophysiology 13:
495–506.
18. Blume, W. T., and M. Kaibara, eds. 1999. Atlas of pedi-
atric electroencephalography. 2nd ed. Philadelphia:
REFERENCES Lippincott-Raven.
19. Kellaway, P. 2003. Orderly approach to visual analysis:
1. American Electroencephalographic Society. 2006. Characteristics of the normal EEG of adults and
Guideline two: Minimum technical standards for children. In Current practice of clinical electroen-
pediatric electroencephalography. Journal of Clinical cephalography, ed. J. S. Ebersole, and T. A. Pedley,
Neurophysiology 23:92–6. 3rd ed., 100–59. Philadelphia: Lippincott Williams &
2. Hahn, J. S., and B. R. Tharp. 2005. Neonatal and pedi- Wilkins.
atric electroencephalography. In Electrodiagnosis in 20. Westmoreland, B. F., and D. W. Klass. 1990. Unusual
clinical neurology, ed. M. J. Aminoff, 5th ed., 85–129. EEG patterns. Journal of Clinical Neurophysiology
New York: Elsevier Churchill Livingstone. 7:209–28.
3. Mizrahi, E. 2004. Atlas of neonatal electroencephalog- 21. Ebersole, J. S., and T. A. Pedley, eds. 2003.
raphy. Philadelphia: Lippincott Williams & Wilkins. Current practice of clinical electroencephalo-
4. Scher, M. S. 2005. Electroencephalography of the graphy, 3rd ed. Philadelphia: Lippincott Williams &
newborn: Normal and abnormal features. In Elec- Wilkins.
troencephalography: Basic principles, clinical appli- 22. Loiseau, P. 2001. Benign focal epilepsies of child-
cations, and related fields, ed. E. Niedermeyer, and hood. In The treatment of epilepsy: Principles and
F. Lopes da Silva, 5th ed., 937–90. Philadelphia: practice, ed. E. Wyllie, 3rd ed., 475–84. Philadelphia:
Lippincott Williams & Wilkins. Lippincott Williams & Wilkins.
5. Holmes, G. L., S. L. Moshé, and H. R. Jones. 23. Frost, J. D. Jr., and R. A. Hrachovy. 2003. Infantile
2006. Clinical neurophysiology of infancy, childhood, spasms. Boston: Kluwer Academic Publishers.
and adolescence, 3–129. Philadelphia: Butterworth- 24. Hrachovy, R. A., and J. D. Frost Jr. 2006. The EEG
Heinemann/Elsevier. in selected generalized seizures. Journal of Clinical
6. Clancy, R. R., A. G. C. Bergqvist, and D. J. Dlugos. Neurophysiology 23:312–32.
2003. Neonatal electroencephalography. In Cur- 25. Markand, O. N. 2003. Lennox–Gastaut syndrome
rent practice of clinical electroencephalography, ed. childhood epileptic encephalopathy. Journal of Clin-
J. S. Ebersole, and T. A. Pedley, 3rd ed., 160–234. ical Neurophysiology 20:426–41.
Philadelphia: Lippincott Williams & Wilkins. 26. Kellaway, P. 2000. The electroencephalographic fea-
7. Shewmon, D. A. 2000. Coma prognosis in children. tures of benign centrotemporal (rolandic) epilepsy of
Part II: Clinical application. Journal of Clinical Neuro- childhood. Epilepsia 41:1053–6.
physiology 17:467–72. 27. Panayiotopoulos, C. P. 2005. The epilepsies: Seizures,
8. Mizrahi, E. M., and P. Kellaway. 1998. Diagnosis syndromes and management. Oxfordshire, UK: Bladon
and management of neonatal seizures. Philadelphia: Medical Publishing.
Lippincott-Raven. 28. Kivity, S., T. Ephraim, R. Weitz, and A. Tamir.
9. Tharp, B. R. 2002. Neonatal seizures and syndromes. 2000. Childhood epilepsy with occipital paroxysms:
Epilepsia 43:2–10. Clinical variants in 134 patients. Epilepsia 41:
10. Mizrahi, E. M., and B. R. Tharp. 1982. A characteristic 1522–33.
EEG pattern in neonatal herpes simplex encephalitis. 29. Watemberg, N., I. Linder, R. Dabby, L. Blumkin, and
Neurology 32:1215–20. T. Lerman-Sagie. 2007. Clinical correlates of occip-
11. Niedermeyer, E., and F. Lopes da Silva, eds. ital intermittent rhythmic delta activity (OIRDA) in
2005. Electroencephalography: Basic principles, clin- children. Epilepsia 48(2):330–4.
ical applications, and related fields, 5th ed. Philadel- 30. Westmoreland, B. F. 2005. The EEG in cere-
phia: Lippincott Williams & Wilkins. bral inflammatory processes. In Electroencephalog-
12. Westmoreland, B. F. 1999. The electroencephalogram raphy: Basic principles, clinical applications, and
in tuberous sclerosis. In Tuberous sclerosis complex, related fields, ed. E. Niedermeyer, and F. Lopes
da Silva, 5th ed., 323–37. Baltimore: Williams & ciples & practice, editor-in-chief E. Wyllie, 4th ed.,
Wilkins. 441–54. Philadelphia: Lippincott Williams & Wilkins.
31. Andermann, F. ed. 1991. Chronic encephalitis 33. Neville, B., and J. H. Cross. 2006. Continuous spike
and epilepsy: Rasmussen’s syndrome. Boston: wave of slow sleep and Landau–Kleffner syndrome.
Butterworth-Heinemann. In The treatment of epilepsy: Principles & practice,
32. Dubeau, F. 2006. Rasmussen’s encephalitis (chronic editor-in-chief E. Wyllie, 4th ed., 455–62. Philadel-
focal encephalitis). In The treatment of epilepsy: Prin- phia: Lippincott Williams & Wilkins.
Chapter 12
Ambulatory
Electroencephalography
Jeffrey R. Buchhalter
INTRODUCTION CLINICAL APPLICATIONS

INDICATIONS SUMMARY
TECHNOLOGY
INTRODUCTION Purpose and Role of AEEG

Ambulatory electroencephalography (AEEG) • AEEG provides a means of recording
is a clinical neurophysiology tool that is most brain activity in a continuous fashion for a
useful when a diagnosis cannot be made on prolonged period of time in the outpatient
the basis of the clinical history, neurologic setting in a cost-efficient manner.
examination, and routine outpatient awake and • AEEG can define the etiology of an
sleep electroencephalographic (EEG) studies. epileptic event, seizure classification, and
Recently, AEEG has been used to define localization.
the frequency of epileptiform discharges in a • AEEG can have a high yield if patients are
nonepileptic population.1 The advantages of carefully selected for recording.
AEEG include the ability to perform pro-
longed recording in the patient’s natural envi-
ronment and to record multiple cycles of sleep
that have not been induced with medication. INDICATIONS
Also, the cost saving compared with the cost
of inpatient evaluation is substantial. The util- Typically, the questions being asked relate to
ity of the specific type of study performed will the potential epileptic etiology of an event
be determined by the question being asked.2 (spell classification) in a patient with or with-
The technology has evolved from cassette tapes out known seizures, the number of seizures
with limited memory capacity to digital sys- that are occurring in the awake and sleep states
tems with remontaging capability, spike and (seizure quantification), determination of par-
seizure detection and extracerebral channels tial vs. generalized onset (seizure classifica-
for recording heart rate, eye movements, oxy- tion), and determining the precise region of
gen saturation, respiratory excursion and video. seizure onset (ictal localization). The likelihood
187
of recording useful information depends on TECHNOLOGY

the clinical scenario that prompted the routine
EEG and the frequency of the event. The fol- Although single-channel ambulatory cardiac
lowing are situations in which AEEG may be monitoring was reported in 1949, it was not
indicated: until the 1970s that the miniaturization of
cassette recording allowed AEEG to become
a reality. Devices capable of recording four
Clinical Scenario 1: Parents report that their
channels of physiologic data were modified to
9-month-old child intermittently stiffens
record EEGs. The addition of preamplifiers
and stops breathing for 10 seconds within
reduced the noise level from 30 to 50 μV to
2 or 3 hours after the onset of sleep each
5 μV, so that EEG background activity and
night. Routine EEG with sleep recording
higher amplitude epileptiform transients could
for 15 minutes is normal.
be distinguished from background noise.3 The
Clinical Scenario 2: Teachers report that a
problem of prolonged playback times caused
6-year-old boy stares one or two times a
by the limited response time of mechanical
day during school. A routine awake and
pens was solved by adaptation of an inkjet
sleep EEG is normal.
printer that allowed playback of a 24-hour
Clinical Scenario 3: A 27-year-old woman
recording in 24 minutes. Thus, by 1980, it was
describes a numb feeling in her left arm
possible to record four channels for 24 hours
and states that she “blanks out” three or
on a lightweight, analog cassette recorder. This
four times a week. A routine awake and
system was used to quantify 3-Hz spike-wave
sleep EEG is normal.
discharges in children with childhood absence
epilepsy, to lateralize EEG abnormalities, and
Clinical Scenario 1 poses the problem of to distinguish pseudoseizures from epileptic
an infant who may have nocturnal seizures, seizures.
intermittent aspiration, or respiratory obstruc- However, substantial technical problems
tive disease. AEEG could clarify the event included limited spatial resolution and diffi-
by demonstrating epileptiform abnormalities culty with artifact detection because of the lim-
associated with the event or by recording the ited number of channels, inability to reformat
lack of chest wall movement and nasal air- montages, and inferior recording characteris-
flow, with the use of additional polysomno- tics when compared with those of standard
gram channels, or by showing no abnormalities hardwired systems. To estimate the effect of
at all. limited channels, a series of rigorous studies
Clinical Scenario 2 describes the frequent were performed in epilepsy patients undergo-
problem of the “staring child.” The differential ing standard inpatient EEG monitoring. The
diagnoses that could be refined by recording distribution of epileptiform discharges was
the event with AEEG in the classroom setting determined,4 the montage of the 4-channel
include normal daydreaming, complex partial AEEG was accordingly optimized,5 and the
seizures, and absence seizures. If the event is detection of epileptiform abnormalities by the
a seizure, precise quantitation is possible, as two techniques was compared.6 It was found
is the detection of subclinical (electrographic) that even a limited number of recording chan-
seizures. nels could detect approximately 75% of the
Clinical Scenario 3 could be caused by abnormalities. The subsequent introduction of
a complex partial seizure, a pseudoseizure, an 8-channel system markedly improved the
or cardiovascular or migraine phenomena. ability to identify artifacts and improved local-
Recording the event with AEEG could demon- ization.7 This was enhanced by the introduction
strate epileptiform abnormalities consistent of computerized, off-line analysis of the ana-
with a seizure or suggest vascular insufficiency log data. These algorithms were demonstrated
as indicated by focal or generalized slowing. to provide excellent detection of epileptiform
If each event has onset in a discrete brain rhythmic activity.8
region (e.g., a temporal lobe), then this indi- Analog recording on cassettes was replaced
vidual could be considered for further surgi- by the application of digital, computer-based
cal evaluation if the seizures are refractory to technology that has rendered the record-
antiepileptic medications. ing quality of AEEG comparable to that of
Ambulatory Electroencephalography 189
in-laboratory EEG. Since the initial report external environment in a manner not feasible
of this type of system,9 an impressive array with AEEG. Also, the reduction or discontin-
of commercially available products with var- uation of antiepileptic medication to facilitate
ious modifications have become available. seizure recording in the inpatient setting would
All systems use digital recording that allows not be safe in the home.
computer-based reformatting of the montage
at the time of review. Basic recording fea- Key Points
tures vary slightly between systems with regard
to the number of recording channels (16–32), • Basic recording features of AEEG equip-
sampling rate (1–500 Hz), analog-to-digital ment vary with regard to the number of
conversion (12–22 bit), frequency bandwidth recording channels (16–32), sampling rate
(DC–70 Hz), and common mode rejection of (1–500 Hz), analog-to-digital conversion
greater than 100 dB. Continuous recording can (12–22 bit), frequency bandwidth (DC–
be performed from 24 to 60 hours depend- 70 Hz), and common mode rejection of
ing on the battery life and data compression greater than 100 dB.
techniques used. Seventy-two hours of record- • Despite the significant technical advances,
ing is possible if events or samples are stored several limitations of AEEG remain, such
in an intermittent rather than a continuous as electrode stability, which cannot be
mode. Most systems incorporate epileptiform assured in the home setting as it can be
transient and seizure detection algorithms. in the EEG laboratory or epilepsy moni-
Systems vary significantly with regard to the toring unit.
availability of pulse oximetry, specific chan-
nels for polysomnography, and simultaneous
video recording. Clinical events are recorded CLINICAL APPLICATIONS
in an event calendar and indicated by push-
ing an event marker button. One system pro- There are two fundamental types of clinical
vides an audio channel for the patient or applications of AEEG. The first type is deter-
attendant to indicate the time and nature of mining whether an event is of epileptic etiology
the event. The data are downloaded from the (including persons with known seizures). Com-
AEEG storage medium to a standard desktop mon examples include loss or impairment of
personal computer (PC) on which proprietary consciousness, behavioral disturbances, motor
software allows review with montage reformat- phenomena, and sensory experiences.11 The
ting, ability to alter the filter and sensitivity second type is in persons with known seizure
settings, and analysis of detected events. disorders in which the AEEG may clarify par-
Simultaneous video to accompany the AEEG tial vs. generalized onset, quantification of elec-
recording is not available on the majority of trographic seizures, and localization of ictal
most commercially available systems. How- onset for possible epilepsy surgery. The utility
ever, one manufacturer (Digitrace) has mar- of AEEG is best demonstrated when compari-
keted a system with video for several years son can be made with the same population pre-
and another video option has recently become viously (or simultaneously) studied with either
available (Lifelines Neurodiagnostic Systems, routine awake and sleep EEG or inpatient,
Inc.). In order for the video component to video-EEG recording. The following review
be useful, the patient must remain within the highlights the applications noted above.
relatively small field of view of the portable An early study with the 4-channel cassette
camera. In an authoritative review of long- recorder of 100 children and adults with tem-
term monitoring in epilepsy, it was noted that poral lobe epilepsy indicated that AEEG was
no technology assessment exists for the use of at least three times more effective in record-
video with AEEG.10 ing a seizure than routine EEG.12 Laterality,
Despite the significant technical advances, but not precise localization of the seizure focus,
several limitations of AEEG remain. Electrode could be determined in many of the patients.
stability cannot be assured in the home setting Applications of AEEG in the pediatric pop-
as it can be in the EEG laboratory or epilepsy ulation have included distinguishing epileptic
monitoring unit. A trained EEG technologist from nonepileptic spells, predicting outcome
or nurse can test the patient’s response to the after neonatal seizures, quantifying absence
seizures, and characterizing infantile spasms. clarify the epileptic etiology of an event. This
An 8-channel system was used to study 95 question was approached in a large study of
infants and children who had clinically likely 2221 patients (most of them adults).18 AEEG
(n = 40) or unlikely (n = 55) seizures.13 In the recorded typical clinical events in approxi-
known seizure group, recorded seizures aided mately one-third of the study population. Of
in the quantitation of seizure frequency. Ambu- these patients, one-third had EEG findings
latory recordings captured events in 24 of the that correlated with the ictus. The record-
suspected pseudoseizure group and demon- ing duration ranged from 1 to 8 days, with
strated no ictal EEG changes. A study of 90% of patients having an event within 2 days.
infants who had neonatal seizures demon- This is the only study that provides informa-
strated that a combination of findings on the tion about the likely yield of AEEG mon-
routine EEG performed at the time of the itoring with time, although interpretation is
seizure followed by subsequent AEEG pre- limited because of the lack of clinical informa-
dicted the risk of seizure recurrence at 4 tion about the patients, thereby limiting the
months of age in 92% of infants.14 This study ability to predict who may benefit from more
showed how routine EEG and AEEG might prolonged study. A related question is how
complement each other. the frequency of events relates to the like-
To compare the utility of recording tech- lihood of recording one using AEEG. Some
niques in typical absence epilepsy, 25 chil- insight is gained by a study of 157 children
dren with this epilepsy syndrome had a routine who reported having discrete events at least 3
awake EEG, followed by an 8-hour AEEG in days per week.19 In this population, 89% of the
the awake state.15 These studies were repeated patients had typical events during 1–4 days of
1 month after the initiation of treatment. study of which 76% were nonepileptic.
Although the initial prolonged study did not Relatively few recent investigations have
add to diagnostic accuracy, at 1 month, only directly compared routine EEG and AEEG.
four children had spike-wave abnormalities on In a study of 344 patients, 16-channel,
the routine EEG and 10 children had epilep- computer-assisted AEEG proved “clinically
tiform discharges on the AEEG despite having useful” (defined as detecting an epileptiform
valproic acid levels in the therapeutic range. It abnormality or recording clinical event that did
appears that in this syndrome AEEG is very not have an epileptiform EEG correlate) in
useful when the clinician and parents believe 74% of the study population.20 A similar pro-
that seizures have been eliminated. portion (67%) of AEEG studies was useful in
AEEG has been shown to provide diag- 191 patients who had normal or nonspecific
nostic and prognostic information in children routine EEG recordings. A subsequent investi-
with infantile spasms. This severe symptomatic gation revealed that management was affected
epilepsy syndrome of childhood is character- directly by the results of the 16-channel AEEG
ized by frequent clinical and electrographic recording in approximately 80% of the patients
seizures. In a cohort of 74 infants with infan- referred.21
tile spasms, AEEG detected partial seizures The utility of AEEG in quantifying seizures
in 51% in addition to the generalized spasms, was demonstrated by a study that analyzed 595
and the partial seizures were associated with an recordings of which 47 demonstrated partial
unfavorable outcome.16 This study contributed seizures.22 Sixty-two percent of the seizures
to the recognition that a generalized seizure were noted by the patients as indicated by
disorder can be associated with focal seizures push buttons and/or diary entries. However,
that indicate a relatively poor outcome. More 11 seizures (23%) were not noted or recorded
recently, the utility of the computerized, 16- by the patient. These findings provide a clear
channel, digital AEEG with a seizure detec- indication of how frequently patients underes-
tion algorithm was demonstrated in a pop- timate seizure occurrence. The implication is
ulation of children to differentiate epileptic that if the clinical suspicion is high, the clini-
from nonepileptic events.17 The seizure was cian may wish to request an AEEG even if the
detected by the computer only and not by reported event frequency may appear to make
clinical observation in 5 of 26 recordings. the likelihood of recording an event low.
One issue that remains unclear is how long A potential application of AEEG is the
ambulatory monitoring should be performed to detection of abnormalities that occur during
Ambulatory Electroencephalography 191
sleep. A recent study compared the detec- Key Points

tion of abnormalities in routine sleep-deprived • AEEG is an effective tool for determining
EEGs vs. AEEG in patients with normal or
the possible epileptic etiology of events
nonspecific awake recordings.23 Approximately
involving alterations in conscious, behav-
the same percentage of interictal epilepti-
ioral, sensory, and motor activity when suf-
form transients was found by both tech-
ficient information has not been obtained
niques during sleep. However, whereas the
from routine awake and sleep EEG.
sleep-deprived EEG recorded no seizures, the • AEEG is also useful in the classification,
AEEG recorded seizures in 15% of patients.
quantification, and localization of seizures.
Furthermore, for approximately half of the • Technical barriers to high-quality recording
seizures, patients did not indicate aware-
have been eliminated by computer-assisted
ness of their seizures by activating a push
digital systems with seizure detection
button. This study highlights the value of
algorithms.
prolonged recording and seizure detection • Out-of-hospital AEEG should not be used
algorithms.
when the dose of medication is decreased
The role of AEEG in the selection of candi-
or intracranial electrode implantation is
dates for epilepsy surgery is evolving. As noted
required.
above, the early 4-channel system did not • AEEG is likely to provide substantial cost
provide the necessary localizing information.6
savings compared with the cost of misdiag-
Similar results were found in a study in which
nosis of nonepileptic events, inappropriate
an 8-channel AEEG recording with sphenoidal
medication use, physician utilization, and
electrodes was compared directly with a stan-
inpatient EEG evaluation, but these have
dard 16-channel recording in an inpatient pop-
not been evaluated formally.
ulation undergoing study for epilepsy surgery.24
The AEEG system was comparable to the stan-
dard system for hemispheric lateralization, but
it was not as reliable for precise localization of
seizure onset. However, it has been suggested SUMMARY
that the use of AEEG systems with at least 16
channels in a carefully selected population of Ambulatory EEG (AEEG) can provide a cost-
patients may provide the necessary information efficient means of recording prolonged, con-
to proceed directly to epilepsy surgery with- tinuous EEG in the outpatient setting. The
out inpatient electrophysiologic evaluation.25 quality of digital systems is comparable to
A report of seven patients who had nonle- in-lab recording—number of available chan-
sional temporal lobe epilepsy, resection based nels, sampling rate, frequency bandwidth,
upon only AEEG recording and excellent out- sharp transient/seizure detection algorithms,
come is supportive of this approach.26 The and the ability to reformat studies. AEEG is
authors were careful to note the small num- an effective technology to distinguish epilep-
ber of patients involved in this retrospective tic from non-epileptic events in addition to
study and that AEEG cannot be used in cir- classification, quantification, and localization
cumstances that would compromise the safety of seizures. Limitations of the technique are
of the patient, for example, when the dose of the inability to immediately secure loose elec-
an antiepileptic drug is decreased to facilitate trodes and to decrease antiepileptic drugs
the recording of seizures or when intracranial without direct observation by EEG technolo-
electrodes are required. gists and nursing personnel.
In the future, the technology developed for
AEEG may be utilized for seizure interven-
tion.27 This presumes a “closed loop” system
in which the detection or anticipation of an REFERENCES
ictal onset in the near or immediate future will
lead to an intervention such as direct stimula- 1. Schachter, S. C., M. Ito, and B. B. Wannamaker,
et al. 1998. Incidence of spikes and paroxysmal rhyth-
tion of the brain to depolarize the susceptible mic events in overnight ambulatory computer-assisted
cortex or local infusion of an immediate acting EEGs of normal subjects: A multicenter study. Journal
antiepileptic compound. of Clinical Neurophysiology 15:251–5.
2. Schomer, D. L. 2006. Ambulatory EEG telemetry: in typical absence seizures. Brain & Development
How good is it? Journal of Clinical Neurophysiology 13:223–7.
23:294–305. 16. Plouin, P., O. Dulac, C. Jalin, and C. Chiron. 1993.
3. Ives, J. R. 1975. 4-Channel 24 hour cassette recorder Twenty-four-hour ambulatory EEG monitoring in
for long-term EEG monitoring of ambulatory patients. infantile spasms. Epilepsia 34:686–91.
Electroencephalography and Clinical Neurophysiol- 17. Foley, C. M., D. K. Miles, A. Legido, and W. Grover.
ogy 39:88–92. 1993. Diagnostic value of computerized outpatient
4. Leroy, R. F. and J. S. Ebersole. 1983. An evaluation long-term EEG monitoring in children and adoles-
of ambulatory, cassette EEG monitoring: I. Montage cents (abstract). Epilepsia 34(Suppl. 6):139.
design. Neurology 33:1–7. 18. Tuunainen, A., U. Nousiainen, E. Mervaala, and
5. Ebersole, J. S. and R. F. Leroy. 1983. An evaluation of P. Riekkinen. 1990. Efficacy of a 1- to 3-day
ambulatory, cassette EEG monitoring: II. Detection of ambulatory electroencephalogram in record-
interictal abnormalities. Neurology 33:8–18. ing epileptic seizures. Archives of Neurology 47:
6. Ebersole, J. S., and R. F. Leroy. 1983. Evaluation 799–800.
of ambulatory cassette EEG monitoring: III. Diag- 19. Olson, D. M. 2001. Success of ambulatory EEG
nostic accuracy compared to intensive inpatient EEG in children. Journal of Clinical Neurophysiology
monitoring. Neurology 33:853–60. 18:158–61.
7. Ebersole, J. S., and S. L. Bridgers. 1985. Direct 20. Morris, G. L. III, J. Galezowska, R. Leroy, and R.
comparison of 3- and 8-channel ambulatory cassette North. 1994. The results of computer-assisted ambu-
EEG with intensive inpatient monitoring. Neurology latory 16-channel EEG. Electroencephalography and
35:846–54. Clinical Neurophysiology 91:229–31.
8. Koffler, D. J., and J. Gotman. 1985. Automatic detec- 21. Morris, G. L. 1997. The clinical utility of computer-
tion of spike-and-wave bursts in ambulatory EEG assisted ambulatory 16 channel EEG. Journal of Med-
recordings. Electroencephalography and Clinical Neu- ical Engineering & Technology 21:47–52.
rophysiology 61:165–80. 22. Tatum, W. O. T., L. Winters, and M. Gieron,
9. Ives, J. R., N. R. Mainwaring, and D. L. Schomer. et al. 2001. Outpatient seizure identification: Results
1992. An 18-channel solid-state ambulatory EEG of 502 patients using computer-assisted ambula-
event recorder for use in the home and hospital envi- tory EEG. Journal of Clinical Neurophysiology 18:
ronment (abstract). Epilepsia 33(Suppl. 3):63. 14–9.
10. Velis, D., P. Plouin, J. Gotman, and F. L. da Silva. 2007. 23. Liporace, J., W. T. Tatum, G. L. Morris III, and
Neurophysiology IDSo. Recommendations regard- J. French. 1998. Clinical utility of sleep-deprived ver-
ing the requirements and applications for long-term sus computer-assisted ambulatory 16-channel EEG
recordings in epilepsy. Epilepsia 48:379–84. in epilepsy patients: A multi-center study. Epilepsy
11. Gilliam F., R. Kuzniecky, and E. Faught. 1999. Ambu- Research 32:357–62.
latory EEG monitoring. Journal of Clinical Neuro- 24. Tuunainen, A., and U. Nousiainen. 1993. Ictal record-
physiology 16:111–15. ings of ambulatory cassette EEG with sphenoidal elec-
12. Ives J. R., and J. F. Woods. 1979. A study of 100 trodes in temporal lobe epilepsy: Comparison with
patients with focal epilepsy using a 4-channel ambula- intensive videomonitoring. Acta Neurologica Scandi-
tory cassette recorder, ISAM 1979: Proceedings of the navica 88:21–5.
third international symposium on ambulatory monitor- 25. Schomer, D. L., J. R. Ives, and S. C. Schachter.
ing. London: Academic Press. 1999. The role of ambulatory EEG in the evaluation
13. Aminoff, M. J., D. S. Goodin, B. O. Berg, and of patients for epilepsy surgery. Journal of Clinical
M. N. Compton. 1988. Ambulatory EEG record- Neurophysiology 16:116–29.
ings in epileptic and nonepileptic children. Neurology 26. Chang, B. S., J. R. Ives, D. L. Schomer, and
38:558–62. F. W. Drislane. 2002. Outpatient EEG monitoring in
14. Kerr, S. L., D. W. Shucard, M. H. Kohrman, and the presurgical evaluation of patients with refractory
M. E. Cohen. 1990. Sequential use of standard and temporal lobe epilepsy. Journal of Clinical Neurophys-
ambulatory EEG in neonatal seizures. Pediatric Neu- iology 19:152–6.
rology 6:159–62. 27. Waterhouse, E. 2003. New horizons in ambula-
15. de Feo, M. R., O. Mecarelli, G. Ricci, and M. F. Rina. tory electroencephalography. IEEE Engineering in
1991. The utility of ambulatory EEG monitoring Medicine & Biology Magazine 22:74–80.
Chapter 13
Prolonged Video
Electroencephalography
Cheolsu Shin
INTRODUCTION Psychogenic Seizures or

EQUIPMENT Nonepileptic Behavioral Events
CLINICAL APPLICATION Classification of Seizure Type
Epileptic Vs. Nonepileptic PVEEG and Surgical Evaluation
Events
SUMMARY
INTRODUCTION manifestations of seizures when they view the

event only once. Capturing the seizure on
Prolonged video electroencephalography video allows repeated review of the event and
(PVEEG) has become an essential tool in the comparison with subsequent events. PVEEG
evaluation of patients with epilepsy and other provides physicians and allied health staff an
paroxysmal or episodic events.1–3 Often, the extended period of observation that is often
clinical history alone is not adequate for mak- helpful in searching out psychosocial issues
ing an accurate diagnosis, partly because of that may contribute to the patient’s condition.
inadequate observer history—the patients fre- PVEEG can be used to differentiate epilep-
quently are unaware and have to report a tic from nonepileptic events, and it can help
second-hand history that usually is sketchy at classify seizure type. PVEEG is also impor-
best. Even with activation procedures such as tant in localizing the epileptogenic region in
sleep, sleep deprivation, hyperventilation, and patients being considered for epilepsy surgery.
photic stimulation, epileptiform abnormalities With modifications of this technique, neonates,
may not appear on routine electroencephalo- infants and children, patients in intensive care
graphic (EEG) recordings, which may last an units, and patients with other episodic events
hour at most. Accurate diagnosis often requires such as sleep disorders can be evaluated.
capturing the clinical spell in conjunction with This chapter describes various techniques of
simultaneous electrophysiologic and behav- PVEEG and discusses its clinical applications.
ioral monitoring. PVEEG allows behavioral In 1949, Hunter and Jasper,4 Schwab et al.,5
correlation of the patient’s clinical spells with and Stewart et al.6 described systems for cin-
the EEG. The most experienced examiners ematographic and EEG recordings. The early
may miss subtle or even obvious clinical split-screen systems used mirrors to record
193
simultaneously the patient’s behavioral mani- EEG recorded on one of the audio channels
festations and the EEG. They were bulky and or a system that records the EEG digitally on
could record only 1 hour of cinema before the the videotape or a computer with time synchro-
film reel had to be changed, making the sys- nization. Most new systems utilize all digital
tems quite impractical. The development of formats for both video and EEG signals. How-
the videotape player and more compact camera ever, there are pitfalls to be aware of even with
units led to greater acceptance of the proce- digital signal processing.9 All digital does not
dure. Even today, a simple prolonged EEG always mean all correct. Although the digiti-
can be obtained using a video camcorder and zation may be accurate, display of the EEG
a standard EEG. An obvious disadvantage of tracing is on the computer monitor that has
this technique is that paper must be printed limited resolution. Sometimes, it is necessary
for the entire monitoring session. Currently, to review 5-second epochs instead of the usual
video EEG units with digitally encoded EEG 10-second epochs to allow for correct repre-
stored on a computer system or the video- sentation of higher frequency signals on the
tape itself are commercially available from monitor. There may also be artifacts of digitiza-
many sources. Most systems use telemetered tion such as aliasing of the signal. This can be
EEG with either cable or radio telemetry dealt with through the use of the antialiasing
and remote control video cameras, allowing filter, but manufacturers may differ in some of
relatively free movement of patients in the these details. In cases of intracranial monitor-
monitoring unit. ing, the simultaneous display of large number
of channels up to 128 make it difficult to see
Purpose and Role of Prolonged Video EEG the low-amplitude, fast-frequency signals. In
• Video and EEG correlation of the parox- those circumstances, reviewing select number
of channels in different subgroups is necessary.
ysmal events for diagnosis of epileptic vs.
Infrared cameras are used at night to achieve
nonepileptic conditions.
• Classification of seizure types. reasonable quality video recording even in
• Localization of seizure onset zone for a darkened room. The patient may be con-
nected to the recording equipment with a long
epilepsy surgery.
EEG cable or the EEG data may be sent
through a radio transmitter. With cable teleme-
try, patients have limited mobility. However,
EQUIPMENT the technical quality of cable or hardwired sys-
tems is usually superior to that of radioteleme-
The variability among the PVEEG systems tered systems. With radiotelemetry, patients
used in laboratories throughout the world is have greater mobility, which may or may not
considerable because of the different manu- be an advantage given the circumstances of
facturers.7 Each manufacturer uses custom- a patient’s seizures, seizure frequency, and
designed hardware and software to encode and severity.
decode the EEG signal during acquisition and With rapidly advancing computer and net-
analysis. This makes it difficult to exchange working technology, both in hardware and in
the raw EEG data between different systems. software, most systems now use digital tech-
Miniaturization of electronics and computer nology at least for EEG signal processing.
equipment allows PVEEG on an outpatient EEG data can be stored on optical disks or
basis because the patient and family can eas- other computer storage devices. The cost of
ily transport the system. This can be used for these digital PVEEG systems can be high,
patients with frequent events that may not but because of their superior recording, data
require detailed testing by medical person- processing, and analysis capabilities, they are
nel.8 However, in most patients, PVEEG is gaining wider acceptance as an industry stan-
performed in an inpatient epilepsy unit where dard. Digital PVEEG allows online processing
trained medical personnel provide continu- of EEG activity, with automatic detection of
ous surveillance along with video and EEG seizures and interictal epileptiform activity.10, 11
monitoring. Many software programs are available for this
Currently, several systems are available com- online detection, but all of them require ver-
mercially for PVEEG. Older versions used ification off-line because movement artifacts
a VHS video system with either multiplexed or rhythmic or sharp sleep transients can
Prolonged Video Electroencephalography 195
trigger the detection paradigm.12, 13 Montage subtle or subclinical electrographic seizures

reformatting, filtering, frequency analysis, tim- and test the patient during the seizure and
ing analysis, and correlation studies can be alert the treatment team to possible status
performed off-line with these systems. The epilepticus when overt clinical activity may
ability to analyze a single seizure carefully in not be obvious. Competent technical personnel
many ways and to view it with several different are essential for conducting specialized tests
montages has decreased the number of actual such as ictal single-photon emission computed
seizures that need to be recorded. Because of tomography (SPECT), in which it is vital to rec-
the risks involved in recording seizures and in ognize the seizure rapidly, either by behavior or
tapering seizure medications and the expense by EEG.
of the procedure, it is important to obtain
as much information from as few seizures as Key Points
possible.
Video signals are also being digitized for net- • Progress in digital equipments and data
work transmission as well as for digital storage. processing has led to a major improve-
This enables remote monitoring of patients, ment in PVEEG, but there are pitfalls in
for example, from an intensive care unit, and digital analysis to be aware of.
the monitoring includes both video and EEG. • Competent technical personnel in atten-
However, because video signals are such large dance for inpatient PVEEG is essential
data sets, a larger and faster capacity network for high-quality data acquisition as well as
system is necessary for online access and the testing of patients during the paroxysmal
quality of the video may be inferior to that of events.
an analog video system. One could also moni-
tor digital EEG data online through the inter-
net access as long as the hospital information CLINICAL APPLICATION
technology system insures security of the net-
work access. Video signal online again requires
very high capacity network and therefore is of Epileptic Vs. Nonepileptic Events
poor quality through the internet. These prob-
lems are likely to be overcome with rapidly PVEEG often can help answer the fundamen-
advancing technology. tal question of whether a spell is an epilep-
Portable PVEEG systems can also be used tic event.14 Many different paroxysmal events
in intensive care units, neonatal units, and psy- can mimic seizures clinically. Distinguishing
chiatric facilities. These portable systems can epileptic and nonepileptic events is critical for
be helpful in capturing events that are too determining effective therapy, because anti-
infrequent to be seen on routine EEG but seizure medications are rarely beneficial for
do not need online monitoring. The systems conditions other than seizures and have poten-
can include a video camera and recorder syn- tially significant risks.
chronized to the EEG. The data are analyzed Recurrent episodes of loss of consciousness
off-line later, as in PVEEG monitoring of out- can result from various nonepileptic causes,
patients. including syncopal attacks from aortic steno-
Most of these systems are designed so that sis, cardiac arrhythmias, vasovagal depression,
continuous video EEG monitoring can pro- orthostatic hypotension, and hyperventilation.
ceed without the continuous attendance of Vertigo and nonspecific dizzy spells may also be
technical personnel. However, most inpatient difficult to differentiate from seizures clinically.
PVEEG monitoring units have technical per- PVEEG with simultaneous electrocardiogra-
sonnel dedicated to monitoring several patients phy (ECG) and blood pressure monitoring may
continuously around the clock. The person- help to delineate the nonepileptic nature of
nel perform essential functions such as testing these events. Rarely, cardiac arrhythmias can
and examining patients during and after the be triggered by an epileptic discharge (ictal
seizure or spell and operating remote control syncope); this can be documented by PVEEG.
cameras to focus or zoom in on the patient, Various movement disorders can be con-
and they ensure high-quality EEG recordings fused with seizures. Simultaneous surface elec-
by repairing faulty electrodes or connections. tromyographic (EMG) recording of agonist
Also, the technical personnel can recognize and antagonist muscles with continuous EEG
recording can be helpful in diagnosing move- psychiatric treatment and discontinuation of

ment disorders and ruling out an epilep- unnecessary and potentially dangerous med-
tic cause. ical therapy. PVEEG may be advantageous
Sleep disorders and seizures may be confused economically in the diagnosis and treatment
with one another. PVEEG allows behavior to of psychogenic seizures, because patients with
be correlated with the EEG and other variables such a condition consume a significant amount
such as respiration, eye movements, surface of medical resources, including the costs of
EMG, ECG, and oxygen saturation. PVEEG repeated visits to the emergency department.18
may also help differentiate transient ischemic In most cases, psychogenic seizures are not
attacks, hypoglycemic attacks, and migraine under conscious control. Factitious disorders,
attacks from seizures. Blood glucose monitor- such as Munchausen’s syndrome or actual
ing can be performed at the time of the attacks willful malingering, are rare. Gates and col-
to rule out hypoglycemia. leagues19 compared the clinical manifestations
PVEEG can be especially helpful in of psychogenic generalized tonic–clonic events
differentiating types of childhood spells. with those of true generalized tonic–clonic
Daydreaming, breathholding, migraines, night seizures and found that out-of-phase clonic
terrors, and other parasomnias can be con- movements of the extremities, pelvic thrust-
fused with seizures. Occasionally, nocturnal ing, lack of eye manifestations, side-to-side
epileptic seizures are misdiagnosed clinically as head movements, and early vocalizations were
physiologic parasomnias. Although outpatient more common in the psychogenic group. Psy-
monitoring under the supervision of parents chogenic seizures also tend to last longer than
may be adequate for classifying these types 2 minutes; patients with psychogenic seizures
of spells, inpatient setting usually yields more often have long attacks, with frequent rest peri-
reliable data.15 ods during attacks. Also, these patients often
Repetitive motor activities are common in respond to some degree to verbal or noxious
patients with altered mental status in inten- stimulation during the generalized movements.
sive care units. Many of these movements are However, caution is needed in making a clini-
manifestations of underlying cerebral or spinal cal judgment on the basis of the history alone,
lesions or reflections of toxic and metabolic because many of these ictal features can occur
derangement. However, some of these move- in epileptic seizures, especially ones with an
ments may represent epileptic events, possibly extratemporal focus.20
subtle status epilepticus, which would explain Although a spectrum of psychiatric diag-
the altered mental status. Because untreated noses are represented in patients with psy-
generalized status epilepticus can cause per- chogenic seizures and as many as one-third of
manent neuronal damage, this possibility needs them have a history of sexual or physical abuse
to be kept in mind and pursued for appropri- or assault, some may not have any easily iden-
ate diagnosis and treatment with antiseizure tifiable psychiatric or psychologic problem.
medications. PVEEG with a portable system In addition, these psychiatric and psycho-
or online remote monitoring can be helpful in logic factors are also associated with patients
diagnosing these movements. with epilepsy. To complicate matters further,
some epileptic patients also have psychogenic
Key Points seizures, and in some series, 10%–40% of
patients with psychogenic seizures also had
• There are nonepileptic and nonpsy-
epileptic seizures.21 Therefore, the correla-
chogenic causes to many paroxysmal tion of each type of ictal event with the
events that can be diagnosed with PVEEG. EEG is essential to exclude epilepsy as a
cause.
PVEEG allows careful analysis of behav-
Psychogenic Seizures or ioral and EEG manifestations in patients with
Nonepileptic Behavioral Events psychogenic seizures. Muscle and movement
artifacts are usually prominent, but there is no
Psychogenic seizures are a difficult diagnos- ictal EEG change. After the movement stops,
tic problem.16, 17 Accurate diagnosis of psy- the normal background EEG usually returns
chogenic seizures can lead to appropriate immediately. Postictal slowing and suppression
of the EEG, which are typically present after Classification of Seizure Type
a true generalized tonic–clonic seizure, do not
occur after a psychogenic seizure. A caveat is PVEEG may lead to a reclassification of the
that the patient needs to be tested during the seizure type in many patients with uncon-
ictal events to assess the responsiveness and trolled seizures, especially if the diagnosis is
memory processing. If there is no alteration in question, thus improving medical manage-
of consciousness, the possibility of a simple ment.
partial seizure cannot be ruled out. Surface Distinguishing primary generalized from
EEG changes may not be seen with the sim- secondary generalized seizures is often diffi-
ple partial seizures, and this may be true of the cult only on the basis of the clinical history
majority of cases.22 and routine EEG. Some patients have rapid
The timing of the spells is also helpful infor- secondary spread and may have an interictal
mation because psychogenic seizures are more EEG that shows generalized spike-and-wave
likely to occur during the day. If they occur discharges (secondary bilateral synchrony).
at night, it is during wakefulness. Although However, ictal recording may demonstrate
some patients may claim that the spells occur that seizure onset is focal. Other patients
out of sleep, they invariably wake up first with true absence or true generalized tonic–
(as shown by the change in EEG pattern clonic seizures may mistakenly be thought
from sleep to wakefulness) and then have the to have focal epilepsy, because of the asym-
spell.23 Epileptic seizures during sleep tend to metrical manifestation of generalized dis-
occur directly out of sleep, without intervening charges. By clinical history alone, absence
wakefulness. Laboratory studies on changes seizures may be indistinguishable from com-
in the serum levels of prolactin or neuron- plex partial seizures. In virtually all patients
specific enolase can corroborate the PVEEG with untreated absence seizures, hyperventi-
studies.24, 25 Epileptic generalized tonic–clonic lation will activate 3-Hz spike-and-wave dis-
seizures or complex partial seizures—but not charges. The diagnosis may be more difficult
psychogenic seizures—are associated with a to make in patients with infrequent spells or
significant increase in the levels of these sub- those taking medication and may benefit from
stances above the baseline values for that PVEEG. Medications can be tapered when the
patient. patient is in the hospital and being carefully
Induction or provocation of psychogenic observed.
seizures is a matter of controversy.26 Certain Monitoring can also help to differenti-
triggering situations can be used if the his- ate temporal from extratemporal seizures.27–29
torical information about the reliable trigger- Many patients with simple partial seizures
ing factors is clear. Some clinicians have used have no EEG accompaniment. However, most
saline injections or a tuning fork with a strong patients with complex partial seizures have
suggestion that a seizure will occur. However, an ictal EEG change. Temporal lobe seizures
to interpret the results, it has to be verified often begin with an attenuation of scalp activ-
that the induced spell is the same type as the ity, followed by a rhythmic discharge, usually
noninduced spell. in the theta range that increases in amplitude
and becomes more widespread. Postictally, a
Key Points focal slowing often occurs over the temporal
region where the seizure began. Tachycardia is
• PVEEG is necessary for the accurate observed in most patients with complex partial
diagnosis of nonepileptic behavioral seizures of temporal lobe origin, even before
spells. any significant motor activity is apparent. Lat-
• In many patients, nonepileptic behavioral eralized ictal posturing of the upper extrem-
spells coexist with epileptic seizures, mak- ity contralateral to the ictal onset is observed
ing it important to analyze all different with many seizures of temporal onset.30 Also,
spell types. there is forced turning of the head away from
• Correct diagnosis of nonepileptic behav- the ictal focus just before secondary gener-
ioral spells using PVEEG allows for more alization.31 Postictally, there may be paresis
efficient and appropriate care of these or neglect on the side contralateral to the
conditions. seizure focus.
Frontal lobe seizures tend to be shorter and who are being considered for epilepsy surgery.36
to cause less postictal confusion than temporal The most important part of the presurgical
lobe seizures. Frontal lobe seizures are associ- evaluation probably is the recording of the
ated with frequent falls, because of the rapid patient’s typical seizures on continuous EEG
bilateral spread. Focal tonic, or fencing, pos- with video monitoring.37 This serves two pur-
tures may be seen and may suggest a focus in poses at the outset: first, to establish that the
the supplementary motor area. Certain frontal refractory habitual seizures are indeed epilep-
lobe seizures may mimic absence seizures and, tic and not a nonepileptic behavioral spell
at times, have been referred to as pseudoab- and, second, to establish the localization of the
sences. Frontal lobe seizures often begin with epileptic focus electrophysiologically. Because
low-amplitude fast activity but can be associ- seizures occur sporadically and unpredictably,
ated with some frontal sharp waves or spikes. treatment with antiepileptic medications is
PVEEG can often help identify these frontal usually withdrawn rather rapidly to expedite
lobe seizures, although at times the movement the recording of seizures.38 Some concern
artifacts from hypermotor behavior may make exists about whether seizures recorded during
the identification of ictal discharge difficult. In acute withdrawal of medication faithfully rep-
that situation, other factors should be consid- resent the patient’s habitual seizure pattern. An
ered, such as stereotypic behavior, onset during additional risk is the possibility of a secondar-
sleep, or postical slowing of EEG. ily generalized tonic–clonic seizure in patients
Inpatient PVEEG can be helpful in distin- who had only complex partial seizures and
guishing various spells and multiple seizure the possibility of a generalized tonic–clonic or
types in patients with Lennox–Gastaut syn- complex partial status epilepticus. The PVEEG
drome.32 Patients with this syndrome can have monitoring unit should be well equipped and
different types of epileptic seizures as well the personnel should be expertly trained in
as stereotyped mannerisms, tics, and other the management of these neurologic emergen-
movements that are not epileptic. Because cies. Ready intravenous access must be main-
these phenomena tend to occur almost daily tained by way of a heparin-lock system, and
in this patient population, only a relatively intravenous formulations of benzodiazepines
short session of inpatient PVEEG monitoring (lorazepam) or fosphenytoin should be imme-
is needed. After the various spells and seizures diately available. Patients and families need
have been classified, outpatient management is to be well informed about these issues before
simpler because antiepileptic drug adjustments or at the time of admission to the monitoring
are not necessary for recurrent nonepileptic unit.
spells. Through education and reassurance of PVEEG allows careful analysis of clinical
parents and caretakers, emergency department and EEG changes and can be used with scalp,
visits or unnecessary rectal administration of sphenoidal, foramen ovale, or implanted depth
diazepam can be avoided. or subdural strip or grid electrodes. With
computer-assisted recordings, montage refor-
Key Points matting, and off-time analysis of seizures, there
may be less need for invasive EEG recordings.
• Correct seizure classification through For surgical monitoring, digital video EEG
PVEEG makes possible more rational systems with 64–128 channels are commonly
pharmacological treatment of the epilep- used. One important caveat is that computer
tic condition. monitor does not allow adequate resolution
of each channel activity when all the record-
ing channels are displayed. Low-amplitude fast
PVEEG and Surgical Evaluation beta or gamma activity, frequently indicative of
cortical seizure focus, may be ignored, lead-
Surgical treatment frequently eliminates or ing to misplaced seizure onset localization. One
decreases the frequency and severity of needs to view subgroups of channels with man-
seizures in many partial epilepsies.33, 34 Seizures ageable numbers (e.g., 20) with higher sensi-
arising from the temporal lobe are especially tivity to detect low-amplitude signals of seizure
amenable to surgical treatment.35 A compre- onset zones. For focal cortical resection, typical
hensive evaluation is performed on patients seizures must be shown to arise from a single
region of the brain. An adequate number of Key Points

seizures must be recorded so that the seizure • PVEEG is an essential first step in
can be localized confidently. If the patient
the evaluation of medically intractable
has multiple seizure types, then all of them
epileptic syndromes toward surgical solu-
need to be captured and analyzed to determine
tion and is necessary for intracranial elec-
whether the seizures have multifocal onset
trocorticographic monitoring.
or a single focus with different propagation • Personnel, both technical and nursing,
paths.
should be well versed in dealing with the
When intracranial electrodes are implanted
more severe motor seizures and status
and the identified epileptogenic zone is near a
epilepticus that may occur during medica-
functional cortical region (e.g., motor, sensory,
tion withdrawal.
language, or visual cortex), the determination • Intracranial electrodes can be used for
of cortical function is necessary. A train of
functional mapping of areas surrounding
electric pulses (usually biphasic rectangular
the epileptic focus to be resected during
pulses at 60 Hz from a stimulation isolation
presurgical PVEEG.
unit) can be applied through pairs of elec-
trodes through the special interface device
coupled to the PVEEG input stage. Dur-
ing the delivery of the stimulation, positive SUMMARY
motor, sensory, or visual phenomena can be
observed or reported by the patient. For the PVEEG has become an essential diagnostic
language area, stimulation causes the arrest tool in the management of epileptic seizures,
of language function, such as a pause in psychogenic seizures, and other paroxysmal
reading a sentence aloud or naming objects. events. Various monitoring systems are avail-
EEG recording must be performed to verify able, mostly all digital systems. PVEEG mon-
that the stimulation did not cause an after- itoring allows psychogenic seizures and other
discharge (electrographic seizure), because nonepileptic paroxysmal events to be differen-
that would suggest a propagated effect away tiated from epileptic seizures. It also enables
from the stimulated electrodes. This can be accurate classification of most seizure types
done in the PVEEG monitoring unit out- and is a necessary part of the evaluation of
side the operating room and obviate intra- patients being considered for surgical treat-
operative cortical mapping with the patient ment of epilepsy. Also, this type of monitoring
awake. allows extended observation of patients and
In evaluating medically refractory partial can lead to a better understanding of their
epilepsy, PVEEG from the scalp electrodes medical problems and any superimposed psy-
alone is not always adequate for localiz- chosocial issues. All these factors contribute to
ing seizure onset. Subtraction ictal–interictal improved quality of life for patients and their
SPECT coregistered on magnetic resonance families and more efficient delivery of appro-
images (SISCOM) can be useful for either priate medical care in these often very complex
avoiding further invasive intracranial monitor- situations.
ing or guiding the placement of intracranial
electrodes.39, 40 The ictal SPECT scan needs to
be performed with exquisite temporal urgency.
The sooner the tracer is injected after the onset REFERENCES
of the seizure (usually within tens of seconds),
the more likely the scan will reflect the foci of 1. Engel, J. Jr, J. Burchfiel, and J. Ebersole, et al. 1993.
ictal increase in blood flow. Therefore, it is crit- Long-term monitoring for epilepsy. Report of an
IFCN committee. Electroencephalography and Clin-
ical to monitor the patient closely for signs of ical Neurophysiology 87:437–58.
seizure onset, either by behavior or by EEG. 2. Kaplan, P. W., and R. P. Lesser. 1990. Long-term
This is possible only in the setting of inpatient monitoring. In Current practice of clinical electroen-
PVEEG with appropriately trained monitoring cephalography, ed. D. D. Daly, and T. A. Pedley, 2nd
ed. 513–34. New York: Raven Press.
personnel and ready access to and availabil- 3. Mizrahi, E. M. 1999. Pediatric electroencephalo-
ity of the radioactive tracer from the nuclear graphic video monitoring. Journal of Clinical Neuro-
medicine department. physiology 16:100–10.
4. Hunter, J., and H. H. Jasper. 1949. A method of analy- epileptic patients. The American Journal of Psychiatry
sis of seizure pattern and electroencephalogram: A cin- 137:705–9.
ematographic technique. Electroencephalography and 22. Bare, M. A., T. H. Burnstine, R. S. Fisher,
Clinical Neurophysiology 1:113–14. and R. P. Lesser. 1994. Electroencephalographic
5. Schwab, R. S., M. W. Schwab, D. Withee, and changes during simple partial seizures. Epilepsia
Y. C. Chock. 1954. Synchronized moving pictures 35:715–20.
of patient and EEG. Electroencephalography and 23. Bazil, C. W., and T. S. Walczak. 1997. Effects of sleep
Clinical Neurophysiology 6:684–6. and sleep stage on epileptic and nonepileptic seizures.
6. Stewart, L. F., H. H. Jasper, and C. Hodge. 1956. Epilepsia 38:56–62.
Another simple method for simultaneous cinemato- 24. Rabinowicz, A. L., J. Correale, R. B. Boutros,
graphic recording of the patient and his electroen- W. T. Couldwell, C. W. Henderson, and
cephalogram during seizures. Electroencephalography C. M. DeGiorgio. 1996. Neuron-specific enolase
and Clinical Neurophysiology 8:688–91. is increased after single seizures during inpatient
7. Lagerlund, T. D., G. D. Cascino, K. M. Cicora, video/EEG monitoring. Epilepsia 37:122–5.
and F. W. Sharbrough. 1996. Long-term electroen- 25. Tumani, H, M. Otto, and O. Gefeller, et al. 1999.
cephalographic monitoring for diagnosis and man- Kinetics of serum neuron-specific enolase and pro-
agement of seizures. Mayo Clinic Proceedings 71: lactin in patients after single epileptic seizures. Epilep-
1000–6. sia 40:713–18.
8. Gilliam, F., R. Kuzniecky, and E. Faught. 1999. Ambu- 26. Dericioglu, N., S. Saygi, and A. Ciger. 1999. The value
latory EEG monitoring. Journal of Clinical Neuro- of provocation methods in patients suspected of having
physiology 16:111–15. nonepileptic seizures. Seizure 8:152–6.
9. Epstein, C. M. 2006. Digital EEG: Trouble in 27. Marks, W. J. Jr, and K. D. Laxer. 1998. Semiology
paradise? Journal of Clinical Neurophysiology 23: of temporal lobe seizures: Value in lateralizing the
190–3. seizure focus. Epilepsia 39:721–6.
10. Gotman, J. 1982. Automatic recognition of epilep- 28. Williamson, P. D., V. M. Thadani, and J. A. French,
tic seizures in the EEG. Electroencephalography and et al. 1998. Medial temporal lobe epilepsy: Videotape
Clinical Neurophysiology 54:530–40. analysis of objective clinical seizure characteristics.
11. Gotman, J. 1985. Seizure recognition and analysis. Epilepsia 39:1182–8.
Electroencephalography and Clinical Neurophysiol- 29. Geyer, J. D., T. A. Payne, E. Faught, and I. Drury.
ogy Supplement 37:133–45. 1999. Postictal nose-rubbing in the diagnosis, lat-
12. Salinsky, M. C. 1997. A practical analysis of computer eralization, and localization of seizures. Neurology
based seizure detection during continuous video- 52:743–5.
EEG monitoring. Electroencephalography and Clin- 30. Kotagal, P, A. Bleasel, E. Geller, P. Kankirawatana, B.
ical Neurophysiology 103:445–9. I. Moorjani, and L. Rybicki. 2000. Lateralizing value
13. Spatt, J, G. Pelzl, and B. Mamoli. 1997. Reliability of asymmetric tonic limb posturing observed in sec-
of automatic and visual analysis of interictal spikes in ondarily generalized tonic–clonic seizures. Epilepsia
lateralising an epileptic focus during video-EEG mon- 41:457–62.
itoring. Electroencephalography and Clinical Neuro- 31. Theodore, W. H., R. J. Porter, and P. Albert,
physiology 103:421–5. et al. 1994. The secondarily generalized tonic–
14. Mohan, K. K., O. N. Markand, and V. Salanova. clonic seizure: A videotape analysis. Neurology 44:
1996. Diagnostic utility of video EEG monitoring 1403–7.
in paroxysmal events. Acta Neurologica Scandinavica 32. Bare, M. A., T. A. Glauser, and R. H. Strawsburg.
94:320–5. 1998. Need for electroencephalogram video confir-
15. Connolly, M. B., P. K. Wong, Y. Karim, S. Smith, and mation of atypical absence seizures in children with
K. Farrell. 1994. Outpatient video-EEG monitoring in Lennox–Gastaut syndrome. Journal of Child Neurol-
children. Epilepsia 35:477–81. ogy 13:498–500.
16. Krumholz, A. 1999. Nonepileptic seizures: Diagnosis 33. Engel, J., ed. 1993. Surgical treatment of the epilep-
and management. Neurology 53:S76–83. sies, 2nd ed. New York: Raven Press.
17. Scott, D. F. 1982. Recognition and diagnostic aspects 34. Penry, J. K., and R. J. Porter. 1977. Intensive monitor-
of nonepileptic seizures. In Pseudoseizures, ed. ing of patients with intractable seizures. In Epilepsy:
T. L. Riley, and A. Roy. 21–33. Baltimore: Williams & The 8th international symposium, ed. J. K. Penry,
Wilkins. 95–101. New York: Raven Press.
18. Martin, R. C., F. G. Gilliam, M. Kilgore, E. Faught, 35. Kilpatrick, C., M. Cook, A. Kaye, M. Murphy, and
and R. Kuzniecky. 1998. Improved health care Z. Matkovic. 1997. Non-invasive investigations suc-
resource utilization following video-EEG-confirmed cessfully select patients for temporal lobe surgery.
diagnosis of nonepileptic psychogenic seizures. Journal of Neurology, Neurosurgery, and Psychiatry
Seizure 7:385–90. 63:327–33.
19. Gates, J. R., V. Ramani, S. Whalen, and R. Loewenson. 36. Sutula, T. P., J. C. Sackellares, J. Q. Miller, and
1985. Ictal characteristics of pseudoseizures. Archives F. E. Dreifuss. 1981. Intensive monitoring in refrac-
of Neurology 42:1183–7. tory epilepsy. Neurology 31:243–7.
20. Geyer, J. D., T. A. Payne, and I. Drury. 2000. The 37. Serles, W., Z. Caramanos, G. Lindinger, E. Pataraia,
value of pelvic thrusting in the diagnosis of seizures and C. Baumgartner. 2000. Combining ictal surface-
and pseudoseizures. Neurology 54:227–9. electroencephalography and seizure semiology
21. Ramani, S. V., L. F. Quesney, D. Olson, and R. J. improves patient lateralization in temporal lobe
Gumnit. 1980. Diagnosis of hysterical seizures in epilepsy. Epilepsia 41:1567–73.
38. Quiroga, R. C., L. Pirra, C. Podesta, R. C. Lei- extratemporal epilepsy surgery outcome. Neurology
guarda, and A. L. Rabinowicz. 1997. Time distribution 55:1668–77.
of epileptic seizures during video-EEG monitoring. 40. So, E. L., T. J. O’Brien, B. H. Brinkmann, and
Implications for health insurance systems in develop- B. P. Mullan. 2000. The EEG evaluation of single
ing countries. Seizure 6:475–7. photon emission computed tomography abnormali-
39. O’Brien, T. J., E. L. So, and B. P. Mullan, et al. 2000. ties in epilepsy. Journal of Clinical Neurophysiology
Subtraction peri-ictal SPECT is predictive of 17:10–28.
Chapter 14
Electroencephalographic Special
Studies
Gregory A. Worrell and Terrence D. Lagerlund
INTRODUCTION Cross-Spectral Analysis

QUANTITATIVE METHODS OF Interpolation Techniques
ELECTROENCEPHALO- Topographic Displays (Mapping)
GRAPHIC ANALYSIS Multivariate Statistical Methods
Fourier (Spectral) Analysis and Topographic Analysis
Spike, Sharp-Wave, Cortical Projection Techniques
High-Frequency Oscillation and Source Dipole Localization
Seizure Detection
MAGNETOENCEPHALOGRAPHY
Montage Reformatting
SUMMARY
Cross-Correlation Analysis
INTRODUCTION • Assists in making serial comparisons

between EEGs performed on the same
Digital computers can aid in extracting infor- subject at different times or between two
mation from electroencephalographic (EEG) subject groups.
waveforms that is not readily obtainable with
visual analysis alone. These computers may also
be used for quantitation of key features of
waveforms, which may be useful in accurate
EEG interpretation and also in making serial QUANTITATIVE METHODS OF
comparisons between EEGs performed on the ELECTROENCEPHALOGRAPHIC
same subject at different times or between two ANALYSIS
subject groups in scientific investigations. Dig-
ital computers may also partially automate the
interpretation of EEGs, particularly in prolon- Fourier (Spectral) Analysis
ged monitoring for epilepsy. This chapter dis-
cusses some of the uses of computers in EEG. The technique of spectral analysis is described
in Chapter 4. Electroencephalographic spec-
Purpose and Role of EEG Special Studies tral analysis has been used to assess the dom-
inant background frequencies in normal and
• Allows for quantitation of key features disease states, including dementia, cerebral
of waveforms to assist in accurate EEG infarction, cerebral neoplasms, and various
interpretations. toxic and metabolic disorders. Most abnormal
203
conditions increase the amount of slow-wave Spike, Sharp-Wave,

activity in an EEG and may decrease the High-Frequency Oscillation and
amount of faster frequency activity. Quantita- Seizure Detection
tive measures that may be used include (1) the
percentage power in the delta, theta, or delta + The detection of interictal and ictal epilep-
theta bands; (2) the percentage power in the tic discharges in an EEG by computer algo-
alpha, beta, or alpha + beta bands; (3) vari- rithms has received much attention because of
ous other power ratios (e.g., alpha to delta); the increasing use of prolonged EEG moni-
(4) the mean frequency in the entire spec- toring for epilepsy. In prolonged EEG mon-
trum or in a portion of the spectrum, such itoring, many EEG channels (21 or more in
as the alpha band; and (5) the spectral edge most cases for scalp, and often more than 100
frequency, which is often taken to be the fre- for intracranial recordings) are recorded over
quency below which 95% of the power in the many hours, days, or weeks, generating huge
entire spectrum occurs. Spectral analysis has amounts of data that must be reviewed to
also been used in intraoperative monitoring to assess the frequency, nature, and localization of
assess changes in depth of anesthesia or cere- epileptic discharges. Computer algorithms may
bral ischemia. Although spectral analysis per se be used to locate candidate interictal or ictal
can be applied only to one channel at a time, discharges for further study by an electroen-
useful comparisons can be made between the cephalographer and, thus, the need to scan
spectra of recordings in different regions of many pages of EEG tracings manually may
the head. As expected, focal brain lesions tend be avoided. The techniques of pattern recog-
to produce focal changes in power spectra, nition are discussed in Chapter 4. Spike and
whereas diffuse processes produce generalized sharp-wave detection generally relies on the
changes. Quantitative measures of asymmetry sharpness criteria applied to individual wave-
can be calculated, such as the left-to-right ratio forms on a channel-by-channel basis, although
of power in the entire spectrum or in indi- more advanced algorithms may use context
vidual frequency bands, such as alpha, delta, information and multichannel correlation to
and so forth. Some studies of power spec- improve specificity. High-frequency oscillation
tra during transient cerebral events, such as and seizure detection algorithms are more
epileptic seizures, have also been conducted. complex, because of the great variability in
The power spectra of a partial complex seizure types of ictal discharges, and generally are
of temporal lobe origin, for example, tend to based on detecting a sudden change in rhyth-
be complex, with several frequency compo- micity and amplitude of background activity
nents (often a fundamental with one or more simultaneously in several channels.
harmonics, that is, integral multiples of the Current state-of-the-art commercially avail-
fundamental frequency) whose amplitudes in able software has a reasonably good overall
various regions of the head may differ. The sensitivity, although the sensitivity may be inad-
spectra also evolve over time during the course equate for the seizures of certain patients.
of the seizure. Consequently, prolonged monitoring systems
usually do not rely exclusively on computer
detection of seizures but also make use of
Key Points
trained observers or patient and family mem-
• Spectral analysis is the decomposition bers to recognize and log seizure occurrence.
of an EEG time series into a sum The specificity of currently used algorithms is
of sinusoidal functions of different also fairly good; however, in many prolonged
frequencies. monitoring situations, a large variety of arti-
• Quantitative measures can be determined facts arising from patient’s activities and vari-
for EEGs in different disease states using ous waveforms of cerebral origin, for example,
spectral analysis. V waves in sleep, cause false detections that
• Spectral analysis can be used in intra- may outnumber true epileptic discharges by an
operative monitoring to assess changes order of magnitude for some patients. Thus,
in depth of anesthesia or cerebral a physician or technician must review all dis-
ischemia. charges detected by the computer, but the
Electroencephalographic Special Studies 205
ability to markedly decrease the amount suppose that recorded EEG data include the
of data to be reviewed still makes auto- following channels:
mated detection algorithms of great practical
value.
There has been considerable interest recently Channel Derivation
in interictal and ictal high-frequency oscilla-
tions (80–1000 Hz) recorded from intracra- 1 Fp1 − Cz
nial subdural and depth electrodes. Because 2 Fp2 − Cz
3 F3 − Cz
these oscillations are of very low amplitude
4 F4 − Cz
they are largely obscured by lower frequency 5 C3 − Cz
activity. Studies using automated detection of 6 C4 − Cz
high-frequency oscillations have demonstrated 7 P3 − Cz
a strong correlation of high-frequency oscilla- 8 P4 − Cz
tions with epileptogenic brain.1 9 O1 − Cz
10 O2 − Cz
11 A1 − Cz
Key Points 12 A2 − Cz
• Spike and sharp waves are recognized
as EEG signatures of epileptogenic brain Then, a new referential montage with ipsilat-
and are strongly associated with seizure eral ear reference can be created by subtract-
disorders. ing pairs of channels as follows:
• High-frequency oscillations (80–1000 Hz)
are oscillatory local field potentials from
intracranial EEG recordings. Channels Derivation
• Seizures are an abnormal brain activity
characterized by excessively synchronized 1–11 Fp1 − A1
rhythmic discharges of a population of 2–12 Fp2 − A2
neurons. 3–11 F3 − A1
• Computer algorithms have been devel- 4–12 F4 − A2
5–11 C3 − A1
oped for automated detection of epilepti- 6–12 C4 − A2
form spikes, sharp waves, high-frequency 7–11 P3 − A1
oscillations, and seizures. These algo- 8–12 P4 − A2
rithms make possible the data mining of 9–11 O1 − A1
large EEG data sets. 10–12 O2 − A2
New reference electrodes can also be cre-

ated by averaging data from two or more exist-
Montage Reformatting ing electrodes. For example, suppose an EEG
is recorded with a Cz reference. A new deriva-
Montage reformatting allows EEG montages tion such as O1 − A1/2 , where A1/2 represents
to be selected when the data are reviewed the average of the ear electrodes, can be cal-
independently of the montage used to acquire culated as follows: O1 − A1/2 = (O1 − Cz ) −
and to store the data. The same EEG segment 0.5(A1 − Cz ) − 0.5(A2 − Cz ). Other types of
can be viewed using a variety of different mon- montages, such as a common average reference
tages. To generate a derivation such as X1 − X2 , montage (in which the reference for each elec-
which does not exist in the EEG as recorded, trode is the average potential of all recorded
the computer looks for two existing channels electrodes) or a laplacian (source) montage3, 4
one of which records X1 against a reference (in which the reference for each electrode is
electrode and another which records X2 against the average of the four nearest neighbors to
the same reference, and then subtracts the two. that electrode), may also be generated easily by
That is, X1 −X2 = (X1 −R)−(X2 −R), where R is the computer. The same ictal discharge viewed
the reference. This allows new referential and with various montages (one as recorded, three
bipolar montages to be formed.2 For example, after reformatting) is shown in Figure 14–1.
Figure 14–1. An EEG recorded during a seizure. A, As recorded from a C3 /C4 average reference (AV on figure) that
is active. B, Reformatted to average ear reference (A1/2 in figure). C, Reformatted to longitudinal bipolar montage. D,
Reformatted to longitudinal laplacian montage (Nav, 4-neighbor average; N3, 3-neighbor average). (From Lagerlund, T.
D. 1991. Montage reformatting and digital filtering. In Epilepsy surgery, ed. H. Luders, 318–22. New York: Raven Press.
By permission of Lippincott Williams & Wilkins.)
Key Points number 1 but delayed by time T, then the

cross-correlation function is a maximum at
• Digital EEG makes montage reformatting
t = T, and periodically thereafter if both sig-
fast and efficient. nals are periodic. Thus, examination of cross-
correlation functions may be used as a means
to assess small interchannel time differences
(Fig. 14–2); one possible use of this is in
Cross-Correlation Analysis determining whether a bilateral epileptic dis-
charge in fact represents secondary bilateral
Cross-correlation analysis quantifies the rela- synchrony or is generalized from onset.
tionship between EEG signals recorded from
different derivations. The cross-correlation Key Points
function is the correlation between EEG signal
number 1 and EEG signal number 2 delayed • Cross-correlation analysis quantifies the
by time t, expressed as a function of t. If relationship between EEG signals recor-
EEG signal number 2 is identical to number ded from different derivations.
1, the cross-correlation function is the same • Examination of cross-correlation func-
as the autocorrelation function described in tions may be used as a means to assess
Chapter 4. If signal number 2 is similar to small interchannel time differences.
A B 2. Phase spectrum—the relative phase angle

I = 10 μV (0◦ –360◦ ) between corresponding fre-
quency components of signal number 1
T5-AV * and those of signal number 2, expressed
as a function of the frequency.
The phase angle, in turn, is related to timing

O1-AV differences between the two signals. Because
the phase angle between two sine or cosine
waves delayed by a fixed time T equals 360◦
multiplied by the product of the time delay T
OZ-AV and the frequency f , the time delay between
two signals can be calculated from the phase
spectrum φ by the formula5
O2-AV
φ
T=
360◦ × f
T6-AV Thus, if a plot of phase angle vs. frequency is

linear over some range of frequencies, the time
difference between the signals can be derived
0 500 1000 –500 0 500 by taking the slope of the plot divided by 360◦
Time (ms) Lag time (ms) (Fig. 14–3).
Figure 14–2. A, Autocorrelation functions and, B, cross-
correlation functions for a 1-second epoch of alpha activity
recorded using a C3 /C4 average reference (AV on figure).
The autocorrelations demonstrate the rhythmicity of the
Interpolation Techniques
signal but cannot be used to assess interchannel time
differences. The cross correlations demonstrate a system- Interpolation techniques are used to estimate
atically increasing time delay between channels as one goes the electric potential at intermediate scalp
from right, T6 , to left, T5 , across the head. positions from its known value at each elec-
trode position. This procedure is necessary
when constructing maps of scalp potential, and
it may form a preliminary step in the applica-
Cross-Spectral Analysis tion of various techniques of spatial, or topo-
graphic, analysis of EEG, such as calculating
The cross-power spectrum between EEG sig- accurate estimates of the laplacian of the scalp
nal number 1 and number 2 is the Fourier potential, multivariate statistical analyses, cor-
transform of the cross-correlation function. tical projection techniques, and source dipole
From the cross-power spectrum, one can cal- localization. Some interpolation methods that
culate two additional functions: are commonly used include the following:
1. Nearest neighbor inverse distance

1. Coherence function—a normalized func- weighted.6, 7
tion that takes on values between 0 and 1. 2. All-electrode inverse distance weighted.8
It is defined as the absolute square of the 3. Rectangular two-dimensional splines,
cross-power spectrum of signal numbers which conceptually minimize the bending
1 and 2, divided by the product of the energy of an infinite, elastic plate con-
power spectrum of signal number 1 and strained to pass through known points.9, 10
that of signal number 2. It is a conve- 4. Rectangular three-dimensional splines,
nient measure of the degree of correlation which interpolate potentials in three
between similar frequency components in dimensions, after which results may be
the two signals (0%–100%), expressed as projected as needed onto spherical or
a function of the frequency. ellipsoidal surfaces.11, 12
5. Spherical surface splines, which are at one instant in time; it shows the distribu-
applied to a spherical surface instead of tion of potentials on the head surface at that
a rectangular one.6, 9, 13 time and may facilitate localization of EEG
6. Spherical harmonic expansion, the equiv- abnormalities (while making it more difficult
alent of a Fourier series in spherical to appreciate their variation in time). Because
instead of rectangular coordinates.14, 15 potentials are actually measured at only a few
7. Single or multidipole source models, points on the head (that is, at the electrode
which localize source dipoles and then locations), an interpolation technique must be
predict the scalp potential these sources used to estimate potentials at all other scalp
would generate. points;7 thus, the information conveyed by a
topographic display is only as accurate as the
Note that methods 1–3 are based on a rect- interpolation technique used (Fig. 14–4). Var-
angular planar model of the scalp surface, and ious methods of topographic display include
methods 5 and 6 assume a spherical head and three-dimensional plots (potential on z axis vs.
should be more accurate. Only method 1 (the x and y coordinates), contour plots (connect-
simplest and least accurate) has been used ing all points on the head with the same value
widely in commercial EEG systems. of potential), gray scale intensity plots (degree
of darkness at each point on a head map cor-
responds to the potential at that point), and
color plots (color at each point on a head map
Topographic Displays (Mapping) corresponds to the potential at that point). As
a general rule, topographic maps should be
Electroencephalographic topographic, or spa- used only as an adjunct to ordinary time series
tial, maps are a way of displaying EEG data EEG displays and not as a replacement for
that differ significantly from the conventional them, because 75% or more of the information
multichannel amplitude-vs.-time plots (mon- derived from an EEG depends on the tem-
tages). However, topographic mapping is not poral rather than the spatial characteristics of
in itself a method of EEG analysis, because waveforms.
it only displays the raw data in a different In addition to maps of unprocessed EEG,
way. In its simplest form, a topographic map topographic mapping has been used to dis-
of scalp potential is a snapshot of the EEG play the spatial distribution of the results of
R.S. 82–776
A
A 3–5
2 sec 300 μV
B 3–5
46 points
B 180
Spectrum Correlation: 96%
Phase
50 Hz 50 Hz Slope: 4.7°/Hz
–180
Delay: 13 ms
1
Spectrum Coherence
0
50 Hz 50 Hz
Figure 14–3. A, Onset of a seizure discharge recorded with a bipolar montage from intracerebral electrodes in a patient
with epilepsy: A, A 3–5 are in amygdala and B 3–5 are in hippocampus. B, Power spectra of each channel (left) and
phase angle and coherence spectra (right); the upper coherence graph is the coherence itself, the lower graph is the lower
boundary of the 99% confidence interval of the coherence. For the range of frequencies indicated by the arrows, the phase
is linearly related to frequency with slope 4.7◦ per Hz, corresponding to a delay of 13 ms between channels. Positive slope
indicates that the first channel is leading. (From Gotman, J. 1983. Measurement of small time differences between EEG
channels: Method and application to epileptic seizure propagation. Electroencephalography and Clinical Neurophysiology
56:501–14. By permission of Elsevier Scientific Publishers.)
A
FP1 FP2
FP1 FP2
F7 F8
F3 F3 F4 F4
FZ
C4
C3
T3 C3 CZ C4 T4
P3 P4
F7 F8
P3 PZ P4
T5 T6 T4
T3
T5 O1 O2 T6
OZ
O2
O1
FZ
192 ms
CZ
20 μV
PZ 250 ms
OZ
B C D
18.38 13.97
6.40 6.30
9.31 10.19 7.55
4.37 9.50 11.75 11.13 2.31
11.23 14.26 10.80

9.49 8.04
16.65 10.01
14.36
Figure 14–4. Example of topographic map construction for visual evoked potential signals recorded referentially from
20 scalp electrodes. Each evoked potential is divided into 128 intervals of 4 ms. A, Individual evoked potentials for the
indicated electrode locations. B, Mean voltage values at each electrode location for the 4-ms interval beginning 192 ms
after the stimulus (corresponding to the vertical line in A). C, The grid of interpolation points (64 × 64) used; each of the
4096 points is assigned a voltage value by linear interpolation from the three nearest known points. D, The topographic map
of this evoked potential, using an equal interval intensity scale to represent voltage at each location. (From Duffy, F. H.,
J. L. Burchfiel, and C. T. Lombroso. 1979. Brain electrical activity mapping (BEAM): A method for extending the clinical
utility of EEG and evoked potential data. Annals of Neurology 5:309–21. By permission of the American Neurological
Association.)
various techniques of multichannel EEG anal- • Various methods of topographic display

ysis. For example, maps of power spectra— include three-dimensional plots, contour
power within a specific frequency band at var- plots, gray scale intensity plots, and
ious scalp locations—can be generated. Simi- color plots.
larly, maps of the laplacian of the scalp poten- • Topographic maps should be used only
tial, of principal components of the scalp as an adjunct to ordinary time series
potential (from multivariate statistical analy- EEG displays and not as a replacement
sis), or of estimated cortical potentials (from for them.
a cortical projection technique) can also be
produced.
Multivariate Statistical Methods
Key Points
and Topographic Analysis
• Visualization of EEG potentials and
sources on topographic maps and patient- The purpose of multivariate statistical methods
specific MRI aids in analysis of EEG data. of EEG topographic analysis is to achieve data
reduction from many simultaneously recorded a time series (montage format) or as a topo-
EEG signals. This is possible because of graphic display (map format). This technique
the redundancy of multichannel EEG data— has proved capable of resolving two adjacent
many activities or waveforms appear simulta- dipole sources in the cerebral cortex that could
neously in many different channels. Mathemat- not be resolved by inspection of the scalp
ical techniques such as factor analysis, principal EEG signals and has been used successfully
component analysis, eigenvector analysis, and to localize median nerve somatosensory evoked
independent component analysis are used to potentials and interictal epileptic discharges
reduce the observed EEG signals in multiple on the cortical surface, with confirmation by
channels to a minimum number of indepen- recordings from subdural electrode grids.
dent, or orthogonal, component signals. Indi-
vidual components may be displayed spatially
as topographic maps or temporally as derived Key Points
EEG channels. Although these methods may • Cortical projection techniques such as
provide data reduction, their major drawback is spatial deconvolution and deblurring are
that the resultant independent signals are not designed to reverse the smearing effect of
always recognizable as traditional pure EEG the skull on scalp EEG by using a model
activities such as alpha or mu, and compari- of volume conduction in the head.
son between analyses made on the same EEG • This technique has proved capable of
at different times or on different subjects is resolving two adjacent dipole sources in
difficult. Also, these methods generally do not the cerebral cortex that could not be
provide information on the nature and location resolved by inspection of the scalp EEG
of physiologic generators of EEG. However, signals.
principal component analysis and independent
component analysis may be used to create a
type of spatial filter that can aid in the removal
of certain types of artifact, such as ocular move- Source Dipole Localization
ment and electrocardiographic artifact, from
an EEG recording.16, 17 The ultimate goal of localization of abnor-
mal activity, such as epileptic discharges, from
Key Points EEG is to find the intracranial sources generat-
ing a given distribution of scalp potentials. This
• The purpose of multivariate statistical is sometimes called the inverse problem (the
methods of EEG topographic analysis is to forward problem refers to finding the distribu-
achieve data reduction from many simul- tion of scalp potentials resulting from a known
taneously recorded EEG signals. distribution of intracranial sources). Although
the forward problem has a unique solution, the
inverse problem does not; that is, there are an
Cortical Projection Techniques infinite number of different sets of intracranial
generators that could produce any given distri-
Cortical projection techniques such as spatial bution of scalp activity. To constrain the prob-
deconvolution18 and deblurring19 are designed lem, certain physiologically based assumptions
to reverse the smearing effect of the skull on must be made about the number and approxi-
scalp EEG by using a model of volume con- mate location of generators. Most approaches
duction in the head (such as a three-sphere to solving the inverse problem have concen-
model of brain, skull, and scalp or an anatomi- trated on finding the location, orientation, and
cally based boundary element or finite element strength of a single dipole generator whose
model constructed from magnetic resonance potential field best matches actual data. This
scans of the patient’s head). The electric poten- is done with a least-squares minimization algo-
tial at selected points on the brain surface is rithm, which varies the dipole coordinates and
calculated from the electric potential at the direction to minimize the sum of squares of
scalp surface, thus noninvasively providing a the differences between the predicted and
distribution of electrical activity at the cortical the actual potentials at each electrode location
surface. Cortical potentials may be displayed as on the head.20, 21 The assumption of a single
Figure 14–5. Dipole modelling of a spike discharge recorded with a sternoclavicular reference from a patient with
epilepsy. A, Map of measured spike discharge distribution. B, Map of distribution of potential based on fitted dipole. C,
Fitted dipole located in right frontocentral region (long arrow), with short arrow indicating orientation of dipole negativity.
(From Thickbroom, G. W., H. D. Davies, W. M. Carroll, and F. L. Mastaglia. 1986. Averaging, spatio-temporal mapping and
dipole modelling of focal epileptic spikes. Electroencephalography and Clinical Neurophysiology 64:274–7. By permission
of Elsevier Scientific Publishers.)
dipole generator is most useful for small gener- These magnetic fields are generated by cur-
ators, such as the generators of certain evoked rent flowing in neurons, with a small contri-
potential peaks or of some epileptic spikes bution from extracellular current flow in the
(Fig. 14–5). More recently, distributed source volume-conducting medium around the brain
models such as low resolution electromagnetic (generally less than the contribution of intra-
tomography (LORETA) have been proposed cellular currents). These magnetic fields are
and have been used to estimate the loca- extremely small, typically in the femtotesla or
tion of generators of some EEG waveforms.22 picotesla range (10−15 to 10−12 T). They must be
LORETA utilizes a distributed grid of dipole detected by a magnetic gradiometer connected
sources, with considerably more dipole sources to a special type of extremely sensitive ampli-
than recording electrodes, as opposed to sin- fier called a superconducting quantum interfer-
gle dipole models.22 In addition, because of the ence device (SQUID), which must be cooled
reduced complexity of linear source methods by liquid helium. To eliminate noise signals
it is now common to use realistic, patient spe- caused by the much larger magnetic fields asso-
cific, models of the head and brain based on ciated with electrical equipment, power lines,
MRI.23, 24 With source localization methods it is and the earth’s magnetic field, a special mag-
now possible to combine the temporal resolu- netically shielded room is required. For all of
tion of EEG with the spatial resolution of other these reasons, MEG is a very expensive tool.
imaging modalities, such as MRI and SPECT. Another disadvantage of MEG, compared with
The integration of these imaging modalities has EEG, is that it cannot be used readily for the
shown promise as a clinical tool for localization long-term recordings needed to capture and
of epileptogenic brain.25, 26 localize an epileptic seizure, because the sub-
ject’s head must be kept immobilized near the
Key Points magnetic gradiometer array during the entire
recording. Until recently, the number of chan-
• With source localization methods it is now nels available in commercial MEG instruments
possible to combine the temporal resolu- was relatively small, although some systems
tion of EEG with the spatial resolution now available have more than 100 channels;
of other imaging modalities, such as MRI the spatial resolution of these devices is quite
and SPECT. good. Because magnetic fields created by a
• The integration of these imaging modal- current source are always oriented along a tan-
ities has shown promise as a clinical tool gent to a circle around the line of current flow,
for localization of epileptogenic brain. MEG is insensitive to radially oriented cur-
rents in cerebral cortex and is sensitive only to
tangential currents, in contrast to EEG, which
MAGNETOENCEPHALOGRAPHY is sensitive to both (although more sensitive
to radial than to tangential currents). Thus,
Magnetoencephalography (MEG) is the record- in practice, MEG recordings are often com-
ing of the small magnetic fields produced by bined with simultaneous conventional EEG
the electric activity of neurons in the brain. recordings.
The major advantage of MEG is in source REFERENCES

localization. The accuracy of localization of
intracranial sources by MEG is not limited by 1. Gardner, A. B., G. A. Worrell, E. Marsh, D. Dlugos,
the smearing effects of the volume-conducting and B. Litt. 2007. Human and automated detec-
tion of high-frequency oscillations in clinical intracra-
medium, especially the poorly conducting nial EEG recordings. Clinical Neurophysiology
skull, on electric potentials, as occurs in EEG, 118(5):1134–43.
because all the tissues between the sources and 2. Lagerlund, T. D. 1991. Montage reformatting and
the magnetic field detectors are transparent digital filtering. In Epilepsy surgery, ed. H. Luders,
318–22. New York: Raven Press.
to magnetic fields. Thus, when a dipole local-
3. Hjorth, B. 1975. An on-line transformation of EEG
ization algorithm is used with MEG data, a scalp potentials into orthogonal source derivations.
simple homogeneous sphere model of the vol- Electroencephalography and Clinical Neurophysiol-
ume conductor is usually sufficient to obtain ogy 39:526–30.
accurate localization of the source dipole.27 In 4. Wallin, G., and E. Stalberg. 1980. Source derivation
in clinical routine EEG. Electroencephalography and
addition to recording the spontaneous MEG, Clinical Neurophysiology 50:282–92.
one can record evoked magnetic fields in 5. Gotman, J. 1983. Measurement of small time
response to visual, auditory, and somatosen- differences between EEG channels: Method and
sory stimuli, and these may also be submitted application to epileptic seizure propagation. Elec-
to dipole localization algorithms to determine troencephalography and Clinical Neurophysiology
56:501–14.
the location of visual, auditory, and somatosen- 6. Babiloni, F., C. Babiloni, L. Fattorini, F. Carducci,
sory cortical areas. This may be used as part of P. Onorati, and A. Urbano. 1995. Performances of sur-
the surgical planning process in patients with face Laplacian estimators: A study of simulated and
tumors or vascular malformations, in whom real scalp potential distributions. Brain Topography
8:35–45.
the sensory cortical areas may be significantly 7. Duffy, F. H., J. L. Burchfiel, and C. T. Lombroso. 1979.
displaced from their usual or expected loca- Brain electrical activity mapping (BEAM): A method
tion. MEG has been used for localization of an for extending the clinical utility of EEG and evoked
epileptic spike focus before performing resec- potential data. Annals of Neurology 5:309–21.
tive surgery for intractable partial epilepsy.28 8. Lemos, M. S., and B. J. Fisch. 1991. The weighted
average reference montage. Electroencephalography
and Clinical Neurophysiology 79:361–70.
9. Nunez, P. L. 1989. Estimation of large scale neocorti-
cal source activity with EEG surface Laplacians. Brain
Key Points Topography 2:141–54.
10. Perrin, F., J. Pernier, O. Bertrand, M. H. Giard, and
• MEG is the recording of the small mag- J. F. Echallier. 1987. Mapping of scalp potentials by
netic fields produced by the electric activ- surface spline interpolation. Electroencephalography
ity of neurons in the brain. and Clinical Neurophysiology 66:75–81.
• The major advantage of MEG is in source 11. Law, S. K., P. L. Nunez, and R. S. Wijesinghe. 1993.
High-resolution EEG using spline generated surface
localization. Laplacians on spherical and ellipsoidal surfaces. IEEE
• MEG has been used for localization of Transactions on Bio-medical Engineering 40:145–53.
an epileptic spike focus before perform- 12. Srinivasan, R., P. L. Nunez, D. M Tucker, R. B. Sil-
ing resective surgery for intractable partial berstein, and P. J. Cadusch. 1996. Spatial sampling
and filtering of EEG with spline laplacians to estimate
epilepsy. cortical potentials. Brain Topography 8:355–66.
13. Perrin, F., J. Pernier, O. Bertrand, and J. F. Echal-
lier. 1989. Spherical splines for scalp potential and
current density mapping. Electroencephalography and
Clinical Neurophysiology 72:184–7.
SUMMARY 14. Lagerlund, T. D., F. W. Sharbrough, N. E. Busacker,
and K. M. Cicora. 1995. Interelectrode coherences
from nearest-neighbor and spherical harmonic expan-
This chapter reviews several quantitative anal- sion computation of laplacian of scalp potential.
ysis techniques that may be applied to digitized Electroencephalography and Clinical Neurophysiol-
EEG data. The technique of MEG is also ogy 95:178–88.
discussed. Many of these techniques were pri- 15. Pascual-Marqui, R. D., S. L. Gonzalez-Andino,
marily used as research tools; but as they have P. A. Valdes-Sosa, and R. Biscay-Lirio. 1988. Cur-
rent source density estimation and interpolation
become more widely available, they are having based on the spherical harmonic Fourier expan-
an increasing effect on EEG interpretation and sion. The International Journal of Neuroscience 43:
diagnosis. 237–49.
16. Lagerlund, T. D., F. W. Sharbrough, and 22. Pascual-Marqui, R. D., C. M. Michel, and D.
N. E. Busacker. 1997. Spatial filtering of multi- Lehmann. 1994. Low resolution electromagnetic
channel electroencephalographic recordings through tomography: A new method for localizing electrical
principal component analysis by singular value decom- activity in the brain. International Journal of Psy-
position. Journal of Clinical Neurophysiology 14: chophysiology 18:49–65.
73–82. 23. Ebersole, J. S. 1999. Non-invasive pre-surgical eval-
17. Hu, S., M. Stead, and G. A. Worrell. 2007. Automatic uation with EEG/MEG source analysis. Electroen-
identification and removal of scalp reference signal cephalography and Clinical Neurophysiology. Supple-
for intracranial EEGs based on independent comment, 50:167–74.
ponent analysis. IEEE Transactions on Bio-medical 24. Phillips, C., M. D. Rugg, and K. J. Friston. 2002.
Engineering 54(9):1560–72. Anatomically informed basis functions for EEG source
18. Nunez, P. L. 1988. Methods to estimate spatial prop- localization: Combining functional and anatomical
erties of dynamic cortical source activity. In Functional constraints. Neuroimage 16(3 Pt 1):678–95.
brain imaging, ed. G. Pfurtscheller, and F. H. Lopes da 25. Worrell, G. A., T. D. Lagerlund, F. W. Sharbrough,
Silva, 3–9. Toronto: Hans Huber Publishers. et al. 2000. Localization of the epileptic focus by low-
19. Gevins, A., J. Le, P. Brickett, B. Reutter, and resolution electromagnetic tomography in patients
J. Desmond. 1991. Seeing through the skull: with a lesion demonstrated by MRI. Brain Topography
Advanced EEGs use MRIs to accurately measure cor- 12(4):273–82.
tical activity from the scalp. Brain Topography 4: 26. Ding, L., G. A. Worrell, T. D. Lagerlund, and B. He.
125–31. 2007. Ictal source analysis: Localization and imag-
20. Salu, Y., L. G. Cohen, D. Rose, S. Sato, C. Kufta, ing of causal interactions in humans. Neuroimage
and M. Hallett. 1990. An improved method 34(2):575–86.
for localizing electric brain dipoles. IEEE 27. Hari, R. 1994. Comment: MEG in the study of
Transactions on Bio-medical Engineering 37: epilepsy. Acta Neurologica Scandinavica Supplemen-
699–705. tum 152:89–90.
21. Thickbroom, G. W., H. D. Davies, W. M. Carroll, 28. Minassian, B. A., H. Otsubo, S. Weiss, I. Elliott, J.
and F. L. Mastaglia. 1986. Averaging, spatio-temporal T. Rutka, and O. C. Snead III. 1999. MEG localiza-
mapping and dipole modelling of focal epileptic tion in pediatric epilepsy surgery: Comparison with
spikes. Electroencephalography and Clinical Neuro- invasive intracranial electroencephalography. Annals
physiology 64:274–7. of Neurology 46:627–33.
Chapter 15
Electroencephalography in the
Surgical Evaluation of Epilepsy
BACKGROUND PRESURGICAL EVALUATION

PRESURGICAL SELECTION AND WITH CONTINUOUS OR
EVALUATION CHRONIC INTRACRANIAL
Clinical Evaluation MONITORING
Imaging Studies Depth Wire Electrodes
Intracarotid Amobarbital Subdural Electrode
Monitoring
ROUTINE EEG IN THE SURGICAL
EVALUATION OF PATIENTS INTRAOPERATIVE
WITH SEIZURES ELECTROCORTICOGRAPHY
PREOPERATIVE VIDEO-EEG SUMMARY
MONITORING
BACKGROUND considered when treating epilepsy patients4 as

even rare seizures impair the ability to operate
Epilepsy is a relatively common neurologi- a motor vehicle.5 As a result of the impact
cal condition with an estimate of more than on QOL, elimination of all disabling seizures
1 million citizens of the United States suffer- should be the goal of any therapy.
ing from partial epilepsy. Unfortunately more Historically, resective epilepsy surgery for
than 40% of individuals with partial epilepsy the person with intractable epilepsy has some-
are thought to be medically refractory to times been considered a treatment of last
antiepileptic drugs (AEDs) and are consid- resort. The time for the PWE to ultimately
ered to have intractable epilepsy.1, 2 In people be considered for epilepsy surgery has some-
with epilepsy (PWE), disabling and intractable times been measured in decades. A random-
complex, partial, and secondarily generalized ized prospective study comparing surgery for
major motor seizures significantly impair their intractable temporal lobe epilepsy with medi-
quality of life (QOL).3 For example, the abil- cal therapy concluded that resective temporal
ity to legally drive has been identified as an lobe surgery is significantly superior to med-
important indicator for QOL and should be ical therapy alone.6 The American Academy
215
of Neurology practice parameter concerning neuropsychological and possibly psychiatric

surgery for temporal lobe epilepsy endorses evaluations, and frequently intracarotid amo-
the benefits of anteriomesial temporal lobe barbital (ICA) testing.19–23 The surgical evalu-
resection over continued medical treatment ation is focused on identifying and resecting
with AEDs in patients with intractable tempo- the epileptic zone, with the ultimate goal of
ral lobe epilepsy.7 The most common type of seizure freedom. Seizure surgery is most effica-
surgery for epilepsy is the focal resection of cious for patients with temporal lobe epilepsy
the suspected seizure focus, with the major- from mesial temporal sclerosis.10, 22 Determin-
ity being temporal lobectomy. Studies have ing that a patient has medically refractory
shown additional benefits of temporal lobec- seizures usually occurs after 6–12 months’ time
tomy including the postoperative improvement or after failing several AEDs.24 Unfortunately,
of behavioral and cognitive functions.8, 9 Long- not all people with partial epilepsy are appro-
term outcomes have been reported to be priate surgical candidates. Major medical or
favorable for anterior temporal lobectomy with psychiatric illnesses may preclude resective
approximately two-thirds of patients remain- surgery, or even a safe presurgical evaluation.1
ing seizure-free after surgery.10 Selecting favor- In reality, the majority of surgical candidates
able temporal lobe epilepsy patients prior are traditionally young adults and children, but
to surgery can improve postoperative out- surgery may be efficacious and indicated in
come.11–13 Health-related QOL after epilepsy healthy, older individuals.1
surgery may worsen if the patient has con-
tinued postoperative seizures and suffered Key Points
from an associated memory decline due to • Surgical evaluation of patients with epile-
the surgery. If seizures are well controlled
psy is focused on identifying and resecting
postoperatively, the patients’ QOL improved
the epileptic zone.
despite any associated memory difficulties.14 • Seizure surgery is most efficacious for
Temporal lobe epilepsy surgery has also
patients with temporal lobe epilepsy from
been shown to reduce overall health care
mesial temporal sclerosis.
costs,15 and improve the employment status.16
When seizures are controlled after surgery,
outcomes are perceived by the patients as
being improved.17, 18 The preoperative elec- Imaging Studies
troencephalographic (EEG) evaluation is done
to localize and characterize seizures and is Imaging studies, such as specialized MRI with
essential in the preoperative planning process. temporal lobe volumes or high-resolution cuts
through the temporal lobes, are useful for
Purpose and Role of Electroencephalogra- identifying lesions that may be associated with
phy in the Surgical Evaluation of Epilepsy epileptogenic brain tissue.11, 12, 25, 26 Identifying
a lesion that is concordant with the epileptic
• To assist in further localization of epilepto-
zone on EEG is a predictor of good surgical
genic zones using continuous intracranial outcome.12, 13 Brain lesions that are typically
monitoring with depth and grid leads. associated with good epilepsy surgical out-
comes include mesial temporal sclerosis, cor-
tical maldevelopment or dysgenesis, vascular
PRESURGICAL SELECTION AND malformations, and neoplasms.
Ictal single photon emission computed
EVALUATION tomography (SPECT) scan is performed dur-
ing a seizure when the nuclear material is
Clinical Evaluation injected intravenously and the SPECT is per-
formed within a few hours to look for an area
Determining the appropriate epilepsy surgi- of increased blood flow corresponding with
cal candidate requires a team approach and the ictal zones. The sensitivity of the SPECT
generally occurs at a comprehensive epilepsy can be enhanced by performing an interictal
center that is able to provide a multidisci- study and then subtracting the interictal from
plinary approach to the evaluation. The eval- the ictal study. Subtraction of the background
uation includes specialized imaging studies, activity during this process may reveal the
EEG in the Surgical Evaluation of Epilepsy 217
ictal onset zone. After subtraction, the result- • Ictal SPECT scan is a nuclear study per-
ing signal is coregistered on the MRI, and formed during a seizure to assess for an
has been named SISCOM. A positive test has area of increased blood flow correspond-
been shown to be helpful in the localization ing with the ictal zones.
process.27–32 • PET scan utilizes FDG and may reveal
Positron emission tomography (PET) scan areas of hypometabolism that may corre-
may also be useful during the presurgical spond with the ictal zone.
evaluation. PET typically utilizes fluoro-deoxy-
glucose (FDG) and is performed interictally. A
positive test reveals areas of hypometabolism Intracarotid Amobarbital
that may correspond with the ictal zone. An
ambulatory EEG is often done during the PET Intracarotid amobarbital (ICA) or Wada test-
scan to verify that the study is performed during is sometimes performed in the assess-
ing the interictal state. ment of epilepsy patients being considered for
surgery.33, 34 The test is performed by selec-
tively injecting each carotid artery with sodium
Key Points amobarbital to anesthetize areas served by the
particular carotid artery. The basic premise of
• Identifying a lesion on imaging studies the test is that after the injection, the areas that
that is concordant with the epileptic zone are anesthetized have their normal function
on EEG is a predictor of good surgical impaired. During this brief period of anes-
outcome. thesia, the patients’ language and memory are
Sodium amytal injection
Fp1-F7
F7-T3
T3-T5
T5-O1
Fp1-F3
F3-C3
C3-P3
P3-O1
Fp2-F8
F8-T4
T4-T6
T6-O2
Fp2-F4
F4-C4
C4-P4
P4-O2
Monitoring at sodium amytal injection
Sens = 7 μV/mm; HF = 15 Hz; LF = 1.6 Hz Right arm down
Calibration signal 1 sec, 70 μV; Paper speed = 33 mm/sec
Figure 15–1. Selective injection of the left carotid artery with sodium Amytal. Amytal injection corresponds
with left hemisphere slowing, and concurrent weakness in the right arm is clinically observed. HFF = 35 Hz,
LFF = 1.0 Hz, Sensitivity = 7 μV/mm.
evaluated.35–40 Data gathered during this time be distant from the scalp-recording site, the
is used to help determine language lateraliza- recording time may be limited, and technical
tion and memory support. The test may predict issues of signal attenuation by scalp bone and
language and memory function after temporal dura may occur.46, 48, 49
lobectomy, which is particularly important in
dominant hemisphere resective surgery.
During the ICA procedure, the EEG is Key Points
used to ensure that no subtle seizures have • Routine EEG is the first step in the
occurred during the test, verify that the EEG evaluation of the person with suspected
changes are limited to the side of the injec- seizures and intractable partial epilepsy.
tion, and ensure that the electrical activity • Standard activation procedures such as
returns to baseline postinjection. Figure 15–1 photic stimulation, hyperventilation, and
shows typical EEG changes during a left-sided sleep deprivation may improve the diag-
carotid injection.36 Limitations of the ICA test nostic yield of the routine interictal EEG.
include accurate ability in predicting postoper- • Epileptiform activity may not be recorded
ative memory function.38, 39, 41 Functional MRI on routine EEG since the epileptic source
testing (fMRI) may be a potential alternative may be distant from the scalp-recording
to ICA, especially in determining language lat- site, the recording time may be limited,
eralization.42, 43 The utility of alternative testing and technical issues of signal attenuation
to the ICA test, including fMRI, is currently by scalp bone and dura may occur.
unclear.44
Key Points
PREOPERATIVE VIDEO-EEG
• ICA or Wada testing is used to help deter-
mine language lateralization and memory
MONITORING
support.
• During the ICA procedure, the EEG is Video-EEG monitoring in an epilepsy moni-
toring unit (EMU) is used to characterize and
used to ensure that no subtle seizures have
localize the onset of typical events. The ictal
occurred during the test, verify that the
scalp recordings can help to localize the epilep-
EEG changes are limited to the side of the
tiform focus, but also may lead to false localiza-
injection, and ensure that the electrical
tion of the ictal zone, especially when seizures
activity returns to baseline postinjection.
begin in the amygdala, which is at a relatively
long distance from the scalp-recording elec-
trodes.21, 48–50 Seizures beginning in the tempo-
ROUTINE EEG IN THE SURGICAL ral lobe have the highest sensitivity and speci-
EVALUATION OF PATIENTS ficity for localization using scalp recordings.
WITH SEIZURES Another misleading false negative result may
occur when recording simple partial seizures,
The use of routine EEG is the first step in in which ictal scalp recordings may show no
the evaluation of the person with suspected discernable ictal change.51 Foci in the frontal
seizures and intractable partial epilepsy.21, 25, 45 lobe may also not be recorded, since many of
The routine, interictal EEG may provide clues its cortical generators of EEG are located at
to localization and seizure type. Utilizing stan- considerable distances for the scalp-recording
dard activation procedures such as photic stim- electrodes.51
ulation, hyperventilation, and sleep deprivation The placement of supplementary electrodes
may improve the diagnostic yield of the routine may provide a means of improving the yield of
interictal EEG.45–47 Normal interictal EEGs surface recordings. The use of inferior lateral
may be recorded in patients with epilepsy and, temporal leads and so-called 10% (additional
therefore, performing multiple routine EEGs closer spaced) electrodes may help to further
may be helpful in increasing the likelihood of localize ictal and interictal discharges.21, 49, 50
recording abnormalities with each subsequent Sphenoidal electrodes have also been widely
test. Epileptiform activity may not be recorded used in the past during the surgical evaluation.
on routine EEG since the epileptic source may These thin electrodes are placed near the base
of the sphenoid bones bilaterally with a guid- to the EMU or the ICU and monitored in a
ance insertion needle. The utility of sphenoidal continuous manner until typical seizures are
electrodes has been questioned, especially with recorded.53–56 The length of time that monitor-
their invasive nature being less well tolerated ing occurs may vary from a few days to weeks,
than comparable surface leads.50 Nasopharyn- and is mostly based on the number and nature
geal leads inserted in each nostril and left in of seizures recorded. The type of electrodes
place for extended periods were felt to help utilized is determined by the suspected loca-
record discharges from the medial temporal tion of the ictal zone and if brain mapping is
structures. The use of these electrodes has needed.
been shown to be of no more benefit than The main potential complications of intracra-
additional lateral low temporal leads in the nial monitoring include hemorrhage and infec-
presurgical evaluation.52 The use of computer- tion, which occur at a frequency of approx-
ized aids such as seizure- and spike-detection imately 3%. Certain variables have been
programs in the evaluation process may be shown to be important in determining the
useful in analyzing the vast amounts of data rate of complications. The complication rate
gathered in the long-term video-monitoring increases with increasing numbers of elec-
unit. Various strategies are employed by these trodes inserted; using more than 60 electrodes
programs to detect epileptiform activity but leads to more frequent complications.57
their use is an adjunct to the interpretative The two most common types of electrodes
process. used for chronic intracranial monitoring are
depth wire electrodes and grid electrodes.
The usefulness of intracranial EEG in the
Key Points surgery evaluation process was recently con-
• Video-EEG monitoring in an epilepsy firmed, especially in patients when there is a
monitoring unit is used to characterize lack of congruence between EEG and MRI
and localize the onset of typical events. findings.58
• Seizures beginning in the temporal lobe
have the highest sensitivity and specificity
for localization using scalp recordings. Depth Wire Electrodes
• The placement of supplementary elec-
trodes, such as inferior lateral temporal Depth wire electrodes are thin, flexible elec-
leads or sphenoidal electrodes, may pro- trodes that are inserted in the parenchyma of
vide a means of improving the yield of the brain in areas of interest. These electrodes
surface recordings. are commonly used when there is ambiguity
as to lateralization of the seizure onset, typi-
cally when there is a concern about possible
bitemporal seizures. When comparing scalp
PRESURGICAL EVALUATION vs. depth leads, discordance is found in only
WITH CONTINUOUS OR 13% of cases when considering all available
CHRONIC INTRACRANIAL information from the results of the initial com-
MONITORING prehensive epilepsy evaluation utilizing imag-
ing and neuropsychological data. When scalp
Even after the initial evaluation in patients with recordings produce ambiguous results, the use
epilepsy, some of the patients may not immedi- of depth electrodes may demonstrate that
ately qualify for resective surgery due to inad- the patient indeed is a surgical candidate.53–55
equate lesion localization. When this occurs, if Depth wire recordings have distinct advan-
the patient is still considered a surgical candi- tages over scalp leads as there is no movement
date intracranial electrodes may be required. or muscle artifact when utilizing depth leads.55
The location of the indwelling electrodes is Typical placement of a left temporal depth
predetermined and is tailored to the individual lead is depicted in Figure 15–2. The use of
patient based on data collected during initial depth wires may be combined with the use
studies, including scalp recordings, ICA test- of grid electrodes in an attempt to improve
ing, and specialized imaging studies. Once the localization of the ictal zone. Interictal spik-
electrodes are placed, the patient is transferred ing recorded with the depth leads cannot be
may be more frequent from the contralateral

hemisphere. Frequent contralateral interictal
spiking does not necessarily predict a poor out-
come in temporal lobectomy.26, 55, 56 Recording
the patient’s typical seizure with depth leads in
place is often the ultimate determining factor
for localization of the site of seizure onset irre-
spective of seizure onset.26 If seizures occur in
an independent manner from opposite hemi-
spheres, the patient is likely not a good resec-
tion candidate. The number of typical seizures
needed to be recorded utilizing depth elec-
trodes to adequately determine if a patient
is a good surgical candidate ranges from five
to ten.25
Figure 15–2. MRI scan highlighting left temporal depth When considering the electrographic seizure
lead placement used in recording the ictal activity depicted onset, the type of discharge is important in
in Figure 15–3. (Note: Depth leads in the opposite tem- prognostic value. The outcome after surgery is
poral lobes and bilateral frontal lobes are not depicted in more favorable if the onset is more focal and
this image).
precise, as recorded from a single electrode
and then spreading, compared to a seizure
relied upon totally to help determine the ictal onset that is more regional.59 Figure 15–3
zone as up to 20% of cases interictal spikes shows a localized onset seizure beginning in
Seizure onset
1–2 Left
2–3 Temp
3–4 Depth
9–10 Right
10–11 Temp
11–12 Depth
Focal depth seizure onset

Sens = 10 μV/mm; HF = 120 Hz; LF = 1.6 Hz
Figure 15–3. Typical ictal activity in the deepest contact in the left temporal depth lead (fast activity in trace 1). Ictal
activity evolves rapidly to adjacent leads as the seizure progresses. HFF = 120 Hz, LFF = 1.0 Hz, Sensitivity = 75 μV/mm.
a temporal depth lead. The amount of time As with depth wire electrodes, the number
that it takes the electrographic seizure activ- of electrodes implanted impacts the complica-
ity to spread has also been shown to pre- tion rate.57 The appropriate grid or depth wire
dict outcome. If the electrographic seizure candidate must be able to tolerate the intracra-
activity propagates more slowly to the oppo- nial electrodes being in place for a period of
site hemisphere, the surgical outcomes are days to a couple of weeks.
better.25 Subdural grid electrodes provide a unique
opportunity to perform cortical mapping to
determine the presence and location of elo-
quent cortex prior to surgical resection of the
Subdural Electrode Monitoring ictal zone. This localization is accomplished by
using the individual electrodes of the grid that
Multiple subdural electrodes embedded at
are resting upon neocortex to electrically stim-
fixed distances in a flat flexible plastic back-
ulate the ictal and periictal cortex looking for
ground material (grid) are utilized for neocor-
vital functions, including language, vision, and
tical recordings. Similar to the use of depth
motor and sensory function. These studies are
leads, subdural grid monitoring is accom-
relatively easily evaluated with the grid in place
plished after the initial scalp EEG and other
during the monitoring session in the EMU.
tests have identified a preliminary target for
Stimulation is accomplished with the use of an
resection. Grid placement is performed under
external stimulator that applies a small elec-
general anesthesia, following which the patient
trical stimulus in either a bipolar or a unipo-
is monitored in the EMU. The number of elec-
lar strategy utilizing the recording electrodes.
trodes and their location is determined by the
Electrodes overlying the suspected ictal zone
preliminary results of the initial studies and
or eloquent cortex are particularly important.
the anatomical location of the suspected ictal
During the stimulation procedure, the patient
zone. When evaluating the mesiotemporal lobe
performs specific and appropriate tasks for the
structures, depth electrodes placed in the tem-
area in question. With each stimulation, the
poral lobes may be more sensitive in detecting
patient is observed or self-reports changes in
epileptiform activity.59 Figure 15–4 depicts a
the ability to perform those specific tasks dur-
typical grid placement and Figure 15–5 shows
ing the stimulation. If the ictal zone overlaps
a well-localized seizure onset.
Figure 15–4. Skull X-Ray showing a combination of frontal grid electrodes and temporal strip electrodes used to localize
the ictal zone.
Seizure onset
Focal grid seizure

Sens = 75 μV/mm; HF = 120 Hz; LF = 1.6 Hz
Figure 15–5. EEG utilizing neocortical grid in the frontal lobe showing a well-localized seizure onset (arrow). Note the
rapid spread to adjacent electrodes.
with eloquent cortex, resection may not be • The use of depth wires may be com-
possible.60 bined with the use of grid electrodes in
Surgical outcomes for people undergoing an attempt to improve localization of the
chronic intracranial monitoring are good, with ictal zone.
an approximately 80% seizure-free rate if there • The outcome after surgery is more favor-
is an associated MRI lesion that is concordant able if the seizure onset is more focal and
with the ictal zone. Poorer outcomes are noted precise, as recorded from a single elec-
in nonlesional cases with less than 25% being trode and then spreading, compared to a
seizure-free after surgery.26 seizure onset that is more regional.
• Multiple subdural electrodes embedded
Key Points at fixed distances in a flat flexible plastic
background material (grid) are utilized for
• Depth wire electrodes inserted in the neocortical recordings.
parenchyma of the brain in areas of inter- • Subdural grid electrodes provide a unique
est may be used when there is ambiguity opportunity to perform cortical mapping
as to lateralization of the seizure onset, to determine the presence and location of
typically when there is a concern about eloquent cortex prior to surgical resection
possible bitemporal seizures. of the ictal zone.
INTRAOPERATIVE A baseline extracranial EEG recording is per-

ELECTROCORTICOGRAPHY formed before the surgical resection with the
patient awake, and a contralateral ear ref-
Intraoperative electrocorticography (ECoG) is erence is used for the intracranial record-
a technique that involves a multidisciplinary ings. Electrical stimulation is used intraop-
epilepsy team, including the epileptologist, eratively to assist in localizing the epilepto-
neurosurgeon, EEG technologist, anesthesiol- genic zone and functional anatomy. Samples
ogist and possibly neuropsychologist assem- of EEG activity are obtained before and after
bling in the operating room (OR) during cortical resection and interictal epileptiform
the resective procedure. The need for ECoG activity is recorded and localized (Fig. 15–6).
depends on the results of the presurgical The cortex producing epileptiform activity is
evaluation, the localization of the epilepto- then resected and the electrodes are rein-
genic zone, and the type(s) of extraoperative serted along the margin of the resection. The
EEG monitoring. If the epileptogenic zone extent of the focal corticectomy is based on
has been localized with continuous intracra- the ECoG and the presurgical evaluation. The
nial monitoring, ECoG may be less important relationship between the localization of the
in determining the region of cortical resec- epileptogenic zone and the eloquent cortex is
tion. Patients in whom depth electrode studies considered at the time of the operation. A pos-
show a mesiobasal limbic origin of seizures texcision recording is made, with a subdural
may subsequently have en bloc resection of the strip electrode placed posterior to the margin
temporal lobe without ECoG. Traditional focal of the surgical resection (Fig. 15–7). If further
corticectomies have used a tailored resection, epileptiform activity is noted further resection
with ECoG localizing the site of seizure onset. is attempted. The limit of the resection may
When performed at Mayo Clinic, ECoG depend on functional cortical mapping that is
is performed with the patient under general also accomplished in the OR.
anesthesia. The anesthetic agents used include Electrocorticography has several important
nitrous oxide, fentanyl, or a low-volume per- limitations that must be recognized for appro-
centage of isoflurane. A 16- or 32-channel priate interpretation of these studies. Informa-
EEG recording is made. Methohexital given tion about the localization and lateralization
intravenously may be used to enhance or to of the epileptogenic zone must be determined
activate epileptiform activity in the preexci- prior to the ECoG study. Electrocorticogra-
sion recordings. The strategy for intraoperative phy has been used mainly to assess neocortical
EEG monitoring is discussed with the surgi- epileptiform activity, and it may be restricted in
cal team, the anesthesiologist, and the EEG sampling from the orbitofrontal, mesiofrontal,
technologist. The electroencephalographer is and mesiotemporal regions, which may not cor-
present in the OR for immediate interpretation relate with the region of seizure onset. Other
of the EEG data. potential disadvantages of ECoG include the
Several techniques for ECoG are available sampling of predominant interictal EEG activ-
that include the use of subdural strip elec- ity, the restricted spatial distribution of the
trodes or a rigid electrode holder with graphite EEG recording, and usually the brief dura-
tip or cotton wick electrodes.61 After cran- tion of the monitoring. The neurosurgical team
iotomy, electrodes are placed on the cortex must delay the cortical resection until ECoG
and ECoG is recorded while the patient is is performed and the studies are interpreted.
under light general anesthesia. Subdural strip In addition, general anesthesia used during the
electrodes are particularly useful in record- surgical procedure may suppress epileptiform
ings obtained from the inferior temporal lobe activity. High concentrations of certain anes-
and the mesiofrontal region.61 The preexci- thetic drugs, such as isoflurane, may make it
sion ECoG at Mayo Clinic uses three subdu- difficult to document the neocortical extent of
ral strips (each with eight electrode contacts) the epileptogenic zone. Enflurane may acti-
placed on the superior and inferior temporal vate or increase interictal epileptiform activ-
lobe gyri and in the suprasylvian region. Three ity; however, this may be nonspecific and may
depth electrodes are also placed by hand in not correlate with the site of seizure onset.
the region of the amygdala and hippocampus. The false positivity of methohexital-activated
Pre excision ECoG
F-Age: 40 1–25–90
LA-A2
LAH-A2
LPH-A2
IT1-A2
IT2-A2
IT3-A3
IT4-A2
IT5-A2
IT6-A2
IT7-A2
IT8-A2
100 μV
1 sec
Figure 15–6. Electrocorticography performed at the time of left anterior temporal lobectomy. The three upper channels
represent recording from the mesiotemporal region with depth electrodes. Prominent spiking is noted in the mesiotemporal
region and the lateral temporal cortex.
Post excision ECoG

F-Age: 40 1–25–90
PT1-A2
PT2-A2
PT3-A2
PT4-A2
PT5-A2
PT6-A2
PT7-A2
PT8-A2
100 μV
1 sec
Figure 15–7. Postexcision ECoG performed with a subdural strip placed posterior to the margin of the resection. No
definite residual spiking is noted.
224
ECoG has also been observed. Certain ECoG- that essential elements of the evaluation are
recorded postexcision spike discharges are not in agreement before surgery is offered. The
prognostically important.62 use of specialized tests and sometimes invasive
The use of intraoperative ECoG has been EEG recording techniques can assist in localiz-
found to be of little predictive value for a ing the ictal zone for possible surgical resection
good outcome after resective surgery.11, 63 The when standard monitoring techniques are not
presence of residual spiking may be associated conclusively localizing. Epilepsy surgery has
statistically with unfavorable seizure outcome. been shown to be effective in the treatment of
However, the presence of residual spiking may intractable epilepsy, especially in patients with
not preclude a successful surgical outcome temporal lobe epilepsy. Future technology is
after focal corticectomy.64 currently being studied that may dramatically
change how we manage and evaluate people
Key Points with intractable epilepsy.
• During electrocorticography, electrodes
are placed on the cortex and the activity is
recorded while the patient is under light REFERENCES
general anesthesia.
• The cortex producing epileptiform activ- 1. Kwan, P., and M. J. Brodie. 2000. Early identification
ity is surgically resected and the electrodes of refractory epilepsy. The New England Journal of
are reinserted along the margin of the Medicine 342:314–19.
2. French, J. A., and S. Schachter. 2002. A workshop on
resection. antiepileptic drug monotherapy indications. Epilepsia
43(Suppl 10):3–27.
3. Fisher, R. S. 2000. Epilepsy from the patient’s perspec-
tive: Review of results of a community-based survey.
SUMMARY Epilepsy & Behavior 1:S9–14.
4. Gilliam, F., R. Kuzniecky, E. Faught, L. Black, G. Car-
As technology advances, the availability and penter, and R. Schrodt. 1997. Patient-validated con-
use of minute microprocessors may change tent of epilepsy-specific quality-of-life measurement.
the way we evaluate the patient with epilepsy. Epilepsia 38:233–6.
5. Drazkowski, J. F., R. S. Fisher, J. I. Sirven, et al.
Currently trials are ongoing using permanently 2003. Seizure-related motor vehicle crashes in Arizona
implanted electronic devices that are capable before and after reducing the driving restriction from
of recording limited but chronic EEG from 12 to 3 months. Mayo Clinic Proceedings 78:819–25.
depth or grid leads.65 The device has been 6. Wiebe, S., W. T. Blume, J. P. Girvin, and M. Eliasziw.
2001. Effectiveness and efficiency of surgery for tem-
shown to be able to analyze ongoing EEG and poral lobe epilepsy study group. A randomized, con-
detect the onset of seizures by various algo- trolled trial of surgery for temporal-lobe epilepsy. The
rithms and in response provide a small electri- New England Journal of Medicine 345:311–8.
cal shock in an attempt to abort the ictal activ- 7. Engel, J. Jr., S. Wiebe, J. French, et al. 2003. Qual-
ity, while other permanent devices stimulate at ity Standards Subcommittee of the American Academy
of Neurology; American Epilepsy Society; American
scheduled intervals.66 Trending data, brief sam- Association of Neurological Surgeons. Practice param-
ples, and seizures may be easily downloaded eter: Temporal lobe and localized neocortical resec-
for analysis via the Internet. The ability to tions for epilepsy: Report of the Quality Standards
safely record and detect seizures on a chronic Subcommittee of the American Academy of Neurol-
ogy, in association with the American Epilepsy Society
basis may change future treatment strategies. and the American Association of Neurological Sur-
The current study and use of such devices is geons. Neurology 60:538–47.
limited to patients felt not to be resective sur- 8. Radhakrishnan, K., E. L. So, P. L. Silbert, et al. 1998.
gical candidates and are able to be participants Predicators of outcome of anterior temporal lobec-
in ongoing clinical studies. tomy for intractable epilepsy: A multivariate study.
Neurology 51:465–71.
The use of EEG in the evaluation of peo- 9. National Institutes of Health Consensus Conference.
ple with refractory partial epilepsy for possi- 1990. Surgery for epilepsy. JAMA 264:729–33.
ble resective surgery remains integral to the 10. Walczak, T. S., R. A. Radtke, J. O. McNamara, et al.
comprehensive epilepsy evaluation. Surgery 1990. Anterior temporal lobectomy for complex partial
seizures: Evaluation, results, and long-term follow-up
remains a potentially curative procedure for in 100 cases. Neurology 40:413–18.
people with intractable epilepsy. Using a multi- 11. Cascino, G. D., M. R. Trenerry, C. R. Jace Jr.,
disciplinary team to evaluate the PWE requires et al. 1995. Electrocorticography and temporal lobe
epilepsy: Relationship to quantitative MRI and opera- 29. So, E. L. 2000. Integration of EEG, MRI, and SPECT
tive outcome. Epilepsia 36:692–6. in localizing the seizure focus for epilepsy surgery.
12. Cascino, G. D., M. R. Trenerry, E. L. So, et al. 1996. Epilepsia 41(Suppl 3):S48–54.
Routine EEG and temporal lobe epilepsy: Relation 30. Bouilleret, V., M. P. Valenti, E. Hirsch, F. Semah, and
to long-term EEG monitoring, quantitative MRI, and I. J. Namer. 2000. Correlation between PET and SIS-
operative outcome. Epilepsia 37:651–6. COM in temporal lobe epilepsy. Journal of Nuclear
13. Cascino, G. D. 1998. Neuroimaging in partial epilepsy: Medicine 48:991–8.
Structural magnetic resonance imaging. Journal of 31. Wiest, R., J. Kassubek, K. Schindler, et al. 2005. Com-
Epilepsy 11:121–9. parison of voxel-based 3-D MRI analysis and subtrac-
14. Langfitt, J. T., M. Westerveld, M. J. Hamberger, tion ictal SPECT coregistered to MRI in focal epilepsy.
et al. 2007. Worsening of quality of life after epilepsy Epilepsy Research 65:125–33.
surgery: Effect of seizures and memory decline. Neu- 32. Whiting, P., R. Gupta, J. Burch, et al. 2006. A systemic
rology 68:1988–94. review of the effectiveness and cost-effectiveness of
15. Langfit, J. T., R. G. Holloway, M. P. McDermott, neuroimaging assessments used to visualize the seizure
et al. 2007. Healthcare costs decline after successful focus in people with refractory epilepsy being con-
epilepsy surgery. Neurology 68:1290–8. sidered for surgery. Health Technology Assessment
16. Chin, P. S., A. T. Berg, S. S. Spencer, et al. 2007. 10:iii–iv, 1–250.
Employment outcomes following resective epilepsy 33. Gardner, W. J. 1941. Injection of procaine into
surgery. Epilepsia 48(12):2253–7. the brain to relocate speech area in left-handed
17. Spencer, S. S., A. T. Berg, B. G. Vickrey, et al. persons. Archives of Neurology and Psychiatry
2007. Multicenter study of epilepsy surgery. Health- 46:1035–8.
related quality of life over time since resective epilepsy 34. Wada, J. 1949. A new method for determination of
surgery. Annals of Neurology 62(4):327–34. the site of cerebral speech dominance: A preliminary
18. Gilliam, F., R. Kuzniecky, K. Meador, et al. 1999. report on the intracarotid injection of sodium Amytal
Patient-oriented outcome assessment after tempo- [in Japanese]. Igaku to Seitbutsugaki (Medicine and
ral lobectomy for refractory epilepsy. Neurology Biology) 14:221–2.
53:687–94. 35. Wyllie, E., H. Lüders, D. Murphy, et al. 1990. Intrac-
19. Salinsky, M., R. Kanter, and R. M. Dasheiff. 1987. arotid amobarbital (Wada) test for language domi-
Effectiveness of multiple EEGs in supporting the nance: Correlation with results of cortical stimulation.
diagnosis of epilepsy: An operational curve. Epilepsia Epilepsia 31:156–61.
28:331–4. 36. Morrell, M. J., C. A. Phillips, M. J. O’Connor, and
20. Jack, C. R. Jr., W. H. Theodore, M. Cook, and G. M. R. Sperling. 1991. Speech during partial seizures:
McCarthy. 1995. MRI-based hippocampal volumet- Intracranial EEG correlates. Epilepsia 32:886–9.
rics: Data acquisition, normal ranges, and optimal 37. Wada, J., and T. Rasumssen. 1960. Intracarotid injec-
protocol. Magnetic Resonance Imaging 13:1057–64. tion of sodium Amytal for the lateralization of cerebral
21. Quency, L. F., M. W. Risinger, and D. A. Shewmon. speech dominance: Experimental and clinical observa-
1993. Extracranial EEG evaluation. In Surgical treat- tions. Journal of Neurosurgery 17:266–82.
ment of the epilepsies, ed. J. Engel Jr., 2nd ed., 173–95. 38. Kneebone, A. C., G. J. Chelene, D. S. Dinner, R.
New York: Raven Press. I. Naugle, and I. A. Awad. 1995. Intracarotid amo-
22. O’Brien, T. J., E. L. So, B. P. Mullan, et al. 2000. Sub- barbital procedure as a predictor of material-specific
traction peri-ictal SPECT is predictive of extratempo- memory change after anterior temporal lobectomy.
ral epilepsy surgery outcome. Neurology 55:1668–77. Epilepsia 36:857–65.
23. Theodore, W. H., S. Sato, C. V. Kufta, W. D. Gaillard, 39. Loring, D. W., K. J. Meador, G. P. Lee, et al. 1995.
and K. Kelley. 1997. FDG-positron emission tomog- Wada memory asymmetries predict verbal memory
raphy and invasive EEG: Seizure focus detection and decline after anterior temporal lobectomy. Neurology
surgical outcome. Epilepsia 38:81–6. 45:1329–33.
24. Engel, J., Jr., and G. A. Ojemann. 1993. The next step. 40. Loring, D. W., B. P. Hermann, K. J. Meador, et al.
In Surgical treatment of the epilepsies, ed. J. Engel Jr., 1994. Amnesia after unilateral temporal lobectomy: A
2nd ed., 319–29. New York: Raven Press. case report. Epilepsia 35:757–63.
25. Gates, J. R. 1987. Epilepsy presurgical evaluation 41. Loring, D. W., G. P. Lee, K. J. Meador, et al. 1990.
in the era of intensive neurodiagnostic monitoring. The intracarotid amobarbital procedure as a predic-
Advances in Neurology 46:227–47. tor of memory failure following unilateral temporal
26. Schiller, Y., G. D. Cascino, and F. W. Sharbrough. lobectomy. Neurology 40:605–10.
1998. Chronic intracranial EEG monitoring for local- 42. Klöppel, S., and C. Büchel. 2005. Alternatives to the
izing the epileptogenic zone: An electroclinical corre- Wada test: A critical view of functional magnetic res-
lation. Epilepsia 39:1302–8. onance imaging in preoperative use. Current Opinion
27. O’Brien, T. J., E. L. So, B. P. Mullan, et al. 1998. Sub- in Neurology 18:418–23.
traction ictal SPECT co-registered to MRI improves 43. Grote, C. L., and K. Meador. 2005. Has amobar-
clinical usefulness of SPECT in localizing the surgical bital expired? Considering the future of the Wada.
seizure focus. Neurology 50:445–54. Neurology 65:1692–3.
28. O’Brien, T. J., E. L. So, B. P. Mullan, et al. 1999. 44. Abou-Khalil, B. 2007. An update on determination of
Subtraction SPECT co-registered to MRI improves language dominance in screening for epilepsy surgery:
postictal SPECT localization of seizure foci. Neurology The Wada test and newer noninvasive alternatives.
52:137–46. Epilepsia 48:442–55.
45. Niedermeyer, E., and F. Lopes da Silva, eds. electroencephalography in 32 patients with refractory
1987. Electroencephalography: Basic principles, clin- epilepsy. Annals of Neurology 12:248–53.
ical applications, and related fields, 2nd ed. Baltimore: 56. Schiller, Y., G. D. Cascino, N. E. Busacker, and F.
Urban & Schwarzenberg. W. Sharbrough. 1998. Characterization and compar-
46. Daly, D. D. 1990. Epilepsy and syncope. In Current ison of local onset and remote propagated electro-
practice of clinical electroencephalography, ed. D. D. graphic seizures recorded with intracranial electrodes.
Daly, and T. A. Pedley, 2nd ed., 269–334. New York: Epilepsia 39:380–8.
Raven Press. 57. Hamer, H. M., H. H. Morris, E. J. Mascha, et al. 2002.
47. Blum, D. 1994. Prevalence of bilateral partial Complications of invasive video-EEG monitoring with
seizure foci and implication for electroencephalo- subdural grid electrodes. Neurology 58:97–103.
graphic telemetry monitoring and epilepsy surgery. 58. Pondal-Sordo, M., D. Diosy, E. E. Téllez-Zenteno, R.
Electroencephalography and Clinical Neurophysiol- Sahjpaul, and S. Wiebe. 2007. Usefulness of intracra-
ogy 91:329–36. nial EEG in the decision process for epilepsy surgery.
48. Quesney, L. F., M. Constain, D. R. Fish, and T. Epilepsy Research 74:176–82.
Rasmussen. 1990. Frontal lobe epilepsy—A field of 59. Engel, J. Jr., and P. H. Crandall. 1987. Intensive neu-
recent emphasis. American Journal of EEG Technol- rodiagnostic monitoring with intracranial electrodes.
ogy 30:177–93. Advances in Neurology 46:85–106.
49. Quesney, L. F., and P. Gloor. 1985. Localization of 60. Lesser, R. P., H. Lüders, H. H. Morris, et al. 1986.
epileptic foci. Electroencephalography and Clinical Electrical stimulation of Wernicke’s area interferes
Neurophysiology. 37(Suppl):165–200. with comprehension. Neurology 36:658–63.
50. Sharbrough, F. W. 1990. Electrical fields and recording 61. Lesser, R. P., and B. Gordon. 2000. Methodologic con-
techniques. In Current practice of clinical electroen- siderations in cortical electrical stimulation in adults.
cephalography, ed. D. D. Daly, and T. A. Pedley, 2nd In Epileptic seizures: Pathophysiology and clinical
ed., 29–49. New York: Raven Press. semiology, ed. H. O. Lüders, and S. Noachtar, 153–65.
51. Devinsky, O., S. Sato, C. V. Kufta, et al. 1989. New York: Churchill Livingstone.
Electroencephalographic studies of simple partial 62. Engel, J. Jr., and G. A. Ojemann. 1993. The next step.
seizures with subdural electrode recordings. Neurol- In Surgical treatment of the epilepsies, ed. J. Engel Jr.,
ogy 34:527–33. 2nd ed., 319–29. New York: Raven Press.
52. Sperling, M. R., and J. Engel Jr., 1985. Elec- 63. Cascino, G. D., M. R. Trenerry, C. R. Jack Jr.,
troencephalographic recording from the temporal et al. 1995. Electrocorticography and temporal lobe
lobes: A comparison of ear, anterior temporal, and epilepsy: Relationship to quantitative MRI and opera-
nasopharyngeal electrodes. Annals of Neurology tive outcome. Epilepsia 36:692–6.
17:510–13. 64. Fiol, M. E., J. R. Gates, F. Torres, and R. E. Maxwell.
53. So, N., P. Gloor, L. F. Quesney, M. Jones-Gotman, A. 1991. The prognostic value of residual spikes in the
Olivier, and F. Andermann. 1989. Depth electrodes postexcision electrocorticogram after temporal lobec-
investigations in patients with bitemporal epileptiform tomy. Neurology 41:512–16.
abnormalities. Annals of Neurology 25:423–31. 65. Duckrow, R. B., and T. K. Tcheng. 2007. Daily
54. So, N., A. Olivier, F. Andermann, P. Gloor, and L. variation in an intracranial EEG features in humans
F. Quesney. 1989. Results of surgical treatment in detected by a responsive neurostimulator system.
patients with bitemporal epileptiform abnormalities. Epilepsia 48(8):1614–20.
Annals of Neurology 25:432–9. 66. Morrell, M. 2006. Brain stimulation for epilepsy:
55. Spencer, S. S., D. D. Spencer, P. D. Williamson, and Can scheduled or responsive neurostimulation stop
R. H. Mattson. 1982. The localizing value of depth seizures? Current Opinion in Neurology 19(2):164–8.
Chapter 16
Movement-Related Cortical
Potentials and Event-Related
Potentials
Virgilio Gerald H. Evidente and John N. Caviness
MOVEMENT-RELATED CORTICAL Abnormalities in Disease

POTENTIALS Contingent Negative Variation
Introduction
EVENT-RELATED POTENTIALS
Technique
The P300
Normal Waveforms
Individual Variation SUMMARY
MOVEMENT-RELATED CORTICAL they found two more components occurring

POTENTIALS right before movement onset: premovement
positivity (PMP) and motor potential (MP).2 As
for post-movement MRCP components asso-
Introduction ciated with hand movements, Shibasaki et al.
in 1980 identified four different peaks: the
Movement-related cortical potentials (MRCPs) N50, P90, N160, and P300.3 Microcomputers
are electroencephalographic (EEG) potentials and software programs are now available that
that occur around the time of movement facilitate the back-averaging of cortical activity
and are recorded using surface scalp elec- preceding a marked event.
trodes. Slow negative EEG waveforms pre-
ceding self-initiated movements were first
described by Kornhuber and Deecke in 1965.1 Purpose and Role of MRCPs
They recorded EEG and electromyogram
(EMG) simultaneously while making sub- • These potentials have been important in
jects do self-paced repetitive movements and understanding the cortical activity under-
stored the data on magnetic tape. Using this lying normal and abnormal movements;
technique, they identified two premovement their clinical application is limited.
components: the Bereitschaftspotential (BP) • The BP may help identify some functional
and reafferente potentiale (RP). Later on, movement disorders.
229
Technique of time the subject has in order to prepare

for the movement. Approximately 400–650 ms
One of the most challenging tasks in clinical before movement onset, the slope of the neg-
neurophysiology is to obtain MRCPs. A tech- ativity turns more sharply upward. This later
nique termed jerk-locked averaging is used to steeper slope of the BP was differentiated by
obtain and study MRCP. The minimal scalp Shibasaki et al. from the early slower shift
electrode montage should include at least one and was designated the Negative Slope (NS );
to two vertex electrodes (Cz and Fz) and a it terminates approximately 90–100 ms before
pair of lateral electrodes around the area of the movement onset.3 The other terms used to
motor cortex (C3 and C4). The recordings are refer to the NS include NS2 and contralateral
typically referenced to linked ear electrodes. It preponderance of negativity (CPN).
is critical that the electro-oculographic activ- The early component of the BP with a shal-
ity also be recorded. By applying additional lower slope is also referred to as early BP, while
electrodes in grids over sensorimotor cortex, the later segment with a steeper slope as late
the potentials can be mapped topographically. BP (Fig. 16–1).4 The precise identity of neu-
A low-frequency filter with a cutoff in the ral generators of each of the movement-related
range of approximately 0.05 Hz or less must be potentials is a subject of controversy. The early
used to record the slow premovement nega- BP is thought to reflect activity associated with
tive wave. processes of movement preparation and most
A brisk voluntary movement, usually of a likely originates from the supplementary motor
digit or distal limb, acts as the timing event, area (SMA), the pre-supplementary motor area
but the character of the movement must be (pre-SMA), and lateral premotor cortex bilat-
defined. The subject is instructed to keep eyes erally. The late BP is theorized to represent
open (to minimize alpha activity) and fixate on activity related to movement execution at the
a single point (to minimize eye movement arti- primary motor cortex (PMC) and lateral pre-
facts). The subject makes self-paced, repetitive motor cortex. The late BP is maximal over the
movements every 3–10 seconds. The initial rise contralateral central area (approximately C1 or
of rectified EMG activity from the movement C2) for hand movements, and at the midline
triggers data collection. The computer buffer (approximately Cz) for foot movements.
is configured to collect the data for 2–3 sec- The premovement positivity or P50 is pre-
onds before and for 0.5–1 second after the dominant over the ipsilateral hemisphere to
trigger. Data from 100 to 200 such artifact-free the moving hand.2, 3 It was initially proposed
movements are collected and stored for later to be related to suppression of movements of
analysis. the opposite hand in voluntary unilateral hand
Averaging must be performed off-line to movements based on observations that the
ensure proper artifact rejection. Each trace is PMP is not seen with bilateral simultaneous
reevaluated to detect any contamination by eye hand movements.5 However, the PMP is cur-
movement artifact. The tracing must also dis- rently thought to represent a trough between
play a clear EMG takeoff point. If acceptable, two successive negative peaks, and that it may
the tracing is aligned at the EMG onset and actually have no physiological significance.4
computed into the ongoing averaging process. The MP occurs right before the onset of
movement and is localized to a small area in the
contralateral central scalp region correspond-
Normal Waveforms ing to the movement site. The MP was also
named N10 as it occurs on an average of about
The earliest recordable MRCP is the Bere- 10 ms prior to the rectified EMG. This neg-
itschaftspotential or readiness potential, which ative peak most likely represents activity of
is a slowly rising negativity beginning 1.0–2.0 pyramidal tract neurons in the PMC.4
seconds before the onset of movement.1 It is Shibasaki named four post-movement peaks
maximal at the vertex (particularly at the mid- based on the time interval from the peak of
line centro-parietal area), and is symmetrically the averaged, rectified EMG to the peak of
and widely distributed over the scalp regard- the post-movement potential.3 The first post-
less of the site of movement. The onset of movement peak called N50 is a negative peak
the BP can vary depending on the amount localized to the frontal region. The P90 is a
Movement-Related Potentials and Event-Related Potentials 231
Trigger Trigger
Fp1 T3
Fp2 T4
F3 T5
F4 T6
C3 Fz BP
NS
C4 Cz
P3 Pz
P4 FC5
O1 FC6
O2 EOG-V
F7 EOG-H
F8 A1A2-
DME
ACC
Rt TA
– Rt TS
10 μV EMG 30 μV ACC 0.25 G
EEG, EOG
+
500 ms
Figure 16–1. Movement-related potential recorded from multiple scalp locations, with right anterior tibial muscle (TA)
relaxation in a normal subject with no soleus movement (TS). ACC is a simultaneous accelerometer recording. The ini-
tial slow negative phase of the Bereitschaftspotential (early BP) begins 1.7 seconds before movement and is followed by
the steeper negative slope (late BP) 650 ms before movement. Later positivity represents the reafferente potential. Elec-
trooculogram (EOG) shows that no eye movement occurred. (From Terada, K., A. Ikeda, S. Yazawa, T. Nagamine, and
H. Shibasaki. 1999. Movement-related cortical potentials associated with voluntary relaxation of foot muscles. Clinical
Neurophysiology 110:397–403. By permission of Elsevier Science.)
surface-positive waveform that occurs over the et al. noted that in normal subjects, the most
parietal region and is larger over the contralat- reproducible part of the BP was the late com-
eral hemisphere. The N160 is a negative poten- ponent; thus, the late BP would be most useful
tial localized to the contralateral parietal area, in studies quantifying MRCP changes before
and forms a positive–negative complex with and after an intervention.7
P90. The P300 is a positive waveform that cor-
responds to the RP referred to by Kornhuber
and Deecke.1
Abnormalities in Disease
Early small studies reported that the
Individual Variation movement-related potentials in Parkinson’s
disease (PD) were normal.8 Subsequent larger
The BP may be absent in some subjects, possi- studies revealed that the amplitude of the
bly reflecting differences in cortical anatomy. early BP component was reduced in PD
Age-related variations in the premovement patients.9, 10 Similarly, the BP associated with
potentials are minor.5 Although older individ- gait showed decreased activation in PD.11 Dick
uals consistently show slower motor reaction et al pointed out that levodopa modifies BP
times to the onset of motion compared to amplitude.12 It causes an increase in the ampli-
younger subjects, the bulk of slowed response tude of the early part of the BP and of the
appears to arise from slowed motor processes negative peak just before the EMG onset.
rather than from slowness in perceptual pro- Limousin et al., as well as Gironell and col-
cessing or in the readiness potential.6 Evidente leagues, described the effects of pallidotomy
on MRCP in PD patients.13, 14 Post pallidotomy, may be distinguished by the presence of a BP

there was an increase in the NS of the late before an apparently involuntary muscle jerk.28
component of the BP, and no change in the Caution is advised when using the BP to sup-
early component. It was thus suggested that port the diagnosis of a psychogenic jerk, since
pallidotomy improves mainly the later stages of its absence may be due to a difficult recording
movement preparation. situation. Moreover, its absence does not rule
Patients with cerebellar lesions may have out pathology.
abnormalities in MRCP. In Ramsay Hunt cere-
bellar degeneration or in some patients with
Benedikt’s syndrome, no premovement poten- Contingent Negative Variation
tial is noted.15 Those with cortical cerebellar
degeneration, however, usually have normal In the contingent negative variation testing
waveforms. Individuals who suffer from acute paradigm, a stimulus such as a click warns the
cerebellar infarction may have a reduction in subject to prepare to move. After 1–2 seconds,
the late BP phase, consistent with the known a second stimulus, such as a flash of light, sig-
cerebellar projection to the motor cortex.16 nals the patient to begin moving. Contingent
Studies of MRCP in stroke showed variable negative variation is a slow negative poten-
findings related to the locations and sizes of the tial that appears in the interval between the
lesions, particularly whether or not the motor warning stimulus and the second stimulus. The
cortex is involved, and the duration. Most of distribution of this wave is predominately bilat-
these studies have found changes in the late eral and frontal, but it may shift with variations
BP as well as the MP, which usually resolve on in the testing procedure. The neural generators
recovery.17 of contingent negative variation appear to be
Patients suffering from primary dystonia or different from those of the BP.29
from symptomatic dystonia due to lesions of
the basal ganglia or their output pathways Key Points
exhibit with smaller early and late BP com-
ponents.18, 19 Patients with focal hand dysto- • The BP and contingent negative variation
nia have abnormalities in the distribution and are two movement-related cortical poten-
amplitude of the BP that indicate impaired cor- tials, each of which is generated by a
tical activation during voluntary movement and specific paradigm.
relaxation.20, 21 In tardive dyskinesia, the BP is • Abnormalities in the BP have been seen in
increased in amplitude.22 a variety of disorders but are not specific.
Early studies on tics by Obeso et al. showed • Caution is advised when using the BP
that such movements were not preceded by to support the diagnosis of a psy-
a premovement potential, whereas mimicked chogenic jerk.
voluntary jerks were.23 Later studies done by
Karp et al. showed that tics are preceded by
MRCPs, albeit shorter than normal.24 EVENT-RELATED POTENTIALS
Chorea may be technically difficult to do
MRCP studies on because of the random- Whereas standard somatosensory or visual-
ness and unpredictability of the movements. evoked potentials map the cortical response
Shibasaki et al. nevertheless reported that the to a simple sensory stimulus, event-related
chorea in chorea-acanthocytosis was preceded potentials record the cortical activity evoked
by a slow negative potential resembling the BP, by a stimulus charged with cognitive signif-
whereas the chorea in Huntington’s disease was icance. As such, event-related potentials are
not.25 more sensitive to the endogenous reaction to
In patients suffering from Restless Legs Syn- a stimulus than to the physical nature of the
drome, no BP could be elicited before the stimulus. Sutton et al.30 were among the first
restless limb movements, whereas a BP pre- to note a large late cortical positivity in reac-
cedes voluntary leg movements in the same tion to stimuli to which the subject attached
subjects.26 importance. Since that time, numerous tech-
Propriospinal myoclonus is not preceded by niques have been devised to record the cor-
the BP.27 Psychogenic jerks, on the other hand, tical activity surrounding processes such as
Movement-Related Potentials and Event-Related Potentials 233
selective attention,31 memory,32 olfaction,33 and amplitude is decreased and the latency is
facial recognition.34 A comprehensive review of prolonged in all types of dementia.
this topic—spanning the disciplines of physiol-
ogy, psychology, psychiatry, and neurology—is
beyond the scope of this book.
SUMMARY
Special EEG averaging techniques may be
The P300 used to study the cortical processes underlying
movement and cognition. Movement-related
The P300 is the most commonly recorded potentials and contingent negative variation are
event-related potential.35 Generally, an oddball observed before a voluntary movement occurs.
technique of auditory stimulation is used, in The P300 and other event-related potentials
which a standard stimulus, also called frequent provide electrophysiologic correlates of per-
stimulus, is replaced at infrequent intervals by ception and cognition.
a stimulus of different tone, termed the odd-
ball stimulus or rare stimulus. The subject is
instructed to attend to or to count the oddball
stimuli. Only trials triggered by this rare event REFERENCES
are averaged.
At times, the P300 is visible on a single raw 1. Kornhuber, H. H., and L. Deecke. 1965. Hirnpo-
tentialänderungen bei Willkürbewegungen und pas-
tracing. Averaging clearly defines a wave with siven Bewegungen des Menschen: Bereitschaftspo-
a peak latency of approximately 300 ms and an tential und reafferente Potentiale. Pflügers Archive
amplitude of approximately 10 μV. The ampli- 284:1–17.
tude of the wave is increased by many factors, 2. Deecke, L., P. Scheid, and H. H. Kornhuber. 1969.
including the subject’s attentiveness and the Distribution of readiness potential, pre-motion positiv-
ity and motor potential of the human cerebral cortex
unpredictability of the oddball stimulus. The preceding voluntary finger movement. Experimental
P300 has a bilateral, mid-parietal distribution. Brain Research 7:158–68.
However, a single generator for the poten- 3. Shibasaki, H., G. Barrett, E. Halliday, and A. M. Hal-
tial cannot be defined; the wave likely reflects liday. 1980. Components of the movement-related
cortical potential and their scalp topography. Elec-
activity in several areas of the brain.36 The role troencephalography and Clinical Neurophysiology 49:
of the wave in cognition is also debated. It may 213–16.
be the electrophysiologic correlate of selected 4. Shibasaki, H., and M. Hallett. 2006. What is
attention. the Bereitschaftspotential? Clinical Neurophysiology
The P300 is abnormal in many diseases 117:2341–2356.
5. Singh, J., R. T. Knight, D. L. Woods, D. J. Beckley,
in which cortical processing is impaired. The and C. Clayworth. 1987. Lack of age effect on human
amplitude is decreased and the latency is pro- brain potentials preceding voluntary movements. Neu-
longed in all types of dementia.37 Prolonga- roscience Letters 119:27–31.
tion of the P300 latency may also be a pre- 6. Roggeveen, A. B., D. J. Prime, and L. M. Ward. 2007.
Lateralized readiness potentials reveal motor slowing
clinical finding in those at risk of develop- in the aging brain. The Journals of Gerontology. Series
ing Alzheimer’s disease.38 Abnormalities have B, Psychological Sciences and Social Sciences 62:
also been reported in PD,39 mild metabolic P78–84.
encephalopathies, drug intoxications, multiple 7. Evidente, V. G. H., J. N. Caviness, B. Jamieson,
sclerosis, autism, and schizophrenia.39 A. Weaver, and N. Joshi. 1999. Intersubject vari-
ability and intrasubject reproducibility of the Bere-
itschaftspotential. Movement Disorders 14:313–19.
8. Barrett, G., H. Shibasaki, and R. Neshige. 1986. Corti-
Key Points cal potential shifts preceding voluntary movement are
normal in parkinsonism. Electroencephalography and
• Event-related potentials record the corti- Clinical Neurophysiology 63:340–8.
cal activity evoked by a stimulus charged 9. Jahanshahi, M., I. H. Jenkins, R. G. Brown, C. D.
with cognitive significance. Marsen, R. E. Passingham, and D. K. Brooks. 1995.
• The P300 is the most commonly recorded Self-initiated versus externally triggered movements.
I. An investigation using measurement of regional
event-related potential. cerebral blood flow with PET and movement-related
• The P300 is abnormal in many diseases in potentials in normal and Parkinson’s disease subjects.
which cortical processing is impaired; the Brain 118:913–33.
10. Dick, J. P. R., J. X. Rothwell, B. L. Day, et al. 1989. potential. Journal of Neurology, Neurosurgery, and
The Beretischaftspotential is abnormal in Parkinson’s Psychiatry 61:103–6.
disease. Brain 112:233–44. 25. Shibasaki, H., T. Sakai, H. Nishimura, Y. Sato, I.
11. Vidailhet, M., F. Stocchi, J. C. Rothwell, et al. Goto, and Y. Kuroiwa. 1982. Involuntary movements
1993. The Bereitschaftspotential preceding simple in chorea-acanthocytosis: A comparison with Hunting-
foot movement and initiation of gait in Parkinson’s ton’s chorea. Annals of Neurology 12:311–14.
disease. Neurology 43:1784–8. 26. Trenkwalder, C., S. F. Bucher, W. H. Oertel, D.
12. Dick, J. P. R., R. Cantello, O. Buruma, et al. 1987. Proeckl, H. Plendl, and W. Paulus. 1993. Beretis-
The Bereitschaftspotential, L-dopa and Parkinson’s chaftspotential in idiopathic and symptomatic restless
disease. Electroencephalography and Clinical Neuro- legs syndrome. Electroencephalography and Clinical
physiology 66:263–74. Neurophysiology 89:95–103.
13. Limousin, P., R. G. Brown, M. Jahanshahi, et al. 27. Brown, P., P. D. Thompson, J. C. Rothwell, B. L.
1999. The effects of posteroventral pallidotomy on the Day, and C. D. Marsden. 1991. Axial myoclonus of
preparation and execution of voluntary hand and arm propriospinal origin. Brain 114:197–214.
movements in Parkinson’s disease. Brain 122:315–27. 28. Terada, K., A. Ikeda, P. C. Van Ness, et al. 1995. Pres-
14. Gironell, A., A. Rodriguez-Fornells, J. Kulisevsky, B. ence of Bereitschaftspotential preceding psychogenic
Pascual, M. Barbanoj, and P. Otermin. 2002. Motor myoclonus: Clinical application of jerk-locked back
circuitry re-organization after pallidotomy in Parkin- averaging. Journal of Neurology, Neurosurgery, and
son disease. A neurophysiological study of the Bere- Psychiatry 58:745–7.
itschaftspotential, contingent negative variation, and 29. Ikeda, A., H. Shibasaki, T. Nagamine, et al. 1994.
N30. Journal of Clinical Neurophysiology 19:553–61. Dissociation between contingent negative variation
15. Shibasaki, H., F. Shima, and Y. Kuroiwa. 1978. Clin- and Bereitschaftspotential in a patient with cerebellar
ical studies of the movement-related cortical poten- efferent lesion. Electroencephalography and Clinical
tial (MP) and the relationship between the denta- Neurophysiology 90:359–64.
torubrothalamic pathway and readiness potential (RP). 30. Sutton, S., M. Braren, J. Zubin, and E. R. John. 1965.
Journal of Neurology 219:15–25. Evoked-potential correlates of stimulus uncertainty.
16. Gerloff, C., E. Altenmuller, and J. Dichgans. 1996. Science 150:1187–8.
Disintegration and reorganization of cortical motor 31. Heinze, H. J., G. R. Mangun, W. Burchert, et al. 1994.
processing in two patients with cerebellar stroke. Combined spatial and temporal imaging of brain activ-
Electroencephalography and Clinical Neurophysiol- ity during visual selective attention in humans. Nature
ogy 98:59–68. 372:543–6.
17. Jankelowitz, S. K., and J. G. Colebatch. 2005. Move- 32. Begleiter, H., B. Porjesz, and W. Wang. 1993. A neu-
ment related potentials in acutely induced weakness rophysiologic correlate of visual short-term memory in
and stroke. Experimental Brain Research 161:104–13. humans. Electroencephalography and Clinical Neuro-
18. Feve, A., N. Bathien, and P. Rondot. 1994. Abnormal physiology 87:46–53.
movement related potentials in patients with lesions 33. Evans, W. J., L. Cui, and A. Starr. 1995. Olfactory
of basal ganglia and anterior thalamus. Journal of event-related potentials in normal human subjects:
Neurology, Neurosurgery, and Psychiatry 57:100–4. Effects of age and gender. Electroencephalography
19. van der Kamp, W., J. C. Rothwell, P. D. Thompson, and Clinical Neurophysiology 95:293–301.
B. L. Day, and C. D. Marsen. 1995. The movement 34. Hertz, S., B. Porjesz, H. Begleiter, and D. Chorlian.
related cortical potential is abnormal in patients with 1994. Event-related potentials to faces: The effects of
torsion dystonia. Movement Disorders 10:630–3. priming and recognition. Electroencephalography and
20. Deuschl, G., C. Toro, J. Matsumoto, and M. Hallett. Clinical Neurophysiology 92:342–51.
1995. Movement-related cortical potentials in writer’s 35. Picton, T. W. 1992. The P300 wave of the human
cramp. Annals of Neurology 38:862–8. event-related potential. Journal of Clinical Neuro-
21. Yazawa, S., A. Ikeda, R. Kaji, et al. 1999. Abnor- physiology 9:456–79.
mal cortical processing of voluntary muscle relax- 36. Neshige, R., and H. Luders. 1992. Recording of event-
ation in patients with focal hand dystonia studied by related potentials (P300) from human cortex. Journal
movement-related potentials. Brain 122:1357–66. of Clinical Neurophysiology 9:294–8.
22. Adler, L. E., M. Pecevich, and H. Nagamoto. 1989. 37. Goodin, D. S., K. C. Squires, and A. Starr. 1978.
Bereitschaftspotential in tardive dyskinesia. Movement Long latency event-related components of the audi-
Disorders 4:105–12. tory evoked potential in dementia. Brain 101:635–48.
23. Obeso, J. A., J. C. Rothwell, and C. D. Marsden. 1981. 38. Green, J., and A. I. Levey. 1999. Event-related poten-
Simple tics in Gilles de la Tourette’s syndrome are not tial changes in groups at increased risk for Alzheimer
prefaced by a normal premovement EEG potential. disease. Archives of Neurology 56:1398–403.
Journal of Neurology, Neurosurgery, and Psychiatry 39. Rumbach, L., C. Tranchant, J. F. Viel, and J. M.
44:735–8. Warter. 1993. Event-related potentials in Parkinson’s
24. Karp, B. I., S. Porter, C. Toro, and M. Hallett. 1996. disease: A 12-month follow-up study. Journal of Neu-
Simple motor tics may be preceded by a premotor rology Science 116:148–51.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART B
Sensory Pathways
The sensory axons conduct information from from sensory structures in response to specific
the periphery to the central nervous system as stimulation are called sensory evoked poten-
action potentials traveling from peripheral sen- tials. Sensory evoked potentials are classified
sory receptors centrally to the spinal cord and as nerve conduction studies that test peripheral
cortex. Primary sensory neurons located in the nerves (Chapter 17), somatosensory evoked
dorsal root and cranial ganglia send a periph- potentials that test central somatic sensory
eral axon out to the limb or cranial receptors, pathways (Chapter 18), brain stem auditory
and a central axon to the brain stem. Sensory evoked potentials that test peripheral and cen-
axons are therefore among the longest in the tral auditory pathways (Chapters 19–21), and
body traveling as far as from the foot to the visual evoked potentials that test peripheral
neck. The sensory axons make reflex connec- and central visual pathways (Chapter 22).
tions in the brain stem and spinal cord, as well Combinations of these sensory potential
as to the cerebellum and cerebral hemispheres. recordings help to determine localization of
Because the signals from single axons are damage at different levels of the nervous sys-
difficult to record, most electrophysiologic tem and, in some cases, help to define the type
recordings from nerves being tested for pos- of underlying lesion. Advances in the method
sible neurologic or neuromuscular disease are of stimulation and recording have extended the
summated responses made from specific gen- applications of some of these techniques.
erators in response to controlled external stim- Mechanical components of peripheral audi-
ulation. Somatic sensory and somatic motor tory function can be tested separately with
axons can be tested by stimulating along the audiography, acoustic reflexes, and evoked oto-
length of a nerve while recording the sensory or acoustic emissions (Chapter 20). Movement-
motor responses from peripheral nerve or mus- related potentials and event-related potentials
cle. Sensory axons can be isolated for testing (see Chapter 16 in Part A) are also sometimes
by selective stimulation of sensory structures referred to as evoked potentials. Evoked poten-
or by selective recording from generators that tials obtained with stimulation of motor axons
are purely sensory. The potentials recorded are considered in Part C.
237
Chapter 17
Sensory Nerve Action Potentials

Eric J. Sorenson
INTRODUCTION Distance Between Recording

PATHOPHYSIOLOGY OF SNAPs Electrode and Nerve
Normal SNAPs Measurement
Abnormal SNAPs Interelectrode Distance
Clinical Importance of SNAPs
PLANNING THE STUDY AND
METHODS OF STUDY OF SNAPs FINDINGS IN DISEASES
Nerve Stimulation Radiculopathy
Recording the Potential Plexopathy
Averaging Common Mononeuropathies
Myopathy
MEASUREMENTS Disorders of the Neuromuscular
TECHNICAL FACTORS Junction
Noise and Shock Artifact Motor Neuron Diseases
Submaximal Stimulation
Recording and Stimulating Distance SUMMARY
Temperature
INTRODUCTION The main indication for assessing periph-

eral SNAPs is in studying sensory nerves. Sen-
This chapter reviews the electrical princi- sory nerve action potentials remain the most
ples relevant to sensory nerve action potential reliable means of testing peripheral sensory
(SNAP) studies, including the neurophysiol- nerves.1 Sensory nerve action potentials are
ogy intrinsic to the axon and extrinsic ambi- very sensitive to pathologic conditions in a
ent electrical interference and their effects on nerve. Often, alterations in the amplitudes, ter-
these studies, the clinical relevance of SNAPs, minal latencies, and conduction velocities of
and strategies for planning appropriate sensory evoked responses are the earliest abnormali-
testing to maximize diagnostic yield. Also, the ties detected in a peripheral neuropathic pro-
recording of SNAPs is frequently complicated cess.2 The studies provide invaluable data for
by unique technical difficulties and common the localization and classification of a periph-
human errors, which are discussed along with eral neuropathy, and the study results may
a strategy for troubleshooting such problems. indicate a specific pathologic condition. While
Finally, the chapter explains how SNAPs are most of the peripheral nerves that are studied
measured and how the function of a nerve is are mixed nerves with both motor and sen-
quantified. sory axons, sensory components can be isolated
239
by stimulating or recording from pure sensory

branches. Studying a pure population of fibers
provides information that is more meaningful
for interpretation. For practical purposes, this
is only possible for the sensory axons.
Purpose and Role of SNAPs

• Provide objective evidence of the integrity
of the peripheral sensory nerves.
• Assist in identifying and localizing mono-
neuropathies.
• Used to assess sensory involvement in
generalized peripheral neuropathies.
• Helps to distinguish lesions of the spinal
roots (preganglionic) from lesions of the
plexus (postganglionic).
PATHOPHYSIOLOGY OF SNAPs
Electrically evoked nerve action potentials are
the only way clinically to directly study the
function of peripheral nerves. Unlike com- Figure 17–1. Effect of decreasing the interelectrode dis-
tance on the amplitude of the response. As the distance
pound muscle action potentials, these poten- decreases, the amplitude diminishes.
tials are not affected by secondary factors such
as transmission at the neuromuscular junc-
tion or the electrical excitability of the mus-
cle. Except perhaps for sympathetic sweat
sufficient to prevent the G2 electrode from dis-
gland skin potentials,3 the electrical excitabil-
torting the waveform as it is recorded at the
ity of sensory nerve target receptors cannot
G1 electrode (Fig. 17–1). Distances greater
be measured reliably. Although somatosensory
than 5 cm or placement of the G2 electrode
evoked potentials provide information about
somewhere beyond the distal nerve segment
the proximal peripheral sensory pathways (see
introduces the possibility of volume conducted
Chapter 18), SNAPs are the only practical way
responses from distant electrical generators.
to assess sensory peripheral nerves reliably.
The nerve action potential is induced by an
Because of these factors, the measurement
electrical stimulus to the nerve sufficient to
of electrically evoked peripheral SNAPs pro-
generate an action potential in all axons simul-
vides invaluable information about the physi-
taneously. This nerve action potential creates
ology and pathology of the peripheral nervous
an electrical field that propagates along the
system, particularly about sensory nerves.
length of the nerve (Fig. 17–2). The initial por-
tion of the electrical field is positive, which,
as it approaches the G1 electrode, results in
Normal SNAPs the initial downward (positive) phase of the
nerve action potential waveform. As the nerve
Sensory nerve action potentials are recorded action potential continues to propagate, the
using two recording electrodes grounded to a large negative portion of the electrical field
common ground. The active electrode, com- (the region of depolarization) passes beneath
monly referred to as the G1 electrode, is placed the G1 electrode resulting in the large upward
upon the skin superficial to the nerve being (negative) phase of the nerve action potential
tested. The reference electrode, G2, is placed waveform. Finally, as the nerve action poten-
along the length of the nerve, typically 3–5 cm tial travels away from the G1 electrode the
distal to the G1 electrode. This distance is final positive portion of the electrical field
Sensory Nerve Action Potentials 241
+0.2 –1.6 +0.7

–1.0 +0.4
–0.6
–0.4
+ – – +
0 –0.2 0
Figure 17–2. Example of the electrical field generated at the axon during the action potential.
volume conducted responses from distant

electrical generators.
• The amplitude of the SNAP response is
reduced with increasing recording dis-
tance from the nerve.
Abnormal SNAPs
A major advantage of studying SNAPs is their
sensitivity for detecting an underlying abnor-
mality. Frequently, sensory potential abnor-
malities are the earliest findings in a peripheral
Figure 17–3. Triphasic appearance of a nerve action
potential. Note initial positivity (A), followed by the dom- nerve disorders and SNAPs are often more
inant negative peak (B), and ending with a final positive sensitive in detecting peripheral nerve abnor-
phase (C). malities than compound muscle action poten-
tials since they are unaffected by reinnervation.
To interpret the findings of SNAP responses
passes beneath the G1 electrode resulting in accurately, the temporal profile and evolution
the last downward (positive) phase of the nerve of the changes in SNAPs must be under-
action potential waveform. The net effect is stood. With acute nerve injuries characterized
the characteristic triphasic appearance to the by injury to the axons, the SNAP remains nor-
SNAP waveform (Fig. 17–3). The magnitude of mal immediately following the injury. Over the
the electrical field diminishes with the square ensuing days, Wallerian degeneration begins
of the distance from the electrical genera- and the degenerating axons lose their electri-
tor (i.e., the axon membrane). As a result of cal excitability. As a result, the amplitude of the
this diminution with increasing distance, the SNAP begins to decrease, reaching its nadir
appearance of the SNAP changes with a reduc- 5–10 days after the injury.
tion in amplitude and a less apparent triphasic In contrast to lesions producing axonal loss,
appearance with increased distance. if the lesion affects only the myelin sheath
and causes a focal area of demyelination, the
SNAPs distal to the lesion may remain nor-
Key Points
mal. Stimulating and recording across the area
• The normal waveform configuration of an of demyelination, however, may show a delay
SNAP is triphasic. in the conduction velocity across the site; the
• Recording the SNAP with a 3–5 cm sepa- delay will be seen as either a prolonged latency
ration between G1 and G2 maximizes the or slowed conduction velocity. Focal conduc-
amplitude and reduces the possibility of tion block, as defined by a discrete loss of
substantially. However, the distal latency is nor-

mal. If the area of demyelination causes pro-
nounced conduction block or dispersion, no
proximal response may be obtained. However,
true conduction block, defined as loss of ampli-
tude over a discrete segment, cannot be deter-
mined reliably by sensory nerve conduction
studies alone.
If the focal demyelinating lesion occurs only
in the terminal segment of the nerve, the dis-
tal latency is prolonged. The amplitudes are
frequently reduced if the lesion is associated
with conduction block or phase cancellation. A
conduction velocity obtained by stimulating or
recording at two sites proximal to the lesion in
a terminal segment may be slightly decreased
because the largest, fastest conducting fibers in
the area of demyelination are affected. If there
Figure 17–4. Phase cancellation of a normal SNAP with is an area of demyelination proximal to both
a larger amplitude and area with distal stimulation (top stimulation sites, the nerve action potential
tracing) when compared to proximal stimulation (bottom
tracing).
responses may be entirely normal.
In diffuse demyelination, in which the nerve
is affected all along its course, distal laten-
cies are prolonged, conduction velocities are
amplitude across a focal segment usually can- slowed, and amplitudes are reduced. However,
not be established with SNAPs, since, in nor- the electrophysiologic features that are used
mal nerves, nerve action potentials demon- to differentiate an acquired demyelinating pro-
strate an unpredictable reduction in amplitude cess from an inherited process, that is, conduc-
with increasing distance between the stimulat- tion block and dispersion, cannot be assessed
ing and recording electrodes (Fig. 17–4). This accurately with SNAPs.
is caused by phase cancellation, which results In cases of diffuse demyelination, testing
from sensory fibers having a wide range of con- SNAPs over long segments will be more sen-
duction velocities even within the same nerve. sitive; because the effects of subtle slowing will
Understanding how the pathology affects be additive, the changes become more appar-
SNAPs is necessary to classify the neuropathic ent as a longer segment of nerve is tested. Con-
process. This classification has important impli- versely, in the case of focal demyelination, test-
cations for the differential diagnosis of neu- ing over the shortest segment of nerve possible
ropathy.1 The usual finding in axonal neuropa- provides the greatest sensitivity, because subtle
thy is a decrease in amplitude. Because the areas of focal conduction slowing are not aver-
neuropathic process preferentially can affect aged with the normally conducting segments.
the largest fibers and, hence, the fastest con- The difference between focal and diffuse dis-
ducting ones, the conduction velocity may be orders becomes important when selecting the
slightly below the limit of normal. Generally, SNAPs to sample in response to a specific clin-
axonal loss alone should not decrease conduc- ical situation. For example, if focal median
tion velocity less than approximately 70% of the mononeuropathy at the wrist is suspected, sam-
lower limit of normal.4 Similarly, the terminal pling the SNAP across a short distance, as
distal latency of the response may be slightly in the palmar (orthodromic) stimulation tech-
prolonged. nique, provides the greatest sensitivity. How-
If the nerve has a focal area of demyeli- ever, if a diffuse disorder is suspected, a median
nation, the findings depend on the sites of antidromic technique with proximal and distal
stimulation. If the stimulation sites are prox- stimulation is preferred because the amplitude
imal and distal to the area of demyelination, is more reproducible and the conduction veloc-
the conduction velocity usually is decreased ity is sampled over a long segment of nerve.

• SNAPs are often more sensitive in • SNAPs are very effective at distinguishing
detecting peripheral nerve abnormal- lesions central to the dorsal root ganglia
ities than compound muscle action (preganglionic) from lesion peripheral to
potentials. the dorsal root ganglia (postganglionic).
• SNAP may remain normal for the first • SNAP amplitudes remain normal distal to
several days following an acute injury. a conduction block.
• The SNAP amplitude is decreased follow-
ing axonal injuries, reaching its nadir 5–10
day after the injury. METHODS OF STUDY OF SNAPs
• Processes producing demyelination may
produce slowing of conduction veloc- Two main strategies have been developed for
ity or prolongation of terminal latencies studying sensory axons. The first strategy is
in SNAPs. to stimulate the mixed motor and sensory
• Conduction block cannot be reliably iden-
nerve and to record the SNAP from a termi-
tified with SNAPs due to the phase can- nal sensory branch at a site that is distal to
cellation that normally occurs in sensory where the nerve splits into motor and sen-
nerves. sory components. This is called the antidromic
technique, because the direction of the action
potential is opposite (anti-) that of the physio-
Clinical Importance of SNAPs logic action potential (Fig. 17–5). The advan-
tage of the antidromic technique is that it
Since SNAPs are sensitive to peripheral neu- ensures adequate supramaximal stimulation
ropathic lesions, they can provide important of the nerve and, thus, larger amplitudes.5, 6
information for localization. The distribution However, this technique also activates the
of abnormalities can suggest a focal lesion, a motor fibers and generates a muscle action
multifocal process, or a diffuse disease. Also, potential that, through volume conduction,
because of the unique anatomy of sensory neu- may interfere with the SNAP that is being
rons, SNAPs are extremely helpful in differ- recorded.
entiating an intraspinal process from a more The second strategy for isolating sensory
peripheral one. The cell bodies of the sen- fibers in a mixed nerve is to stimulate the nerve
sory neurons form dorsal root ganglia, which distal to the point where it splits into sensory
lie within the intervertebral foramina, where and motor components and to record proxi-
the spinal roots exit from the spinal canal. mally over the mixed nerve. This is called the
Thus, a process that is localized within the orthodromic technique, because the direction
spinal canal is described as preganglionic. In of the action potential is the same as that of the
a preganglionic lesion, the distal sensory axon physiologic action potential (Fig. 17–6). How-
remains intact and connected with the cell ever, the number of fibers activated and the
body. The SNAPs remain normal, even if the amplitude of the responses are more variable
sensory loss is severe. Conversely, any lesion than with the antidromic technique. The main
that affects the nerve by interrupting its axons advantage of the orthodromic technique is that
distal to the intervertebral canal causes a loss it eliminates volume conduction from muscle
of amplitude in SNAPs. This provides invalu- action potentials because no motor fibers are
able information for differentiating a pregan- activated.
glionic lesion such as a radiculopathy from Another, and less optimal technique, is to
a postganglionic lesion such as a plexopa- stimulate a mixed nerve and to record at a fixed
thy or mononeuropathy. However, a postgan- distance over the nerve where it contains both
glionic lesion that does not affect the axons motor and sensory fibers. With this technique,
but produces a pure conduction block will not it is assumed that because the nerve has a small
affect the amplitude of SNAPs if the stim- number of motor axons and the amplitude of
ulation and recording sites are distal to the the SNAP is directly proportional to the total
lesion. number of axons activated, the contribution
Figure 17–5. Median antidromic sensory technique. Stimulation of the median nerve at the wrist will activate both motor
and sensory fibers; however, recording from the digit will isolate the sensory fibers to generate an SNAP. Amplitudes are
higher than classic orthodromic and more reliable than palmar, but there is a volume conducted motor artifact of slower
rise time and longer duration that may need to be minimized by moving recording electrodes distally.
Figure 17–6. Median orthodromic sensory technique. A, Classic digital stimulation of the median nerve activates only
cutaneous sensory fibers allowing for a low amplitude SNAP to be recorded from the mixed motor and sensory median nerve
at the wrist. B, Palmar stimulation gives higher amplitude mixed motor and sensory potential, but is technically difficult.
244
of the motor fibers to the SNAP is negligible. stimulator be as close to the nerve as pos-
Because of the variation from subject to subject sible. To initiate the study, the stimulator is
in the motor and sensory components of mixed placed over the approximate location of the
nerves, normal data obtained with this tech- nerve, with the cathode pointed towards the
nique are more variable than comparable data recording electrodes (distal in antidromic stud-
obtained with the antidromic and orthodromic ies and proximal in orthodromic studies). The
techniques. nerve should be depolarized with the cath-
Some of the peripheral nerves that are avail- ode, not the anode. Depolarization with the
able for testing are pure sensory nerves. For anode distally can cause an electrical conduc-
these nerves, it is necessary only to stimulate tion block, called anodal block, of some of the
the nerve and to record at a fixed distance along axons and result in a submaximal response.
it, using either the antidromic or orthodromic The action potential is generated at the cath-
technique. Examples of pure sensory nerves ode, if the anode is placed distally, the dis-
are the sural, superficial peroneal, saphenous, tance measurements will not be accurate. As
and medial and lateral antebrachial sensory the current is increased gradually, an SNAP
nerves. Although superficially the technique becomes apparent. At this stage, the stimulator
for studying these nerves appears straightfor- is moved laterally (sliding) to identify the site
ward, technical factors may make it difficult, as at which the response is maximal. At the site
discussed below. of the maximal response, the current is grad-
Some laboratories have utilized magnetic ually increased until the amplitude reaches its
stimulation to activate the peripheral nerves. maximum.
This technique was developed because of the If stimulation creates a large shock arti-
perception that a magnetic stimulus would be fact, several methods can be used to reduce
less painful than electrically elicited responses. the artifact. Rotating the anode off the nerve
However magnetic stimulation does not allow while stimulating it often decreases the shock
for a precise localization of the initiation artifact (Fig. 17–7). If this is ineffective, con-
site of the elicited nerve action potential. firm that the ground is in the appropriate
This makes latencies and conduction veloc- location. Check the impedance of the record-
ities unreliable. Because of this poor relia- ing and stimulating electrodes, and if neces-
bility, magnetic stimulation has fallen out of sary, apply conduction paste to improve the
favor. impedance values. After this has been done,
if an acceptable response cannot be obtained
Key Points in large limbs, consider near-nerve stimula-
• Orthodromic techniques minimizes motor tion with a monopolar needle electrode. With
near-needle stimulation, the monopolar nee-
artifact but SNAP amplitudes are more
dle serves as the cathode and the surface
variable.
• Antidromic techniques produce more reli- electrode, as the anode. Near-nerve stim-
ulation has several advantages. The largest
able SNAP amplitudes but motor artifact
amount of impedance arises from transcu-
may interfere with the waveforms.
• Pure sensory nerves, such as the sural, sup- taneous stimulation. Thus, placing the nee-
dle within the subcutaneous tissue eliminates
erficial peroneal, saphenous, and medial
the transcutaneous resistance. Also, placing
and lateral antebrachial nerves can be
the needle much closer to the nerve allows
studied with orthodromic or antidromic
supramaximal stimulation at a lower level of
techniques.
current.
Key Points
Nerve Stimulation
• Nerve stimulation is performed with the
The sensory nerve should be stimulated with cathode pointed toward the recording
a current that is sufficient to activate all the electrodes Anode block, due to the anode
sensory axons in the nerve but not to cause pointed towards the recording electrodes,
overstimulation. This balance requires that the may result in a submaximal response.
distance of 3.5–4.0 cm maximizes the sen-

sory response and minimizes the amount of
electrical interference. If the electrical back-
ground noise is excessive, assess the impedance
of the electrodes. Any impedance mismatch
in the electrodes causes problems with the
common-mode rejection, and the signals gen-
erated by background noise will be amplified.
If no satisfactory response can be obtained,
confirm that the location of the electrodes is
correct.
Occasionally, the motor artifact may dis-
tort the SNAP waveform recording and, if the
SNAP is absent, the motor artifact may be mis-
taken for a sensory potential. Motor artifact
primarily occurs when applying the antidromic
technique, such as during the ulnar antidromic
study recording from the fifth digit, where
motor and sensory fibers are co-stimulated.
Motor artifact should be recognized by the
longer latency, slower rise time and longer
duration, broader response compared to sen-
sory responses. If motor artifact is present
stimulation of the nerve should be repeated
at a lower stimulus to determine if a supra-
maximal SNAP can be elicited without the
motor activation. If this is unsuccessful then
the G1 and G2 recording electrodes should
be moved slightly more distal on the digit to
Figure 17–7. Effect of rotation of the anode during separate the motor artifact from the SNAP
stimulation. Note the large shock artifact in the upper waveform.
and lower tracings obscuring the onset and amplitude
of the waveform. Rotation of the anode will minimize
the shock artifact resulting in an acceptable recording
(middle trace). Key Points
• SNAPs are recorded using G1 and G2 disc
• Rotating the anode off the nerve can or ring electrodes, placed 3.5–4.0 cm apart.
reduce shock artifact. • Motor artifact may occur with antidromic
• Near nerve stimulation using a monopo- studies, and can be recognized by the
lar needle can be used when a large shock longer latency, slower rise time and longer
artifact cannot be reduced with other duration, broader response compared to
methods. sensory responses.
• If motor artifact is present stimulation of
the nerve should be repeated at a lower
stimulus or the G1 and G2 recording elec-
Recording the Potential trodes should be moved slightly more dis-
tal on the digit to separate the motor
SNAPs are recorded using disc or ring artifact from the SNAP waveform.
electrodes. Standard 5 mm–1 cm plate elec-
trodes are typically used in the limb. When
recording over the digits, ring electrodes
are used that record the electrical poten- Averaging
tials circumferentially from the digit. Record-
ing SNAPs requires appropriate placement of Sensory potentials have much lower amplitudes
the G1 and G2 electrodes. An interelectrode than motor responses; thus, the signal-to-noise
ratio is much lower. Because of this, the sen-

sory responses are much more affected by
background ambient electrical noise, even with
good testing technique. The low signal-to-noise
ratio can be improved by averaging, which
can reduce or eliminate random background
noise. The improvement in the signal-to-noise
ratio is directly proportional to the square root
of the number of responses averaged. This
means that improvement is greatest with the
first few responses that are averaged and lit-
tle more is gained after averaging the first
four or five responses. Sensory nerve action
potentials should routinely be averaged 4–5
times to maximize the quality of the wave-
form while minimizing the distortion from the
background noise.
Figure 17–8. Amplitude measurement of a potential,
with, top, and without, bottom, an initial positive peak.
Key Points
• Averaging is useful to maximize the signal-
to-noise ratio of the recorded SNAP
responses.
• Four or five SNAP responses should be
routinely averaged during a standard sen-
sory nerve conduction study.
MEASUREMENTS Figure 17–9. Onset latency at 2.5 ms and peak latency at

3.2 ms.
Clinically, the most relevant measurements

are amplitude, distal latency, and conduc-
Conduction velocity is the last variable that
tion velocity. Amplitude is measured as a
is assessed clinically. Conduction velocity is
peak-to-peak amplitude, from the nadir of
obtained by dividing the distance between two
the initial positivity to the peak negativity
stimulus sites by the conduction time. The
(Fig. 17–8). Because amplitude decreases sub-
onset latencies are measured and subtracted
stantially with increasing length of the seg-
to obtain conduction time (Fig. 17–10). The
ment, the amplitude that is recorded and com-
resulting time represents conduction time. In
pared is the distal amplitude.
this instance, the onset latencies are chosen
The distal latency that is recorded and com-
because they represent the fastest conducting
pared is the peak latency, not the onset latency
fibers, and the calculated conduction veloc-
of the response. Peak latency is the time from
ity reflects the speed of conduction in these
the stimulus to the peak negative deflection.
fibers.
Onset latencies are more variable and difficult
to measure (Fig. 17–9). The distances for the
Key Points
distal response should be kept fixed so that the
latencies can be compared directly with nor- • The SNAP parameters of interest include
mal values. If the distance is variable, then the amplitude (measured from the max-
the latency must be divided by the distance imal positive deflection to the maximal
(generating a figure analogous to a conduc- negative deflection), distal latency (mea-
tion velocity) to allow for comparisons. This is sure from the stimulus to the peak
not optimal, because it introduces a source of latency), and conduction velocity (mea-
human error. sured from the SNAP onset latency).
Figure 17–10. Measurement of conduction velocity in a proximal nerve segment. The difference in latencies between
the proximal and distal stimulation sites is divided by the distance between the two sites. Latencies are measured to the
initial negative deflection (onset latency).
• The peak latency is the time from the and, fourth, the distance between the elec-
stimulus to the peak negative deflection, trodes and the nerve affect the amplitude. To
and is the most reliable and easy to mea- obtain reproducible and comparable results,
sure latency parameter. the stimulation distance, nerve temperature,
• Onset latencies between two waveforms and recording distance must be controlled and
recorded or stimulated at different sites standardized for each sensory nerve tested.
are used in determining the conduction Each of these factors is discussed below.
velocity since they represent the fastest
conducting fibers.
Noise and Shock Artifact
A main difficulty in recording well-defined
TECHNICAL FACTORS SNAP responses is background electrical noise.
Since the sensory amplitudes are low, the back-
The SNAPs measured in the electrophysiology ground noise appears proportionally larger;
laboratory represent a summation of individ- this is referred to as a lower signal-to-noise
ual action potentials of all the large myelinated ratio. Thus, it is imperative that background
sensory axons in the stimulated nerve. Because electrical activity be minimized. This includes
the responses are recorded from the nerve proper impedance of the electrodes to avoid
and not from the muscle, the amplitudes are any impedance mismatch that would distort
much lower than those of compound muscle the common-mode rejection between the elec-
action potentials. This causes several prob- trodes. Electrode impedance can be mini-
lems that are not usually encountered in motor mized by applying contact paste to the elec-
conduction studies. The amplitude of SNAPs trodes. If disposable electrodes are used,
depends on several factors. First, the ampli- repeated use of the electrodes will reduce the
tude is directly proportional to the number of effectiveness of the contact gel increasing the
axons that are depolarized. Second, the dis- electrode impedance. Disposable electrodes
tance between the recording and stimulation with increasing impedance should be replaced
sites affects the amplitude. Third, the temper- with a new set. Dry desiccated skin has a very
ature of the nerve at the time of the study high impedance that cannot be overcome with
conducting paste. This is most common in cal- disperse. Thus, the closer a stimulus is applied
loused hand and feet. This impedance can be to the recording site, the less the dispersion
markedly improved by gently abrading the skin and the larger the amplitude. The longer the
with an abrasion board or tape to remove the distance, the greater the dispersion and the
dry and desiccated skin. In addition to ensur- lower the amplitude. Because this diminution
ing proper impedance, background voluntary in amplitude with increasing distance is unpre-
muscle activity that interferes with the baseline dictable, it is preferable to record and to com-
must be minimized. pare the distal amplitudes of the responses
The stimulus should be delivered so that instead of the proximal ones.
shock artifact is minimized. This requires
proper ground placement, and it often requires
rotating the stimulator to minimize the effects Temperature
of shock artifact. It is important to place the
stimulator as near the nerve as possible to allow Cool limb temperature prolongs the dura-
supramaximal depolarization with the minimal tion of the depolarization of the axon mem-
amount of current. This will benefit the study brane, thus prolonging the action potential.
by minimizing shock artifact and by reduc- This in turn increases the amplitude of the
ing the risk of overstimulation and creating response. This effect is significant, and limb
muscle artifact. Occasionally, a needle cathode temperature cannot be ignored when perform-
needs to be placed near the nerve to pro- ing nerve conduction studies. SNAPs are more
vide the appropriate amount of stimulation. susceptible to temperature effects than the
Sources of external electrical activity need to motor compound muscle action potentials. A
be eliminated, including any electrical equip- cool limb also has the effect of slowing con-
ment within the vicinity of the study, particu- duction, which is apparent in the prolonged
larly fluorescent lighting. Incandescent lighting distal latency and slowed conduction velocity
does not produce the same electrical inter- (Fig. 17–11). A temperature correction algo-
ference and, thus, is preferred to fluorescent rithm is used in some electrophysiology labora-
lighting for rooms in which nerve conduction tories to correct for differences in limb temper-
studies are performed. ature. These algorithms tend to be inaccurate.
Although time-consuming, it is preferable to
sufficiently warm a cool limb before perform-
Submaximal Stimulation ing the studies.
The primary variable of interest when record-

ing SNAPs is the number of axons depolarized. Distance Between Recording
The fact that the amplitude of the response is
directly proportional to the number of axons Electrode and Nerve
stimulated allows one to estimate clinically if
the number of sensory axons is normal or The magnitude of an electrical field will
reduced. In order to accurately assess the num- decrease proportional to the square of the dis-
ber and integrity of all of the sensory axons tance from the electrical generator. As the G1
within a nerve, all fibers must be depolarized. electrode is merely recording the strength of
Providing adequate stimulus intensity is nec- the electrical field at that location, the further
essary to ensure accurate assessment of the the electrode is from the nerve (the genera-
response. tor) the lower the amplitude of the waveform
recorded. For most standard sensory studies,
this is not a critical factor. However, in some
obese patients or those with exceptionally large
Recording and Stimulating hands or digits, the increased distance between
Distance electrode and nerve may affect the amplitude
of the response. In less commonly performed
Because the conduction velocity of sensory sensory conduction studies, for example, lat-
axons varies markedly, the longer the conduct- eral and medial antebrachial cutaneous nerves
ing segment the more the responses tend to and the saphenous nerve, careful attention to
A 2 ms
Temp: 24.8°
33.0 mA
A
1 20 μV
A
2 17.3 mA
20 μV
N: 4
B 2 ms
Temp: 35.1°
A 33.0 mA
1 20 μV
N: 5
A
33.0 mA
2
20 μV
N: 5
Figure 17–11. Effect of limb temperature on the sensory nerve action potentials. Note the higher amplitude, larger area
responses with longer latencies in the cool limb (A) when compared to the warm limb (B).
anatomical landmarks is required to ensure can change the conduction velocity calcula-
that the recording is made from the same tion. Recall that relatively short distances are
location each time, because even the slightest used to calculate conduction velocity. There-
movement can affect the distance between the fore, standard positioning of the limb and stan-
electrode and the nerve and, thus, affect the dard anatomical landmarks need to be used in
amplitude. every study.
Measurement Interelectrode Distance

Human error can affect accurate recording of The interelectrode distance is important when
SNAP parameters. These are perhaps the easi- performing SNAP studies. The G1 and G2
est factors to correct and to control. Measuring electrodes need to be far enough apart to
distance is a source of human error. The dis- prevent phase cancellation and a reduction
tances that are chosen affect the amplitude, in the amplitude of the SNAP7 (Fig. 17–12).
latency, and conduction velocity, and careful However, placing the electrodes too far apart
attention has to be paid to measuring distances increases the problem with electrical inter-
because an error of even a few millimeters ference by reducing the effectiveness of the
recording electrodes, and any deviation from
10 μV
these locations with the recording electrodes
1 ms will alter the response amplitudes noticeably.
Because the amplitude of the response is
inversely proportional to the square of the dis-
tance between G1 and the nerve, an error of
even a few millimeters will alter the amplitude
markedly (Fig. 17–13). Also, stimulating the
nerve from a distance requires excessive cur-
rent, which can cause excessive shock artifact,
nonspecific excitation of other nerves, or direct
muscle stimulation, leading to problems with
volume conduction.
Figure 17–12. Sensory nerve action potential with G1 Key Points

and G2 separated by, top, 3.5 cm, and, bottom, 1 cm. Note
the decrease in amplitude. Bar markers indicate measure-
• Shock artifact can be reduced by main-
ments for latency and amplitude. taining proper impedance with conduct-
ing paste and gentle skin abrasion, or by
rotating the stimulator.
common-mode rejection. The interelectrode • Submaximal nerve stimulation may lead to
distance must be kept constant. a falsely low SNAP amplitude.
Placement of the recording and stimulating • Cool limb temperatures have the effect
electrodes needs to be standardized. In most of prolonging the distal latencies, slowing
nerve conduction studies, specific landmarks the conduction velocity, and increasing the
are used for placing the stimulating and the amplitudes.
• Standard electrode placement and G1–G2
separation should be used to record the
maximal SNAP.
• Careful limb measurements are necessary
10 μV
for accurate interpretation of the SNAP

1 ms parameters.
PLANNING THE STUDY AND

FINDINGS IN DISEASES
The specific sensory nerves to be tested in a
patient must be selected on the basis of the
clinical findings and the differential diagno-
sis of the presenting complaints. In general it
is advisable to test nerves supplying the sen-
sory distribution that are clinically affected.
Although this seems intuitive, often elec-
tromyographers are tempted to restrict testing
to the more common nerves because these
nerves are more familiar and less complicated.
For example, in a patient who complains of
numbness in the dorsum of the foot, selecting
the superficial peroneal sensory nerve, rather
Figure 17–13. Sensory nerve action potential with, top than the sural, would be more appropriate.
to bottom, G1 and G2 moved away from the nerve at 5-mm
increments. Note the decrease in amplitude and prolonga-
The temptation of only studying the most com-
tion of the rise time. Bar markers indicate measurements mon sensory nerves should be avoided because
for latency and amplitude. testing these less familiar nerves can provide
invaluable localizing information that often sensory nerve conduction studies are not reli-
cannot be obtained otherwise. This requires able and there is no good sensory study to
knowledge of the techniques unique to these exclude a preganglionic lesion of the L3 or
nerves and the technical problems that tend L4 roots from a postganglionic lesion of the
to occur with these studies. The importance of lumbar plexus.
maintaining skill in this area cannot be overem- The superficial peroneal sensory, sural sen-
phasized. Retaining these skills requires that sory, and saphenous sensory nerves are pure
these less familiar sensory nerves be tested reg- sensory nerves, and anatomical landmarks have
ularly, not only once or twice a year. Some been established for several techniques for
examples of these less familiar sensory nerves stimulating and recording from these nerves.
are the superficial peroneal sensory nerve and However, the anatomical location of these
the medial and lateral plantar nerves in the nerves varies, and amplitudes vary significantly
lower extremities, the lateral and medial ante- from person to person. Also, for each of these
brachial nerves, and the dorsal ulnar cutaneous nerves, the normal amplitude values diminish
nerve in the upper extremity. Testing of each of with age, and the amplitudes become increas-
these nerves in the appropriate clinical setting ingly difficult to obtain. Because of this, it
can add substantially to the quality of the study. is important to compare the responses with
those of the opposite side in any case in which
responses cannot be obtainable or the ampli-
Radiculopathy tude is equivocal for a person of that age.
Normally, the side-to-side asymmetry may be
Cervical and lumbar radiculopathies are among as much as 50%.9 These less common SNAP
the most common diagnoses of patients studies should be conducted for the common
referred to the electrophysiology laboratory. referral diagnoses in order to maintain the skill
SNAPs are most useful in confirming that the needed to perform such tests with confidence
lesion is preganglionic (i.e., intraspinal). The and to obtain valid and reliable results.
most common lumbar radiculopathies are at
the L5 and S1 levels, followed by the L4 and
L3 levels. Because a peroneal neuropathy may Plexopathy
mimic an L5 radiculopathy, one should con-
sider studying the superficial peroneal sensory The selection of nerves to be tested in a per-
nerve. As previously mentioned, the dorsal root son with suspected plexopathy should be based
ganglia lay within the intervertebral foramina, on the most likely localization determined on
and an intraspinal lesion in general does not routine neurologic examination. In cases of
disrupt the continuity between the cell body brachial plexopathy, the specific site of involve-
and its axon leaving the SNAPs intact. How- ment often cannot be localized on the basis of
ever, the location of the dorsal root ganglia can clinical findings alone. Tailoring the study to
vary. The dorsal root ganglia of lower lumbar the areas of suspected involvement increases
and upper sacral roots may actually be located substantially the yield of the nerve conduction
within the spinal canal in 40% of patients.8 studies. Although brachial plexus lesions can
Thus, in these patients, radiculopathy caused be patchy in distribution, a clinical examina-
by lateral herniation of an intervertebral disk tion often suggests one of three patterns: upper
may cause axonal damage peripheral to the trunk/lateral cord, middle trunk/posterior cord,
dorsal root ganglion, reducing the amplitude or lower trunk/medial cord. In the upper
of SNAPs. This has been demonstrated in trunk/lateral cord distribution, the lateral ante-
the superficial peroneal sensory nerve in L5 brachial cutaneous sensory nerve needs to be
radiculopathies.8 In an S1 radiculopathy, the studied in addition to the median nerve. The
sural sensory nerve should be selected for lateral antebrachial cutaneous sensory nerve
testing because it is located within the der- represents the termination of the musculocu-
matomal distribution of the S1 nerve root. taneous nerve and, in all cases, is a branch
Because a femoral neuropathy may mimic an from the upper trunk and lateral cord. If a mid-
L3 or L4 radiculopathy, the saphenous sensory dle trunk/posterior cord lesion is suspected, a
nerve could be considered to exclude a post- superficial radial sensory response in addition
ganglionic lesion. Unfortunately the saphenous to the median sensory response will enable a
more complete assessment of the cutaneous have been developed14 and are described in
distribution from this segment of the brachial Chapter 18. In some normal subjects over the
plexus. If a lower trunk/medial cord lesion age of 60 the lower extremity SNAPs may not
is suspected, a medial antebrachial cutaneous be elicitable. In older subjects with a low-
nerve study in addition to an ulnar sensory amplitude or absent lower extremity SNAP, the
nerve study is necessary to adequately assess response should be compared to that of the
the cutaneous distribution of the lesion. As contralateral limb.
with some sensory nerves in the lower extrem- Asymmetry of the amplitudes of greater than
ity, these uncommon nerve studies become 50% should be considered abnormal. If the
increasingly difficult to perform the older the SNAP is absent bilaterally then no conclusions
patient is, and side-to-side comparisons should can be drawn. Localization of a lumbar plex-
be made for any responses that cannot be opathy often relies on the findings of needle
obtained or have an equivocal amplitude. electromyography.
Using the median SNAP recorded from the
index finger is unreliable in localizing a lesion
of the brachial plexus.10 The median SNAP Common Mononeuropathies
from the index finger is subject to frequent
anatomical variation. In approximately 80% Median and ulnar neuropathies are among the
of cases it is derived from the middle trunk most common diagnoses referred to the elec-
of the brachial plexus and in the remaining trophysiology laboratory. A median mononeu-
20% from the upper trunk.11 More reliable ropathy at the wrist (carpal tunnel syndrome)
localization can be performed by studying the can be easily identified via a prolongation of
lateral antebrachial cutaneous nerve (which the median SNAP distal latency across the
arises from the upper trunk and lateral cord carpal tunnel. This is the most commonly iden-
in all cases); medial antebrachial cutaneous tified abnormality of the median nerve. Several
nerve (which arises from the lower trunk and techniques have been described that assess
medial cord in all cases); and the superfi- slowing of conduction in the median nerve
cial radial sensory nerve (which arises from at the wrist. Two techniques are used almost
the posterior cord in all cases).11 Unfortu- exclusively in clinical practice: the median sen-
nately there is no reliable sensory nerve in the sory antidromic technique (stimulating at the
distribution of the middle trunk. Distinguish- wrist and recording from the index finger) and
ing between an upper trunk and lateral cord median sensory orthodromic technique (stim-
or the lower trunk and medial cord cannot ulating the median nerve in the palm and
be reliably done with SNAPs and must rely recording over the wrist).
upon the needle electromyography findings. In the antidromic technique, the record-
Interestingly, the SNAPs may not be helpful ing site is over one of the digits supplied by
in cases of idiopathic inflammatory brachial the median nerve, commonly the second digit
neuritis.12 This syndrome most commonly (index finger), and the stimulation sites are
affects multiple mononeuropathies rather than proximal at the wrist and at the elbow. The
discrete sections of the brachial plexus. It has advantage of this technique is that the ampli-
a predilection to affect motor predominate tudes are more reliable. However, because the
nerves such as the anterior interosseous, long antidromic technique involves a longer dis-
thoracic, suprascapular, and phrenic nerves.13 tance, it is less sensitive to subtle slowing of
In lumbosacral plexopathy, the anatomical conduction across the wrist and therefore less
patterns are not as discrete as they are in sensitive to mild cases of carpal tunnel syn-
the upper limb. Clinically, lumbosacral plex- drome. This antidromic technique is usually
opathies often can be divided into two dis- applied to more severe cases of carpal tunnel
tribution patterns: lumbar plexus and sacral syndrome where the median motor responses
plexus. In most cases, the sacral plexus can are already abnormal.
be sampled with the sural and superficial per- In milder cases of carpal tunnel syndrome,
oneal sensory nerves. Reliable techniques have the orthostatic or palmar technique is pre-
not been developed to sample the cutaneous ferred. In the palmar (orthodromic) tech-
branches in the lumbar plexus. Techniques for nique, the stimulation site is the palm and the
dermatomal somatosensory evoked potentials recording sites are proximal to the wrist and
at the elbow. This technique is more sensi- show slowing across the wrist, the carpal tun-
tive for focal slowing because the distal dis- nel syndrome will be more severe and the
tance is shorter. The sensitivity can be further antidromic technique, which is easier to per-
improved by comparing the median palmar form and more reliable than palmar ortho-
sensory distal latency to the analogous palmar dromic studies, should be performed. If there
sensory study of the ulnar nerve. If stimulated is no slowing of conduction across the wrist
and recorded over the same distance (typically in the motor studies then the carpal tunnel
8 cm) then the difference in the distal latencies syndrome will be mild and the more sensitive
between these two nerves should be 0.3 ms or palmar orthodromic technique should be used.
less. Differences in latencies larger than this Various grading scales have been developed
indicate focal slowing in the median nerve and to quantify severity of the median mononeu-
in the proper setting diagnostic of a median ropathy at the wrist using nerve conduction
mononeuropathy at the wrist (i.e., carpal tun- studies.15
nel syndrome). One additional advantage is Ulnar neuropathies occur most commonly at
that the proximal and distal recordings can the elbow; however, definitive localization of
be obtained with one set of electrical stimula- the lesion often can be difficult, especially the
tions, decreasing the number of shocks com- more chronic the condition. The most common
pared with that of the antidromic technique. technique for studying the ulnar sensory nerve
However, with the orthodromic technique, the is the antidromic method, with the recording
distribution of axons activated at supramaxi- site over the 5th digit and the stimulation sites
mal stimulation is more variable, resulting in proximal at the wrist and above the elbow.
a less reliable amplitude response. In these The palmar orthodromic technique used in
mild cases of carpal tunnel syndrome the dis- studying median SNAPs is not often used to
tal latency is more relevant than the amplitude assess ulnar neuropathy, because it does not
(Fig. 17–14). appreciably increase the sensitivity of the ulnar
When a case of carpal tunnel syndrome is studies.
suspected, the median motor nerve is often Conduction block in the ulnar nerve across
tested first, which helps guide the decision of the elbow cannot be proven reliably with
which median sensory conduction would be sensory studies, but if conduction block is
most useful to perform. If the motor responses present, the proximal sensory amplitudes may
be much reduced or even unobtainable. In the
case of pure conduction block, the responses
elicited with distal stimulation may demon-
strate normal amplitudes and latencies. Stim-
20 μV
1 ms ulation above and below the elbow may be

helpful in demonstrating focal slowing of con-
duction across the elbow by isolating the slow-
ing across a much shorter segment. This may
be further refined using inching techniques.
If there is no demyelination and only axonal
injury, the only finding on an ulnar antidromic
sensory study may be a reduction in ampli-
tude, a finding that is not helpful in localization.
If this is the case, additional sensory studies
10 μV
1 ms can be used to further localize the lesion. For

example, a lesion can be localized to a site
above the wrist if the dorsal ulnar cutaneous
nerve is abnormal, since it branches from the
main ulnar nerve in the forearm. However, the
reliability of dorsal ulnar cutaneous responses
Figure 17–14. Example of median (upper tracing) and
ulnar (lower tracing) palmar sensory studies in mild is less than that of other nerves more com-
median mononeuropathy at the wrist. Note that the dif- monly tested and this nerve should be com-
ference in the peak latencies between the two studies is pared with that on the contralateral side.16
nearly 1 ms. Occasionally, a lower trunk or medial cord
brachial plexopathy may appear clinically sim- sensory involvement in addition to its primary
ilar to an ulnar neuropathy. If both the ulnar manifestation as a presynaptic defect of neu-
antidromic study and the dorsal ulnar cuta- romuscular transmission. Botulism primarily
neous sensory nerves are abnormal, the medial affects the bulbar region and may demonstrate
antebrachial cutaneous nerve should be tested. alteration of the blink reflexes. Whether this
Abnormal findings in this nerve suggest a more is due to involvement of the afferent sensory
proximal lesion of the medial cord or lower trigeminal nerve or the efferent facial motor
trunk of the brachial plexus. As with the dorsal nerve is not clear.17 Abnormalities within the
ulnar cutaneous nerve, the lateral antebrachial limb SNAPs have not been well described.
cutaneous nerve should be compared with that
of the contralateral side.
Peroneal, sciatic, and femoral neuropathies Motor Neuron Diseases
are common diagnoses referred to the electro-
physiology laboratory. The sensory nerves to be The most common motor neuron disease
tested for each of these mononeuropathies are encountered in clinical practice is amyotrophic
the superficial peroneal, sural, and saphenous lateral sclerosis (ALS). The primary defect in
nerves, respectively. The antidromic technique ALS is degeneration of the motor neurons in
is typically used. If the lesion causes a pure con- the motor cortex and the anterior horn of the
duction block above the level of stimulation, spinal cord. The SNAP amplitudes remain nor-
the SNAP may be normal. This is not uncom- mal in most cases of ALS. There have been
mon in a peroneal neuropathy at the head of recent reports suggesting that the sural SNAPs
the fibula. If the amplitude in any of these may be mildly reduced in a minority of cases.18
sensory studies is abnormal or borderline, the Like ALS, spinal muscular atrophy also results
contralateral side should be tested. in progressive loss of the motor neurons in
the anterior horn of the spinal cord. As in
ALS, the SNAP remain normal. Spinal and
Myopathy bulbar muscular atrophy (SBMA), also known
as Kennedy’s syndrome, clinically presents as
Sensory nerve action potentials are typically a progressive lower motor neuron syndrome
unaffected in most myopathies. Exceptions which may strongly resemble ALS. SBMA is
to this include disorders with multisystem an X-linked disorder, occurring only in men,
involvement where the peripheral nerves, mus- caused by a trinucleotide repeat in the andro-
cle, and often other organ systems are involved gen receptor gene. In addition to the pathol-
simultaneously. This may be seen in systemic ogy in the lower motor neurons, SBMA also
inflammatory disorders such as the mixed con- affects the sensory neurons within the dorsal
nective tissue disorders and vasculitis. Other root ganglia. While the lower motor neuron
disorders than can affect the peripheral sensory findings dominate the clinical presentation of
nerves (and therefore the SNAPs) in addition SBMA the involvement of the dorsal root gan-
to muscle include some congenital metabolic glia neurons is known to reduce the SNAP
disorders, such as mitochondrial diseases, sys- amplitude in a majority of those affected.19, 20
temic hematological disorders such as amy- This involvement of the SNAPs may be the
loidosis, and certain environmental toxins and only neurophysiological feature to distinguish
medications (e.g., hydroxychloroquine). SBMA from ALS.
Key Points
Disorders of the Neuromuscular • SNAP testing needs to be performed in
Junction the distribution of the sensory deficits.
• The SNAP remains preserved in radicu-
Most disorders of neuromuscular junction lopathies.
transmission (autoimmune myasthenia gravis, • The SNAP is abnormal in clinically
Lambert–Eaton myasthenic syndrome and the affected distributions in plexopathies.
congenital myasthenic syndromes) do not • SNAPs are very sensitive at detecting
affect the SNAP. Botulism can present with mononeuropathies.
• SNAPs are normal in common myopathies 2. Buchthal, F., and A. Rosenfalck. 1971. Sensory poten-
and disorders of the neuromuscular tials in polyneuropathy. Brain 94:241–62.
junction. 3. Handwerker, H. O. 1996. Sixty years of C-fiber record-
ings from animal and human skin nerves: Historical
• SNAPs are normal in ALS and spinal mus- notes. Progress in Brain Research 113:39–51.
cular atrophy; abnormal SNAP in a male 4. Albers, J. W. 1993. Clinical neurophysiology of gener-
patient with motor neuron disease should alized polyneuropathy. Journal of Clinical Neurophys-
raise suspicion for SBMA. iology 10:149–66.
5. Melendrez, J. L., L. J. MacMillan, and J. Vajsar. 1998.
Amplitudes of sural and radial sensory nerve action
potentials in orthodromic and antidromic studies in
SUMMARY children. Electromyography and Clinical Neurophysi-
ology 38:47–50.
6. Wilbourn, A. J. 1994. Sensory nerve conduc-
Nerve conduction studies are an invaluable tion studies. Journal of Clinical Neurophysiology
addition to clinical electrophysiology testing. 11:584–601.
SNAPs are a sensitive and specific measure of 7. Wee, A. S., and R. A. Ashley. 1990. Effect of inter-
function in the peripheral sensory pathways. electrode recording distance on morphology of the
These studies confirm whether large myeli- antidromic sensory nerve action potentials at the fin-
ger. Electromyography and Clinical Neurophysiology
nated axons are affected by an underlying 30:93–6.
abnormality. When an area that is affected is 8. Levin, K. H. 1998. L5 radiculopathy with reduced
clinically tested, nerve conduction studies can superficial peroneal sensory responses: Intraspinal and
help to distinguish between a preganglionic extraspinal causes. Muscle & Nerve 21:3–7.
9. Bromberg, M. B., and L. Jaros. 1998. Symmetry of
(i.e., root level or higher) and a postganglionic normal motor and sensory nerve conduction measure-
(i.e., peripheral) process. ments. Muscle & Nerve 21:498–503.
SNAPs are small and technically difficult 10. Moran, S. L., S. P. Steinmann, and A. Y. Shin. 2005.
to record; therefore close attention has to be Adult brachial plexus injuries: Mechanism, patterns of
given to proper technique, including minimiz- injury and physical diagnosis. Hand Clinics 21: 13–24.
11. Ferrante, M. A., and A. J. Wilbourn. 2002. Elec-
ing electrical interference and using proper trodiagnostic approach to the patient with suspected
stimulating and recording methods. When brachial plexopathy. Neurological Clinics of North
stimulating a nerve, stimulate as close to the America 20: 423–50.
nerve as possible to allow a supramaximal 12. Puri, V., N. Chaudhry, K. K. Jain, D. Chowdhury,
and R. Nehru. 2004. Brachial plexopathy: A clinical
response but to minimize the electrical arti- and electrophysiological study. Electromyography and
facts. When recording, adhere to fixed anatom- Clinical Neurophysiology 44:229–35.
ical landmarks, maintain proper interelectrode 13. Jillapalli, D., and J. M. Shefner. 2005. Electrodiag-
distance, and be attentive in measuring stim- nosis in common mononeuropathies and plexopathies.
ulating and recording distances. If satisfactory Seminars in Neurology 25(2):196–203.
14. Dumitru, D., and P. Dreyfuss. 1996. Dermatomal/
responses cannot be obtained with transcuta- segmental somatosensory evoked potential evaluation
neous stimulation, consider near-needle stim- of L5/S1 unilateral/unilevel radiculopathies. Muscle &
ulation in the appropriate setting. Because Nerve 19:442–9.
the amplitudes are low, averaging responses 15. Bland, J. D. 2000. A neurophysiological grading
scale for carpal tunnel syndrome. Muscle & Nerve
enhances the signal and eliminates the random 23:1280–3.
ambient electrical activity. The measurements 16. Dutra de Oliveira, A. L., A. A. Barreira, and
that are most relevant clinically and should be W. Marques Jr. 2000. Limitations on the clinical util-
noted on all studies are the distal amplitude, ity of the ulnar dorsal cutaneous sensory nerve action
distal peak latency, and conduction velocity of potential. Clinical Neurophysiology 111:1208–10.
17. Hatanaka, T., K. Owa, M. Yasunaga, et al. 2000.
the nerve. It must be emphasized that appro- Electrophysiological studies of a child with presumed
priate technique is needed to produce reliable botulism. Childs Nervous System 16(2):84–6.
and valid results. The interpretations drawn 18. Hammad, M., A. Silva, J. Glass, J. T. Sladky, and
from the results of a study can be only as good M. Benatar. 2007. Clinical, electrophysiologic, and
pathologic evidence for sensory abnormalities in ALS.
as the information on which they are based. Neurology 69:2236–42.
19. Polo, A., F. Teatini, S. D’Anna, et al. 1996. Sensory
involvement in X-linked spino-bulbar muscular atro-
REFERENCES phy (Kennedy’s syndrome): An electrophysiological
study. Journal of Neurology 243(5):388–92.
1. Donofrio, P. D., and J. W. Albers. 1990. AAEM 20. Antoni, G., F. Gragnani, A. Romaniello, et al. 2000.
minimonograph #34: Polyneuropathy: Classification Sensory involvement in spinal-bulbar muscular atro-
by nerve conduction studies and electromyography. phy (Kennedy’s disease). Muscle & Nerve 23(2):
Muscle & Nerve 13:889–903. 252–8.
Chapter 18
Somatosensory Evoked Potentials

Jonathan L. Carter and J. Clarke Stevens
INTRODUCTION Factors That Affect the Amplitude

GENERATORS AND ORIGIN OF SEPs and Latencies of the Evoked Response
Neuroanatomic Sites of Origin of
LOCALIZATION
Potentials
Latency Prolongation
METHODS Amplitude Reduction
Nerve Stimulation Variables CLINICAL APPLICATIONS
RECORDING Disorders of the Peripheral
Methods and Montages Nervous System
Volume Conduction and Disorders of the CNS
Near-Field and Far-Field Potentials SEP Findings in Brain Death
Averaging SEPs Recorded in the Intensive
Peak Nomenclature Care Unit
SEP Interpretation SUMMARY
INTRODUCTION Although evoked potentials can be elicited by

physiologic stimuli such as a finger tap or
Somatosensory evoked potentials (SEPs) are tendon stretch, electric stimulation produces
presynaptic and postsynaptic responses consistently higher amplitude evoked poten-
recorded over the limbs, spine, and scalp fol- tials and is the only type of stimulus useful for
lowing the stimulation of peripheral mixed clinical application.
motor and sensory nerves or cutaneous sen- SEPs are used to evaluate the central
sory nerves. SEPs are analogous to standard somatosensory pathway. One of the most fre-
peripheral sensory nerve conduction studies, quent applications is in the evaluation of
although they assess both the peripheral and patients with suspected multiple sclerosis, to
the central somatosensory conduction path- obtain evidence for a clinically or subclinically
ways. Since routine sensory nerve conduc- evident lesion. They also provide objective evi-
tion studies are easier to perform and are dence of central nervous system (CNS) dys-
more reliably recorded than SEPs, the main function when sensory symptoms are vague
value of SEPs is to provide a measurement and the findings on neurologic examination are
of sensory conduction in proximal periph- normal or of uncertain significance. Spinal cord
eral nerves, the spinal cord, and the brain. dysfunction due to causes other than multiple
257
sclerosis is another common indication for the medulla through the brain stem to the ven-
recording SEPs. The localization of sensory tral posterolateral nucleus of the thalamus. The
symptoms to a proximal peripheral nerve, the axons of third-order neurons go from the ven-
spinal cord, or a cerebral site is helpful in diag- tral posterolateral nucleus of the thalamus to
nosis and can suggest where an imaging study the primary somatosensory cortex. The travel-
may show abnormality. SEPs are normal if sen- ing volley of the action potentials propagating
sory complaints are caused by a conversion along these pathways, or the responses gen-
reaction. Magnetic resonance imaging (MRI) erated at the sites of the synapses or within
of supratentorial structural lesions is so sensi- the sensory pathway nuclei may be recorded at
tive and reliable that SEPs are rarely indicated different sites in the limb, spine, and scalp.
for investigation of this area of the CNS. Stimulation of cutaneous sensory nerves or
dermatomes activates large cutaneous affer-
Purpose and Role of SEPs ents with similar anatomic origins as mixed
nerve stimulation, but the SEPs have a lower
• To assess the integrity of the peripheral amplitude than those produced by stimula-
and central somatosensory pathways. tion of mixed nerves because fewer fibers are
• To identify abnormalities in the spinal excited. Also, the peak latencies obtained with
cord, brain stem, or cortex in diseases such the stimulation of sensory nerves are slightly
as multiple sclerosis or cervical stenosis. longer than those obtained when a mixed nerve
• To provide objective evidence of CNS is stimulated, because fast-conducting group Ia
dysfunction when sensory symptoms are muscle afferents are not present in a sensory
vague and the neurologic examination is nerve. Lesions affecting sensory modalities
normal. transmitted by small-diameter sensory fibers
• To assist in localization of sensory symp- or by central pathways in the ventral half of
toms to proximal peripheral nerve (roots), the spinal cord usually do not produce SEP
spinal cord, or cerebral site. abnormalities. Pain-related SEPs have been
recorded by stimulation of small-diameter Aδ
pain fibers with a carbon dioxide laser and ther-
GENERATORS AND ORIGIN mal stimulation, but the acceptance of these
OF SEPs techniques has been limited.2, 3
Key Points
Neuroanatomic Sites of Origin of • The potentials recorded during SEP stud-
Potentials ies represent activity in the proprioceptive
sensory system.
The neuroanatomy of the sensory pathways is • The responses are conducted by large
well known; however, the exact origin of many diameter, myelinated, fast-conducting
of the SEP waveforms is still not clear and it is fibers in the periphery and by the dorsal
evident that some have overlapping generator column-medial lemniscus and spinocere-
sources.1 The potentials recorded with low- bellar pathways in the CNS.
current stimulation of a mixed nerve repre- • Responses obtained with stimulation of
sent activity in the proprioceptive system, con- cutaneous sensory nerves or dermatomes
ducted peripherally by large-diameter, myeli- have a lower amplitude than those with
nated, fast-conducting cutaneous and muscle mixed nerve stimulation.
afferents and conducted centrally by the dorsal
column-medial lemniscus and the spinocere-
bellar pathways. There are numerous collat-
erals to the gray matter at all levels. The METHODS
first-order axons contain the fibers within
the peripheral nerves with the cell bodies Nerve Stimulation Variables
lying in the dorsal root ganglion, as well as
axons extending centrally within the spinal Somatosensory evoked responses are typi-
cord to the cuneate or gracile nuclei in the cally obtained with electrical stimulation of a
medulla. The second-order axons travel from peripheral nerve, such as at the wrist or ankle.
Somatosensory Evoked Potentials 259
Stimulating electrodes are fixed in place with median nerves. In the lower limb, stimulation
an elastic strap. The stimulus applied should of the tibial nerve produces more reliable SEPs
be of sufficient intensity to produce a small than stimulation of the peroneal nerve. The
visible twitch of the muscle. Stimulus inten- lowest amplitudes are obtained with stimula-
sities higher than this are painful (making it tion of cutaneous or dermatomal nerves. After
difficult for the patient to relax) and do not cutaneous nerve stimulation, spinal poten-
increase the amplitude of the SEP. Intensities tials usually are absent in normal subjects;
twice the motor threshold are sometimes nec- thus, disease in the CNS cannot be distin-
essary to achieve maximal central responses. If guished from that in the peripheral nervous
the small foot muscles are atrophied, a twitch system.
may not be visible except at high stimulus
intensities. In this case, use a stimulus that is
2–2.5 times sensory threshold. If sensory loss UNILATERAL AND BILATERAL
is marked, higher stimulus intensities can be STIMULATION
used without causing patient discomfort and Tibial spinal, brain stem, and cerebral evoked
may be needed to exceed threshold. In the potentials increase in amplitude with bilateral
upper limb, stimulus rates of 2–5 Hz are well nerve stimulation (Fig. 18–1). Therefore, bilat-
tolerated by most patients, but rates of 1–2 Hz eral stimulation is helpful when the tibial scalp
are used in the lower limb. A slower rate of SEPs are absent or poorly formed or when
0.5 Hz may be required to avoid a flexor with- spine potentials are absent. In this case, bilat-
drawal reflex in a spastic limb. Rates greater eral stimulation may produce adequate subcor-
than 10 Hz may cause an increase in the latency tical potentials and allow central conduction
and a decrease in the amplitude of some time or at least the absolute scalp latencies to
components. be estimated. The response to bilateral stimu-
lation may not be enhanced if the peripheral
and central nerve conduction velocities are not
NERVE STIMULATED
similar on the two sides. A limitation of bilat-
In the upper limb, the highest amplitude SEPs eral tibial nerve stimulation is that a unilateral
are obtained with stimulation of the ulnar and lesion may be more difficult to identify, since
P38 10 ms
Cz-Fz
2 μV
N30 10 ms
C5S-Fz
2 μV
N22
10 ms
L1S-IC
N8 2 μV
10 ms
Knee-IC
2 μV
10 ms
2 μV
Cz-Fz P38
N30 10 ms
2 μV
C5S-Fz
N22 10 ms
2 μV
L1S-IC 10 ms
N8 5 μV
Knee-IC
10 ms
2 μV
Figure 18–1. Tibial nerve stimulation. Top, unilateral and bottom, bilateral stimulation at the same amplification in the
same patient. Note that bilateral stimulation enhances the amplitudes of N22, N30, and P38.
the conduction along the bilateral somatosen- RECORDING

sory pathway may mask a unilateral lesion.
Bilateral median nerve stimulation is rarely Methods and Montages
necessary.
SEPs are typically recorded using 5-mm tin
Key Points disc electrodes taped or pasted on to the skin
surface at the different recording sites. Scalp
• Stimulation is performed with a cutaneous electrodes are fixed with collodion, and spinal
stimulator, fixed over the nerve with an and Erb’s point electrodes are taped in place.
elastic strap. Responses are recorded over the peripheral
• The stimulus intensity should produce a nerve (median, ulnar, or tibial nerve and Erb’s
small twitch of the muscle or be 2–2.5 point), lumbar or cervical spine, and scalp.
times the sensory threshold. Bipolar electrodes have a fixed interelectrode
• Stimulus rates of 2–5 Hz are used in the distance of 35 mm. Electrolyte gel is applied
upper limb and 1–2 Hz in the lower limb. to all electrodes, and impedance is maintained
• The highest amplitude SEPs are obtained less than 5 k. The montages used in Mayo
with ulnar, median, or tibial nerve stimu- neurophysiology laboratories to study SEPs are
lation. listed in Table 18–1. Different montages are
• Bilateral tibial nerve stimulation may used in different electrophysiology laborato-
enhance the identification of subcorti- ries. Despite differences between electrophys-
cal peaks. iology laboratories in the number of channels
Table 18–1 Standard Methods for Recording SEPs

Median or Ulnar Nerve Tibial Nerve
Stimulation Bipolar at wrist, cathode proximal Bipolar at ankle, cathode proximal

Standard recording, Bipolar at elbow: N5 Bipolar at knee: N8
montage: potential EPi–EPc: N9 L1–ICc: N22
C5S–Fz: N11, N13, N14 C5S–Fz: N30
C3 (C4 )–Fz: N20/P25 Cz–Fz: N33/P38
Ground: arm Ground: leg
Optional recording Anterior neck-C5S: P13 C3 –C4 : N33/P38
C3 (C4 )-noncephalic:
P9, P13, P14 far-field
N20/P25 parietal
F3 (F4 )-noncephalic:
N22/P30 frontal
Machine settings Stimulus rate, 1–5 Hz Stimulus rate, 0.5–2 Hz
Stimulus intensity, slight muscle twitch Stimulus intensity, slight muscle twitch
Amplifier sensitivity Amplifier sensitivity
10 μV spine and scalp 10 μV spine and scalp
20 μV peripheral 20 μV peripheral
Analysis time, 50 ms Analysis time, 80 ms
Filter, 30–3000 Hz Filter, 30–3000 Hz
Number averaged, 500 Number averaged, 500
Measurement Height and F-distance Height and F-distance
Limb temperature Limb temperature
Waveform amplitude Waveform amplitude
N5, N9, N13: onset-peak N8, N22, N30: onset-peak
N20: N20–P25 N33: N33–P38
P25: P25–N35 P38: P38–N46
EPi–EPc, Erb’s point ipsilateral to contralateral; C5S, spine of C5 vertebra; L1–ICc, spine of L1 vertebra to contralateral
iliac crest.
and recording montages used, the major recog- Nerve action potentials that travel along nerves
nized potentials are consistent. Helpful guide- or fiber tracts are called traveling waves.
lines for conducting SEP studies are available Potentials that remain localized in areas of
from the American Association of Neuromus- nuclei or synapses are called stationary waves.
cular and Electrodiagnostic Medicine.4 Near-field potentials (NFPs) represent a
propagating nerve action potential that is
recorded as it passes under the recording elec-
Volume Conduction and trodes. The recording electrodes 3 cm apart
Near-Field and Far-Field Potentials that are used in routine nerve conduction stud-
ies primarily record NFPs. The term far-field
Volume conduction principles have important potential (FFP) refers to stationary potentials
implications to the recorded somatosensory generated by nerve action potentials distant
responses. The principles of volume conduc- to the recording site.5 A referential montage,
tion are discussed in Chapter 3. In a volume such as the scalp to Erb’s point, preferen-
conductor, the amplitude of the potential is tially detects FFPs. The advantage of using
related inversely to the square of the distance far-field recordings is that information from
between the generator and the recording point. many different levels of the nervous system can
If the recording electrode is close to the gener- be obtained from a single recording montage.
ator, a second electrode that is a long distance The disadvantage is the excess noise intro-
away will act as an indifferent electrode and a duced by long interelectrode distances, which
high amplitude potential will be obtained. If makes it difficult to obtain accurate latency
the recording electrode is far from the source, measurements.
another electrode at a similar distance will be
almost as active as the recording electrode
itself, in which case the potentials of the two Averaging
electrodes will cancel in a differential amplifier
so that little or no potential will be recorded. The main technical limitation to recording
All electrodes—regardless of where they SEPs is that their amplitude is low compared
are placed on the body—are relatively active. with the noise of motor activity, movement
Therefore, responses generated anywhere artifacts, the electrocardiogram, the electro-
along the somatosensory conduction pathway magnetic activity in the environment, and the
could be recorded by distant electrodes. This electroencephalogram (EEG). Generally, 500–
is analogous to the electric activity of the heart, 2000 stimuli are necessary to display well-
which can also be recorded anywhere on the defined, reproducible waveforms of 1–10 μV.
body. Ideally, one electrode should be as close Averaging summates activity that is time-
to the generator as possible and the other locked to the stimulus trigger, while gradually
electrode as far away as possible to obtain subtracting random background noise. If the
the maximal potential difference between the noise is excessive, increasing the number of
electrodes. However, increasing electrode dis- stimuli averaged does not help to extract a bet-
tance also increases noise, especially muscle ter signal because the signal-to-noise ratio is
artifact, and may introduce additional gener- too small. For example, artifact in the form of
ators. Therefore, it is necessary to compro- large quasirandom triphasic motor unit poten-
mise between the cancellation effect of closely tials may produce continuous variation in the
spaced electrodes and the noise introduced by averaged waveform that does not improve with
long distance between recording electrodes. If continued averaging (Fig. 18–2). If the noise
the generator is proximal to the shoulder or becomes time-locked with the signal, it will be
hip, moving the reference electrode distally enhanced and may be mistaken for a physio-
along a limb does not improve the signal. logic signal. Sixty-cycle electrical artifact can be
As in peripheral nerve conduction stud- reduced by using a stimulation rate that is not a
ies, a bipolar electrode montage that detects factor of 60 (e.g., 2.1 Hz). Averagers deal with
an approaching or departing depolarization intermittent artifacts by rejecting sweeps that
records a positive waveform, a positivity, and contain waveforms exceeding the fixed maxi-
the electrode overlying the depolarization mal amplitude. This helps to decrease or to
records a negative waveform, a negativity. eliminate most artifacts. In addition, inspection
Figure 18–2. Averaged median sensory nerve action potential (SNAP) at three levels with increasing sample number
(note comparison of input signals with actual potential). A, Input signal to be averaged and the averaged SNAPs at the
wrist and elbow. B, Increasing level of contraction obscures the forearm waveform with low level contraction and makes it
unrecognizable at moderate levels, despite 6000 averages. Even larger numbers of averages at moderate contraction, did not
bring out the forearm waveform. C, The patient is flexing the elbow at a low level of activation showing improvement in
averages in the forearm and elbow with increasing numbers of averages.
of the signal being averaged allows the techni- • Potentials that travel along nerve fibers
cian to interrupt averaging if excessive artifact or tracts are called traveling waves, while
occurs. Averaging is resumed after the techni- those generated in nuclei or synapses are
cal problem has been corrected. called stationary waves.
• NFPs represent a propagating action
potential recorded as it passes under the
Key Points
recording electrodes. FFPs represent sta-
• The amplitude of the SEP responses is tionary potentials generated by action
inversely proportional to the square of the potentials distant to the recording site.
distance between the recording electrode • SEP amplitudes are very low com-
and the generator of the response. pared to other electrical generators in
the body; hence, large number of stim- MEDIAN AND ULNAR MIXED
uli must be averaged to obtain repro- NERVE SEPs
ducible responses. Muscle artifact is a
Following stimulation of the median or ulnar
major technical challenge in recording
nerve at the wrist, activity can be recorded
SEPs.
at the elbow, Erb’s point, cervical spine, and
scalp (Figs. 18–3 and 18–4). Several different
peaks are identified with standard recording
Peak Nomenclature montages: N5, N9, N11, N13, N14, and N20.
N5. The N5 potential recorded with a bipo-
The SEP responses are named according to
lar electrode at the elbow represents the
the direction of peak deflection (P = positive
propagating nerve action potential in the
or N = negative) and the latencies of the peak
median or ulnar nerve. The presence of this
response. By standard convention, upward
potential helps to indicate that the stimulation
deflections are labeled as N (negative) and
is adequate and provides an estimate of periph-
downward deflections as P (positive). The
eral conduction velocity. When SEP latencies
number following the N or P refers to the
are prolonged, a motor conduction study in the
average latency at which the particular poten-
arm or leg (or both) is performed to check for
tial is recorded in normal subjects. Thus,
slowing of peripheral nerve conduction.
with a bipolar montage, the negative poten-
tial that is recorded over the brachial plexus N9. The N9 potential recorded with an elec-
approximately 9 ms after the median nerve trode at Erb’s point (2 cm superior to the
is stimulated at the wrist is termed N9 midpoint of the clavicle) referred to an elec-
(Table 18–1). However, peak nomenclature trode in the same location contralaterally rep-
has not been standardized, and slightly differ- resents orthodromic activity in sensory fibers
ent numbering systems are used to identify and antidromic activity in motor fibers pass-
the same evoked potential. Evoked poten- ing through the brachial plexus. Stimulation
tials are measured only when two superim- of the median nerve activates cutaneous sen-
posed averages reveal consistent responses. sory fibers that enter the spinal cord through
If it is clear that reproducible waveforms the upper and middle trunks and posterior
are present when looking at the two aver- roots of C6 and C7. Antidromic median motor
ages, viewing a single combined average may and spindle afferent potentials pass through
make it easier to identify and to measure the the medial cord and lower trunk of the plexus
waveforms. to enter the spinal cord through the anterior
N20 N36
C4'-Fz P25
N20 5 ms
2 μV
C4'-EPi 5 ms
N18 2 μV
5 ms
C3'-EPi 2 μV
N13 5 ms
2 μV
C5S-Fz N13
5 ms
2 μV
C5S-EP2 N13
5 ms
C5S-AC 2 μV
N9
5 ms
EPi-EPc 2 μV
N5
5 ms
Elbow 10 μV
Figure 18–3. Normal 8-channel median SEPs in a 13-year-old child. AC, anterior cervical electrode placed just above the
thyroid cartilage.
N36
N20
5 ms
C3'-Fz 1 μV
N20 P25
C3'-EPi 5 ms
N18 2 μV
C4'-EPi 5 ms
N13 2 μV
C5Sp-Fz 5 ms
N13 2 μV
C5S-EPi 5 ms
1 μV
N13
C5S-AC 5 ms
1 μV
N9
EPi-EPc 5 ms
N5 2 μV
5 ms
Elbow 20 μV
Figure 18–4. Normal 8-channel ulnar SEPs.
and posterior roots of C8 and T1. With ulnar The N13 potential can be recorded in all nor-
nerve stimulation, activity is confined to the C8 mal subjects, whereas the N11 peak is recorded
and T1 segments. Occasionally, the ulnar N9 in approximately 75% of normal subjects and
potential is difficult to record in normal sub- N14 in approximately 15%–20%. Separation of
jects older than 60 years. Most of the potential the spinal N13/P13 dorsal horn potential from
is generated in sensory fibers because the N9 P14 can be facilitated by recording from the
potential is prominent in patients with avulsion anterior neck at the superior border of the
of the roots of the brachial plexus. Conversely, thyroid cartilage with a contralateral elbow ref-
if a peripheral sensory deficit is substantial, the erence or with a C5S-anterior neck montage.
N9 peak may represent antidromic activity in Loss of N13/P13, but not P14 and N20, may
motor, not sensory, fibers. occur when lesions interrupt collateral axons
to dorsal horn neurons without affecting fibers
N11, N13, and N14. An electrode placed over
ascending in the dorsal columns.
the spine of C5 or C7 referred to Fz is the
most common montage for recording activity In the C5S–Fz montage, N14 is sometimes
arising from the cervical spine and brain stem. seen as a small negative potential on the falling
This montage records three negative poten- phase of N13. N14/P15 potentials probably
tials: N11, N13, and N14. The N11 potential arise in the caudal medial lemniscus because
is likely a presynaptic traveling wave that arises they are preserved in cases of thalamic lesion
from activity near the root entry zone of C6 and tend to be abnormal in cases of brain
and C7 and action potentials ascending in the stem dysfunction.13, 14 The N13–P14 interpeak
dorsal columns. N11 is also referred to as the latency assesses cervical cord–brain stem con-
dorsal column volley (DCV).The evidence is duction time.
convincing that N13 is a standing dipole that
is negative when recorded over the posterior N18. N18 is a broad, subcortically generated
neck and positive when recorded preverte- FFP best recorded in an ipsilateral scalp-
brally.6–9 N13/P13 is a dorsal horn postsynaptic to-noncephalic montage. Evidence points to
potential that is elicited by collaterals of the this potential being postsynaptic activity aris-
primary afferent fibers in the lower cervical ing from several generator sources in the
cord. A second potential with the same latency brain stem.13, 15 Studies of patients with brain
occurs at the level of the cervicomedullary stem lesions suggest that N18 reflects exci-
junction; it possibly arises from the cuneate tatory postsynaptic potentials evoked by dor-
nucleus.10–12 The C5S–Fz montage records a sal column axons in the cuneate nucleus or
large N13 potential that is likely an average accessory inferior olive (or both) or, possi-
of the standing dorsal cord potential and the bly, presynaptic afferent depolarization in the
P13/P14 FFP recorded by the scalp electrode. cuneate nucleus.16–18
N20. Scalp potentials with a latency of 20 ms • N5 represents the propagating nerve

and longer reflect the postsynaptic potentials action potential from the median or ulnar
generated by neurons in the hand area of the nerve at the elbow.
primary somatosensory cortex in response to • N9 represents the propagating nerve
the afferent thalamocortical volley. There is action potential passing through the
disagreement about the exact identity of corti- brachial plexus.
cal generators. Whether the N20/P25 peaks are • N11, N13, and N14 represent activity
mediated by separate thalamocortical projec- in the nerve root entry zone and dorsal
tions or by sequential activity in one pathway columns in the cervical spinal cord.
is uncertain. SEPs may occur with selective • N20 represents potentials generated over
involvement of early or late cortical peaks. This the primary somatosensory cortex.
phenomenon has led to the speculation that
the N20 peak is related to vibration and posi-
tion sense (large myelinated fibers), while the TIBIAL MIXED NERVE SEPs
N35 peak is attributed to pain and temperature Following stimulation of the tibial nerve at the
sense (small myelinated fibers).19 The N20/P25 ankle, potentials are recorded from the leg,
complex, as recorded with the bipolar C3 or lumbar and cervical spine, and scalp.
C4 –Fz montage, may be an average of inde-
pendent posterior frontal (P22/N30) and pari- N8. The peripheral nerve action potential
etal (N20/P30) generators, or may represent recorded at the popliteal fossa is labeled N8
a single generator with electrodes recording (Fig. 18–5).
opposite ends of the dipole generator source.
N22. An electrode over vertebra L1 referred
to the iliac crest records a negative, sometimes
Key Points
bifid, potential designated N22. The N22
• SEP waveform nomenclature is derived represents postsynaptic potentials generated
from the direction of the peak deflec- in the dorsal horn of the spinal cord,
tion (P = positive or downward deflection, analogous to the stationary N13/P13 poten-
N = negative or upward deflection) and tial recorded over the neck with stimula-
the average latency of the response. tion of the median nerve. An initial, small,
• The major clinically important waveforms rarely recorded negative peak, N18, is a
with median nerve stimulation are N5, traveling wave that represents conduction
N9, N11, N13, N14, and N20. through the cauda equina and root.20 It
N46
Cz-Fz P38
P58 10 ms
1 μV
P38
C3'-Fz
10 ms
N38 1 μV
C4'-Fz
N30 10 ms
0.5 μV
C5S-Fz
10 ms
N34 1 μV
Fz-C5S 10 ms
1 μV
N22
10 ms
L1S-IC 2 μV
N18
L4S-L1S 10 ms
1 μV
N8
10 ms
Knee-IC 5 μV
Figure 18–5. Normal 8-channel left tibial SEPs in a 27-year-old woman.

is best displayed with an L1S-to-L4S mon- CUTANEOUS NERVE

tage; however, because its amplitude is small, STIMULATION SEPs
the potential cannot be recorded in many
Stimulation of cutaneous nerves such as the
patients.
sural, superficial peroneal, and lateral femoral
N30. The N30 potential recorded over the cer- cutaneous nerves in the lower extremity and
vical spine (C5S–Fz montage) represents activ- the digital, superficial radial, and other nerves
ity in the fasciculus gracilis, the spinocerebellar in the upper extremity readily elicits a scalp
pathways, and possibly the gracile nucleus.21, 22 SEP in normal subjects. However, the ampli-
It is difficult to record in many subjects tude of the potentials is much smaller than
because of muscle artifact and progressive dis- those obtained with mixed nerve stimulation,
persion of the ascending volley. Stimulation and responses are not obtained over the spine.
of the posterior tibial nerve bilaterally is fre- Cutaneous nerve stimulation is used (1) to
quently helpful in eliciting the tibial lumbar assess the integrity of specific cutaneous nerves
and cervical responses (Fig. 18–5). that are not readily studied with conventional
nerve conduction study techniques, (2) to eval-
N34. The N34 (Fz–C5S montage) is a subcorti- uate isolated root function, and (3) to assess
cally generated FFP analogous to N18 follow- patchy numbness for medical–legal reasons.25
ing stimulation of the median nerve. It may
reflect postsynaptic activity from many gener-
ator sources in the brain stem and, perhaps, DERMATOMAL SEPs
thalamus. Dermatomal stimulation is used occasionally to
P38. The activity of the foot area in pri- assess function of the lumbosacral and cervi-
mary somatosensory cortex is ascribed to P38 cal nerve roots. Stimulation sites are the thumb
(also known as P37).23 This potential is usually (C6), adjacent sides of the index and middle
maximal somewhere between the midline and fingers (C7), little finger (C8), the dorsal sur-
the centroparietal scalp locations, contralat- face of the foot between the first and second
eral to the simulated leg. The primary com- toes (L5), and the lateral side of the foot (S1).
ponent may consist of at least two dipoles.24 Stimulation sites and normal values are avail-
Because the orientation of the dipole inside able for the cervical, thoracic, and lumbosacral
the longitudinal fissure is variable, P38 is some- levels.26
times maximal over the ipsilateral scalp. This
is known as paradoxical localization. To be
certain that P38 is absent, record from the ipsi- SEP Interpretation
lateral scalp as well as from the usual midline
location. The first important step in interpreting SEP
studies is close examination of the recording
Key Points to determine whether all the normally appear-
ing components are present. The absence
• The major clinically important potentials of a main ulnar or median SEP compo-
with tibial nerve stimulation are the N8, nent almost always indicates an abnormal-
N22, N30, and P38 potentials. ity. Similarly, the absence of a waveform that
• The N8 potential represents the periph- is easily recorded on the contralateral side
eral nerve action potential recorded at the also indicates an abnormality. Occasionally, the
popliteal fossa. lumbar and cervical responses following tib-
• The N22 potential represents the post- ial nerve stimulation are absent in normal
synaptic potentials generated in the dor- subjects and frequently absent in older and
sal horn of the spinal cord in the cauda obese subjects, particularly if they have diffi-
equina. culty relaxing. Stimulation of the nerve is per-
• The N30 potential represents activity formed twice to assess for reproducibility of
recorded in the dorsal columns in the the recorded responses, which helps to assess
cervical spinal cord. technical reliability. The lack of superimpos-
• The P38 potential represents activity able tracings at the lumbar and cervical levels
recorded over the primary somatosensory often represents a technical limitation rather
cortex. than an abnormality. Subcortical or peripheral
potentials may be low in amplitude or absent, latency of spinal and cortical evoked potentials.
but because of central amplification and sev- Therefore, it is necessary to monitor limb tem-
eral parallel central pathways, a relatively nor- perature to avoid errors. If the temperature of
mal scalp response may still be obtained. Avoid the arm is less than 32 ◦ C and that of the leg less
making statements about pathologic condi- than 30 ◦ C, the limbs should be warmed. How-
tions, because disease-specific changes are not ever, central conduction velocity is affected
observed with SEP studies. only if hypothermia is profound. To assist with
interpretation, median and tibial nerve con-
duction studies are performed if the SEPs are
Factors That Affect the Amplitude abnormal. For example, a peripheral neuropa-
thy can markedly affect the absolute latencies
and Latencies of the Evoked and morphology of evoked potentials.
Response
A number of physiologic and technical fac- SEDATIVE MEDICATIONS
tors can affect the amplitude and latencies of Sedation given to reduce muscle artifact may
the evoked responses, and are important to allow the patient to sleep during the test, but
consider in the interpretation of the study. it can mildly prolong the scalp latencies. Seda-
tive medications do not have any affect on the
AGE latencies of the potentials recorded from the
peripheral nerve or spine.
SEP latencies and amplitudes are affected by
age. Values in children do not reach those of
adults until the age of 8 years. In older age MUSCLE ARTIFACT
groups, there is a small decrease of periph- Muscle artifact can be controlled by having
eral sensory nerve conduction and amplitude, the patient relax in a reclining chair or bed.
which is most marked distally. According to one Because evoked responses recorded at the
study, median nerve central conduction time elbow, Erb’s point, or the knee have a high
(N13–N20) was constant between the ages of amplitude, muscle activity is not usually a
10 and 49 years, increased by 0.3 ms between problem at these locations. However, record-
the 5th and 6th decades, and then remained ing over the lumbar or cervical spine is dif-
stable in normal subjects up to 79 years old.27 ficult because of the motor unit activity of
Mild prolongation of N9–N13 and N11–N13 the paraspinal muscles and the distance from
transit times has been found in comparing sub- the generators. Audio monitoring of all chan-
jects 15–39 years old with those 40–60 years nels is essential; however, the spine derivations
old.28 are most important because they are usually
the noisiest channels. Muscle artifact in the
BODY HEIGHT AND LIMB LENGTH scalp leads is rarely a problem. Sedation of the
patient is helpful, especially patients who are
The latencies of central SEPs are a function tense or spastic; diazepam is routinely given for
of body height and limb length. Therefore, sedation unless it is contraindicated.
the use of absolute latencies has major limita-
tions. The use of interpeak latencies that are
not related to body size eliminates the effect ELECTRIC ARTIFACT
of height. However, when interpeak latencies The two main sources of electric artifact
cannot be measured because all peripheral or in recordings of SEPs are stimulus artifact
subcortical evoked potentials are absent, abso- and 60-Hz alternating current transmitted
lute latencies must be relied upon for inter- to the amplifier by the machine used to
pretation even though the abnormalities are record the evoked potential or through elec-
nonlocalizing. tromagnetic radiation. Stimulus artifact can
be decreased by using a stimulus-isolation
device and a fast-recovery amplifier, by main-
TEMPERATURE
taining proper orientation and contact of
Low limb temperature decreases peripheral the stimulating electrodes, and by avoid-
nerve conduction velocity and prolongs the ing higher than necessary stimulus intensity.
Maintaining recording electrode impedance because the effects of height, limb length, and
less than 5000 by cleaning the skin and temperature are eliminated.
proper grounding eliminates most 60-Hz noise With the stimulation of the median or ulnar
from the SEP. If different types of electrodes, nerve, the absence or delay of N13, with a nor-
for example, surface and needle electrodes, mal N9, suggests a lesion central to the brachial
are used at recording and reference sites, an plexus and caudal to the foramen magnum.
impedance mismatch is created, thus amplify- The loss of N13 is also consistent with a lesion
ing 60-Hz interference. of the low-to-mid cervical cord. Because collat-
erals from the main pathway generate the N13
FILTER SETTINGS dorsal horn potential, it is not uncommon for
a lesion of the dorsal horn to eliminate N13,
Correct filter settings decrease noise without while dorsal column function and the N14 and
reducing the waveforms of interest. A low- N20 potentials are preserved. If N13 is normal
frequency filter setting of 30 Hz and a high- but N20 is delayed or absent, a lesion rostral to
frequency setting of 3 kHz are usually satisfac- the midcervical cord is indicated and is either
tory. Restricting low frequencies with a filter a cortical lesion or a subcortical lesion of the
setting of 150 Hz reduces 60-cycle artifact and ascending somatosensory pathways.
may be useful in some situations. It also allows For tibial nerve SEPs, the absence of a lum-
better visualization of certain peaks (e.g., N11). bar potential (N22) following tibial nerve stim-
However, the 150 Hz setting has the disadvan- ulation suggests a lesion at or distal to this level.
tage of reducing the amplitude of most peaks The presence of the lumbar N22 potential,
(e.g., N13) and slightly shortening peak laten- with delay of the cervical N30 potential, sug-
cies. Use of a 60-Hz “notch” filter is not rec- gests a lesion within the spinal cord between
ommended because SEPs in this range contain these two areas. In the absence of a cervical
important physiologic information. potential, the presence of a lumbar potential
with a delayed or absent scalp component sug-
Key Points gests a nonlocalized lesion rostral to the lumbar
• Absence of any waveform with median or spinal cord.
ulnar nerve stimulation is abnormal. Side-to-side interpeak latency differences
• Absence of lumbar or cervical potentials are also sensitive indicators of abnormality.
with tibial nerve stimulation is commonly Dispersion of SEPs suggests desynchronization
seen due to muscle artifact, especially in of the nerve action potential analogous to that
older individuals. found in demyelinating disease of peripheral
• Sedation or bilateral tibial nerve stimula- nerves; however, this is difficult to quantify and
tion may help with detection of subcortical should be interpreted cautiously. Morphologi-
peaks when recording tibial SEPs. cal peculiarities of waveforms, unaccompanied
• Body height, limb length, temperature, by latency prolongation, should not be inter-
and other technical factors are important preted as an abnormality but rather as an atyp-
in recording SEPs and may affect inter- ical feature of uncertain clinical significance.
pretation of the SEP.
Amplitude Reduction
LOCALIZATION A decrease in the amplitude of SEP waveforms
is helpful; however, the range of normal values
Latency Prolongation is broad, making this measurement less use-
ful than latency measurements. Also, a general
For purposes of clinical interpretation, SEP attenuation of cortical SEP amplitude may be
waveforms are assumed to represent the encountered with a lesion at any level of the
sequential activation of ascending levels of the somatosensory pathway from the periphery to
somatosensory pathway. Interpeak latency pro- the cerebral cortex. It has been suggested that
longations indicate a defect between the gen- a 50% or greater side-to-side difference indi-
erators of the two peaks involved. Interpeak cates a substantial central conduction block or
latency determinations are most desirable axonal loss or both.
Key Points Disorders of the Peripheral

• Interpeak latency determinations are desir- Nervous System
able because they eliminate the effects of
height, limb length, and temperature. PERIPHERAL NEUROPATHY
• Amplitude of the SEP waveforms is less Peripheral neuropathies generally cause pro-
useful than latency prolongation because longation of the absolute latencies of brachial
of the broad range of normal values for plexus, neck, and scalp evoked potentials, but
amplitude. A 50% or greater asymmetry in central conduction time is normal. When the
amplitude between sides may be clinically neuropathy is marked, only low-amplitude and
important. poorly formed scalp potentials are obtained,
• Interpeak latency prolongations imply a and other components are absent (Fig. 18–6).
lesion between the generators of the two In some cases, no response is obtained at
peaks involved. any level, particularly with stimulation of the
• Absence of subcortical peaks, which is posterior tibial nerve. Routine nerve conduc-
more common with tibial nerve SEPs, tion studies are sufficient for the evaluation
may point to a nonlocalized lesion of cen- of most neuropathies. SEPs are useful for
tral pathways especially if a lumbar N22 evaluating the proximal segments of nerve if
response can be recorded. slowing of conduction through the plexus or
roots is suspected. Occasionally, patients with
chronic inflammatory demyelinating neuropa-
CLINICAL APPLICATIONS thy may have slowing primarily at the root
level, with relatively normal peripheral nerve
The usefulness of SEPs in disorders of the cen- conduction.
tral and peripheral nervous system has been In neuropathy associated with sclerosing
reviewed recently.25 myeloma, prolonged interpeak latencies of
Female, age 31 years Bulimia/Anorexia

SNAP absent Normal scalp SEP
Stimulate: Left median
N5
Elbow
EP1-EP2
N13
C7SP-Fz
N13–N20 = 5.6
N20
C4'-Fz
0 ms 80
Figure 18–6. Median SEPs in a patient with a severe axonal peripheral neuropathy from prolonged ingestion of 2 g of
vitamin B6 daily. N5 has a very low amplitude and N9 is absent. The N13 amplitude is very small, but N20 is normal
because of central amplification of the signal. Central conduction (N13–N20) is normal. The antidromic median sensory
nerve action potential (SNAP) recorded from the index finger was absent.
64.8 ms TRAUMATIC BRACHIAL

Cz-Fz PLEXOPATHY
C5S-Fz Following traumatic plexopathy, recordable
L1S-IC
scalp SEPs and the attenuation or absence of
sensory nerve action potentials indicate con-
Knee-IC tinuity between peripheral and central struc-
tures. Conversely, the presence of a normal
Figure 18–7. Guillain–Barré syndrome in a 69-year-old
woman. Note markedly prolonged scalp response after
Erb’s point potential and the absence of cervi-
sural nerve stimulation. Subcortical evoked potentials are cal and scalp responses suggest avulsion of the
absent. Sural nerve action potential at the ankle was normal roots of the plexus (Fig. 18–9). However, local-
(amplitude 9.8 μV, latency 4.1 ms at 14 cm). ization often is not possible for severe lesions
that involve both preganglionic and postgan-
glionic elements of the plexus. The absence of
responses without paraspinal denervation may
N9–N13 have been found and correlated suggest the need for surgical exploration of the
with demyelination in the dorsal root. Simi- plexus. Complete SEP study of an extensive
lar changes have been reported in Guillain– brachial plexus injury may require stimulation
Barré syndrome, but only occasionally in of several nerves, including the ulnar, median,
the absence of changes in F-wave latencies and radial nerves.
(Fig. 18–7).29 In patients with severe sen- With more restricted injuries, it is possible
sory neuropathies, a relatively normal scalp to be more selective in planning the study. The
SEP can often be obtained because of cen- nerve chosen should have roots near the site
tral amplification of the peripheral affer- of injury, as determined clinically and elec-
ent volley, even when the peripheral sensory tromyographically. If only one or two roots or
nerve action potential is absent. If the scalp trunks are involved, stimulation of the median
latencies are within normal limits, it can be nerve can give normal results, because median
inferred that peripheral sensory nerve con- nerve afferents enter the spinal cord through
duction velocities probably are not markedly several roots (C6, C7, C8, and T1). Stimula-
slowed. Central conduction times in heredi- tion of the musculocutaneous nerve has been
tary motor and sensory neuropathy type I are advocated for the study of upper trunk lesions,
usually normal.30 Slowing of spinal sensory but this nerve is difficult to stimulate selectively
conduction occurs in some patients with dia- without activating the radial nerve. If avulsion
betes mellitus. Dermatomatal SEP can be of C7 is suspected, stimulate the radial nerve,
used to evaluate for some cutaneous neu- and for C8 and T1 lesions, stimulate the ulnar
ropathies, where standard nerve conduction nerve. A more precise but time-consuming
studies are less reliable, such as lateral femoral alternative to stimulation of mixed nerve trunks
cutaneous neuropathy (meralgia paresthetica)
(Fig. 18–8).
N20 (–)
C4'-Fz
N13 (–)
Stimulate right thigh
C5S-Fz
N1 N2
N9
10 ms
Cz-Fz 0.2 μv
P1 10 ms
EPi-EPc
P2 0.5 μV N5
Stimulate left thigh Elbow
10 ms
0.2 μV Figure 18–9. The median SEP in a 34-year-old man
Cz-Fz
injured in a motorcycle accident shows a normal N5,
10 ms
0.2 μV a poorly formed N9, and absence of N13 and N20
components, consistent with root avulsion. The median
Figure 18–8. Left meralgia paresthetica. Stimulation of antidromic sensory amplitude was 15.6 μV. The thenar
the skin of the right and left anterolateral thigh. The scalp compound muscle action potential was absent. The ulnar
response after stimulation on the left is absent. SEP showed the same pattern.
is stimulation of “segmental” cutaneous nerves. • SEP may help in localization of brachial

Stimulation of the lateral cutaneous nerve of plexus lesions, but often must be inter-
the forearm and digital nerve afferents permits preted along with results of conventional
evaluation of nerve fibers confined to nerve needle EMG.
roots C5–C8. • SEP is rarely useful in routine evaluation
of radiculopathies.
THORACIC OUTLET SYNDROME
Ulnar SEPs have limited usefulness in the Disorders of the CNS
diagnosis of neurogenic thoracic outlet syn-
drome.31–33 Several patterns have been des- Abnormalities of SEPs have been described
cribed, including a low-amplitude N9, with a in many diseases of the CNS. However, SEPs
prolonged N9–N13 interpeak latency, and a are most helpful in detecting CNS lesions
low-amplitude N13, with or without attenu- when the results of the clinical examina-
ation of the N9 potential. The ulnar nerve tion are normal or equivocal. Diseases of
should be stimulated because stimulation of myelin tend to produce prominent changes in
the median nerve usually gives normal results. latency, whereas diseases of axons preferen-
SEPs are usually normal in patients with vascu- tially affect the amplitude of central poten-
lar or symptomatic thoracic outlet syndrome. tials. The overlap is so marked that patho-
logic conditions cannot be predicted reliably by
RADICULOPATHY changes in SEPs. The following discussion is
limited to disorders in which SEPs appear to
Radiculopathies are evaluated most easily and aid in diagnosis or management.
reliably by needle electromyography. SEPs
evoked by stimulation of a mixed nerve are usu- CERVICAL SPONDYLITIC
ally normal in the case of a single root lesion, MYELOPATHY
because major nerve trunks are formed from
several roots. Segmental sensory stimulation, Tibial SEPs are abnormal in approximately
stimulation of individual digits, or stimulation 75% of cases of cervical spondylitic myelo-
of peripheral nerve branches innervated by pathies, while ulnar SEPs are abnormal in 60%,
a single nerve root are potentially useful for and median SEPs are abnormal in 25% of
the evaluation of patients with disk disease. cases.37–39 The abnormalities consist of loss of
Some investigators have reported a good corre- amplitude, degradation of waveforms, or slight
lation between lumbosacral dermatomal SEP interpeak delays (Figs. 18–10 and 18–11). The
abnormalities and findings on electromyogra- low-amplitude N9 potentials that are found
phy, myelography, and at surgery; but others occasionally may be caused by disease of the
have found the method less useful.34–36 Der- dorsal root ganglia. The median N13 potential
matomal SEP studies are time-consuming and is more likely to be abnormal when compres-
difficult to interpret because the responses can sion occurs at several spinal cord levels rather
be small and poorly formed. Because of these than at either the C4–5 or the C5–6 level
limitations, SEP studies are not performed alone.40 The absence of the cervical peak (N30)
routinely in the evaluation of radiculopathies. after tibial nerve stimulation is common in cer-
vical myelopathy, but it may also be absent in
Key Points normal older persons, making interpretation
a problem. Therefore, the most reliable mea-
• Peripheral neuropathies and other periph- surement is prolongation of the tibial N22–P38
eral nerve lesions typically produce pro- interpeak latency.
longed absolute latencies, but normal SEP testing repeated a year after success-
interpeak latencies unless there is a sep- ful surgical treatment shows persisting abnor-
arate additional lesion of the CNS. malities, and the results are of little value
• SEPs may be recorded from the scalp in assessing the adequacy of the decompres-
even with severe peripheral neuropathies, sion.41 Somewhat surprisingly, the severity of
a phenomenon known as central amplifi- MRI abnormalities, clinical examination find-
cation of the response. ings, and SEP abnormalities show no clear
N20
C4'-Fz 5 ms
1 μV
C4'-EPi 5 ms
2 μV
5 ms
C3'-EPi 2 μV
N13
C5S-Fz 5 ms
1 μV
C5S-EPc 5 ms
2 μV
N13
C5S-AC 5 ms
N9 1 μV
5 ms
EPi-EPc 2 μV
5 ms
N5 10 μV
Elbow
Figure 18–10. Left median SEPs in a 63-year-old man with a slowly progressive quadriparesis caused by cervical
spondylotic myelopathy. The N13–N20 interpeak latency is mildly prolonged at 7.0 ms (normal, 6.7 ms). See MRI in
Figure 18–11.
correlation. However, severe MRI abnormali-

ties are usually associated with abnormal SEPs.
Conversely, some patients with normal imaging
studies may have abnormal SEPs.42
DEMYELINATING DISEASE
Most frequently SEPs are used to evaluate
suspected multiple sclerosis (Fig. 18–12), espe-
cially in documenting a second clinically silent
lesion, for example, in a patient with optic neu-
ritis. SEP testing is also helpful if symptoms
are suggestive of myelopathy when no definite
abnormalities are found on physical examina-
tion. Median SEPs are abnormal in 56%–59%
and tibial SEPs are abnormal in 77%–82% of
all patients with multiple sclerosis. The rates
of abnormality are somewhat higher for lower
limb SEPs because of the greater length of
white matter traversed. The sensitivity of SEPs
in detecting “clinically unsuspected lesions”
ranges from 37% to 41%. The most frequent
finding is prolongation of SEP interpeak laten-
cies, but low-amplitude and dispersed scalp
responses are also common. It may be difficult
to determine whether the early scalp responses
are markedly prolonged or are absent and only
Figure 18–11. MRI of the cervical spine of the same
patient as in Figure 18–10. Note cervical spondylosis with
late potentials are present. The N13 poten-
myelomalacia and spinal cord atrophy at C3–C4. The tial recorded after median nerve stimulation is
patient had a history of previous C3–C4 anterior cervical often absent or attenuated, whereas the scalp
fusion. response is still present. Lesions of the cervical
N20 remain, including the fact that SEPs are more

C3'-Fz
N11
variable in MS patients than in normal con-
C5S-Fz trols due to instability of conduction in partially
N9 demyelinated axons, or conduction block or
EPi-EPc loss of potentials due to axonal damage. Addi-
N5
tionally, SEPs are more difficult to record in
Elbow MS patients who may have significant spasticity
or other factors that introduce technical diffi-
Figure 18–12. Right median SEPs in a patient with mul-
tiple sclerosis. N13 is absent. The N9–N20 interpeak culties.51 The test–retest variability of evoked
latency is quite prolonged at 14.1 ms (normal, 6.8 ms). The potentials in MS patients over time is not well
scalp response has a low amplitude and is dispersed. These known. All of these factors currently limit the
findings are consistent with demyelination. utility of SEPs as outcome measures for MS
clinical trials.
cord seen on MRI usually result in an abnormal SYRINGOMYELIA

SEP.43 Although the overall sensitivity of SEPs
is lower than that of MRI in multiple sclerosis In syringomyelia, ulnar and median SEPs are
(MS), SEPs may be better for detecting spinal usually normal if there is dissociated sensory
cord lesions that are below the resolution of loss and usually abnormal if all sensory modali-
MRI.44 ties are impaired. The absence of N13 suggests
While SEPs are clearly useful in the diagno- involvement of the central gray matter of the
sis of MS, their use as a prognostic marker in cervical cord.52 Because of conduction directly
MS is more controversial. Recent studies have up the dorsal columns, P14 and N20 may
demonstrated a good correlation (R = 0.60, still be recorded when spinal N13 is absent.
Spearman correlation coefficient) between Tibial SEPs most frequently are abnormal,
SEP abnormality “score” and increased disabil- showing an absence of the neck response
ity on the Kurtzke EDSS (Expanded Disability (N30), delayed low-amplitude or absent scalp
Status Scale) as well as the relevant functional potentials, or prolongation of the N22–P38
system score in patients at baseline evalua- latency.32
tion.45 In a cohort of patients followed lon-
gitudinally for a mean of 30.5 months, there
SPINAL CORD TUMORS
was a significant worsening of SEP scores
overall. However, the correlation of worsen- Various types of spinal cord tumor com-
ing SEP scores with worsening EDSS scores monly have associated SEP abnormalities. An
was not statistically significant. Patients with absent or reduced N13 response indicates
more severe global evoked potential scores involvement of dorsal horn gray matter and
at baseline had a significantly greater risk of is correlated with disturbed pain and tem-
progressing over the follow-up period, with perature sensation and reduced reflexes in
a positive predictive value of 72.5% and a the upper limbs. P14 abnormalities are corre-
negative predictive value of 63.6%. Hence, lated with impaired joint and touch sensation.
evoked potential “scores” may be a useful Involvement of the lumbar cord is associated
tool in prognostication especially since they with an abnormal N22 potential. Prolonga-
represent physiological measures of impair- tion of interpeak latencies is also common.53
ment that may be more clinically meaningful SEPs may be normal in slow-growing astrocy-
than many MRI parameters that are currently tomas that infiltrate but do not destroy sensory
being used as outcome measures in MS clinical pathways.
trials.45, 46
Evoked potentials have been used as an MISCELLANEOUS SPINAL CORD
outcome measure in a limited number of MS LESIONS
clinical trials, with some reports of a posi-
tive correlation evoked potential parameters Markedly prolonged conduction times are
and a therapeutic response to the drug being seen in patients with adrenomyeloneuropa-
studied.47, 48 Other studies have failed to find thy and are caused by involvement of dis-
such associations.49, 50 Significant challenges tal and proximal peripheral nerves and the
CNS.54 Slowing is also found in the majority N20

of adrenoleukodystrophy carriers.55 In vitamin C4'-Fz 5 ms
0.5 μV
B12 deficiency, changes in central conduction N13
C5Sp-Fz 5 ms
usually appear before peripheral slowing, sug- 1 μV
N9
gesting that the central process of a dorsal root 5 ms
ganglion cell is affected earlier than the distal EP1-EP2 1 μV
N5
process.56 Early treatment lessens the abnor- Elbow
5 ms
5 μV
malities of SEPs. SEPs may also be abnormal in
the myeloneuropathy due to copper deficiency,
Figure 18–13. Left median SEPs in a 17-year-old man
where they help document central slowing of with Friedreich’s ataxia. The N13–N20 interpeak latency
conduction in addition to the severe axonal is prolonged at 11.3 ms (normal, 6.7 ms). The N9–N13
sensorimotor peripheral neuropathy that usu- interpeak latency is normal.
ally accompanies this condition. This slowing of
central conduction has been reported to corre- P38 = 52.2
10 ms
late with MRI T2 signal changes in the dorsal 0.5 μV
columns of the spinal cord.57–59 Cz-Fz
10 ms
1 μV
C5S-Fz
Key Points N22 10 ms
0.5 μV
• SEPs are most useful clinically in eval- L1-IC
N8
uation of suspected MS, where they are
10 ms
highly sensitive for clinically undetected Knee-IC 5 μV
spinal cord lesions.
• Tibial SEPs are more likely to be abnor-
mal in MS than median SEPs, due to the Figure 18–14. Tibial scalp potentials, in the same patient
greater length of the pathways involved as in Figure 18–13, are low amplitude, and the P38 latency
from the lower extremities. is approximately 52.2 ms. The N22–P38 interpeak latency
• The usual SEP abnormality in MS is quite prolonged at 28 ms (normal, 20 ms).
is prolongation of interpeak latencies
but preserved waveforms with normal
amplitudes. with other autosomal dominant cerebellar atax-
• Other spinal cord lesions may pro- ias.61, 62 In thalamocortical lesions with no sen-
duce absent or reduced amplitude SEP sory loss, SEPs are normal; but if the sensory
responses, with variable effects on inter- loss is complete, the patient will have no N20
peak latencies. or subsequent potentials. Patients with mini-
mal cortical sensory loss may have asymmet-
rical cortical potentials. Evoked potentials in
patients with nonorganic sensory loss should be
BRAIN STEM AND normal.
SUPRATENTORIAL LESIONS
Central conduction times may be prolonged
MYOCLONUS
with lesions of the lemniscal pathway in the
brain stem or the thalamocortical radiation. Myoclonus occasionally is associated with
SEPs are normal in the Wallenberg and Weber very high-amplitude cortical evoked responses
syndromes, because the medial lemniscus is (Fig. 18–15). The giant SEPs are separable
unaffected. Central conduction times in hered- into three types that reflect hyperexcitability
itary motor and sensory neuropathy type I in the afferent or efferent system (or both)
are usually normal.37 In Friedreich’s ataxia, of the somatosensory cortex.63 The amplitude
SEPs typically show absent or low-amplitude may be increased up to 10 times normal in
N9–N13 potentials of normal or near nor- patients with cortical reflex myoclonus, includ-
mal latency, whereas the latency of cortical ing progressive myoclonus epilepsy and epilep-
responses is increased, often with broadening sia partialis continua.64 The cortical SEPs are
of N2060 (Figs. 18–13 and 18–14). Abnormal not exaggerated in patients with myoclonus of
median and tibial SEPs are also seen in patients brain stem or spinal cord origin.
Onset pregnancy cause unknown

F, age 21
Stimulate: Right median Stimulate: Right tibial
– –
Elbow 5
Knee-IC 5
–
EP1-EP2 5 L1SP-IC –
5
–
C5SP-Fz – C5SP-Fz
1 N35 5
N20
–
P38 = 42.3(0.6 – 6.5)
C5'-Fz 5 Cz-Fz
–
P25 = 35.1(0.7– 5.9)
5
P25 P38
0 ms 60 0 ms 100
Figure 18–15. Upper and lower extremity SEPs in a patient with stimulus-sensitive myoclonus, illustrating the charac-
teristic pattern of very high-amplitude cortical potentials recorded from the scalp electrodes.
MOTOR NEURON DISEASE PROGNOSIS IN COMA

Patients with amyotrophic lateral sclerosis may Head Injury
have minor abnormalities of central conduc-
tion, delay, or absence of cortical responses and SEPs are helpful in defining the functional
increased mean values of all potential laten- reversibility of cerebral injuries of any cause.68
cies with median nerve stimulation and stim- In 78% of patients, SEPs can predict out-
ulation of lower extremity motor and sensory come within 3.5 days.69 Surviving patients
nerves65, 66 (Fig. 18–16). Severe abnormalities with persistent asymmetries of cerebral evoked
in patients with suspected motor neuron dis- potentials remain hemiplegic, whereas those
ease should raise suspicion of other conditions with absence of evoked potentials over both
cerebral hemispheres die. In most patients
that may mimic amyotrophic lateral sclerosis,
for example, cervical spondylosis. An exception who have a good recovery, conduction times
to this generalization is X-linked spinobulbar return to normal before day 10, but in those
who remain disabled or who die, conduction
muscular atrophy (Kennedy’s syndrome), in
which the scalp evoked potentials after median times are persistently abnormal or scalp poten-
and tibial nerve stimulation are poorly formed tials are absent. When cortical evoked poten-
tials are absent, it generally is futile to use
or absent, whereas the peripheral and spinal
potentials are relatively preserved.67 heroic measures such as barbiturate coma or
decompressive craniectomy to control refrac-
tory increased intracranial pressure.
However, SEPs are not perfect indicators
N20 = 22.3 (< 22.0)
of recovery. A few patients with poor recov-
C4'-Fz
ery can have normal SEPs. Serial studies have
shown that both conduction time and ampli-
N13
N13–N20 = 7.3 (< 6.7)
tude recover gradually, with differences per-
C5S-Fz sisting between patients who make a good
N9 recovery and those who remain disabled. SEPs
are superior to EEG in determining prognosis
EPi-EPc
of severe head injury, by allowing prediction of
N5 favorable and unfavorable outcomes in a much
Elbow larger number of patients.70, 71
Figure 18–16. Left median SEPs in a 57-year-old man
A battery of auditory evoked potentials,
with amyotrophic lateral sclerosis. There is mild slowing of visual evoked potentials, and SEPs is also an
central conduction (N13–N20 interpeak latency). accurate and reliable prognostic indicator after
severe head injury, with an overall accuracy of associated with deficits in most patients. SEPs
approximately 91%.72–74 Patients are assigned also have long-term predictive value for deficits
to multimodality evoked potential (MMEP) that become apparent at school age.81
groups according to the most abnormal study
obtained in any modality. The scale ranges
from grade 1 (normal) to grade 4 (absence of SEP Findings in Brain Death
activity). As the severity of MMEP abnormality
increases, so does mortality. Of patients with All patients with brain death have bilateral loss
mildly abnormal MMEP scores, 81% have a of median SEP N20 components, but cervi-
return to normal life or have only moderate dis- cal N13 potentials can still be elicited. The
ability, and 76% of those with severely abnor- presence of SEP N13 is helpful because it
mal MMEP scores have a poor outcome. A establishes that the input signal has reached
good outcome is realized by 76% of patients the CNS. However, this finding does not prove
with a grade 1 MMEP, 61% with grade 2, 35% brain death, because rare patients who have
with grade 3, and 0% with grade 4. Overall, severe bilateral supratentorial lesions, drug
87% of patients with a grade 1 MMEP have a intoxication, or severe cerebral edema but no
good-to-moderate outcome at 1 year. clear clinical signs of brain death may also have
a loss of the N20 potential.82
Recent research has suggested that N18 is
Anoxia, Cerebral Hemorrhage, and
a useful indicator of brain stem function. This
Stroke
potential has the advantage of being generated
Outcome has been examined in a series of by the cuneate nucleus in caudal medulla, close
patients with anoxic coma caused by cardiopul- to the respiratory center. N18 is almost always
monary arrest or severe hypotension whose lost in brain death and preserved in recordings
prognoses were uncertain on the basis of clin- from patients who are comatose but not brain
ical findings on day 1.75 Patients with an dead. In contrast, auditory brain stem evoked
obviously good or bad prognosis clinically were responses reflect pontine and midbrain func-
excluded. All 18 patients with absent or low- tion rather than medullary function and can fail
amplitude responses had no recovery. It was to detect remaining brain stem function.83
found that some patients with initially unfa-
vorable appearing EEGs and normal SEPs
may recover and should be supported until SEPs Recorded in the Intensive
the prognosis is more definitive. The combina- Care Unit
tion of SEPs and brain stem auditory evoked
response is useful in assessing the prognosis Recording SEPs in the intensive care unit
of patients with subarachnoid or hypertensive presents technical challenges, usually in the
hemorrhage and cerebral infarction.76, 77 form of high-amplitude 60-Hz artifact, not
encountered in the outpatient laboratory. Sug-
gestions for reducing the artifact include shut-
Infants and Children
ting off all nonessential electric equipment
SEP results similar to those in adults have such as lights, cardiac monitor, cooling blanket,
been found in children comatose from hypoxic- feeding pumps, and blood warmers to ascer-
ischemic encephalopathy, head injury, or other tain if one of these is causing the artifact.
conditions.78 In a study of 127 children who The impedance of the electrodes should be
were comatose because of severe head injury, checked and, if necessary, the electrodes reap-
all 32 who had an absence of brain stem plied or new electrodes substituted. If artifact
auditory evoked responses and SEPs died. Of is present in all channels, replace the ground.
children with normal evoked potential studies, In an unconscious patient or in one under
78% had a good prognosis.79 SEPs recorded in anesthesia, subcutaneous needle electrodes
the first week after admission correlate highly may be applied; this not only saves time but
with outcome assessed 1 and 5 years after also reduces artifact. If muscle artifact is a
severe brain injury.80 SEPs are also accurate problem in a patient who had a head injury and
prognostic tools for newborns with asphyxia. is decerebrate or decorticate and on a respira-
The absence of scalp potentials is a very tor, a single dose of a neuromuscular blocking
poor prognostic sign, and delayed latencies are agent can be given safely. If a paralyzing agent
cannot be given, increasing the gain to 20 μV or nervous system, to identify objectively abnor-
50 μV per division on channels recording from malities in patients with few sensory manifes-
the elbow, Erb’s point, or neck may eliminate tations or none at all, and to monitor function
blocking of the amplifier and still result in recover time.
ognizable peaks, which can be amplified after
the recording has been completed.
Patients with severe head injury often req- REFERENCES
uire neurosurgical procedures for evacuation
of intracerebral hematomas and frequently 1. Lee, E. K., and M. Seyal. 1998. Generators of
short latency human somatosensory-evoked potentials
have intracranial pressure monitors placed. recorded over the spine and scalp. Journal of Clinical
The location of the incisions and drains may Neurophysiology 15:227–34.
require placement of recording electrodes 2. Kakigi, R., H. Shibasaki, K. Tanaka, et al. 1991.
slightly anterior or posterior to the usual loca- CO2 laser-induced pain-related somatosensory evoked
tions. Generally, this does not result in a potentials in peripheral neuropathies: Correlation
between electrophysiological and histopathological
marked change in the morphology of the cor- findings. Muscle & Nerve 14:441–50.
tical evoked potentials, and they are still diag- 3. Kakigi, R., S. Watanabe, and H. Yamasaki. 2000. Pain-
nostically useful. The rate of stimulation may related somatosensory evoked potentials. Journal of
also have an effect in patients with head injury Clinical Neurophysiology 17:295–308.
4. Guidelines for somatosensory evoked potentials. 1999.
or significant spasticity. Recording at 5 per Muscle & Nerve 22(Suppl 8):S123–38.
second occasionally causes a reflexive increase 5. Dumitru, D., and D. L. Jewett. 1993. Far-field poten-
in decerebrate posturing or extensor spasms, tials. Muscle & Nerve 16:237–54.
which can be eliminated by reducing the rate 6. Desmedt, J. E., and G. Cheron. 1981. Prevertebral
of stimulation to 1 or 2 per second and by (oesophageal) recording of subcortical somatosensory
evoked potentials in man: The spinal P13 compo-
using the lowest voltage possible to elicit a nent and the dual nature of the spinal generators.
muscle twitch. Electroencephalography and Clinical Neurophysiol-
ogy 52:257–75.
Key Points 7. Emerson, R. G., M. Seyal, and T. A. Pedley. 1984.
Somatosensory evoked potentials following median
• SEP latencies may be prolonged in a nerve stimulation. I. The cervical components. Brain
variety of brain stem or supratentorial 107:169–82.
8. Jeanmonod, D., M. Sindou, and F. Mauguiere. 1989.
processes, including ALS, spinocerebel- Three transverse dipolar generators in the human cer-
lar ataxias, Freidrich’s ataxia, and other vical and lumbo-sacral dorsal horn: Evidence from
disorders. direct intraoperative recordings on the spinal cord
• Giant cortical SEPs may be seen with surface. Electroencephalography and Clinical Neuro-
physiology 74:236–40.
various forms of cortical myoclonus. 9. Mauguiere, F. 2000. Anatomic origin of the cervical
• Absence of cortical SEP responses or sig- N13 potential evoked by upper extremity stimulation.
nificant asymmetries in coma may predict Journal of Clinical Neurophysiology 17:236–45.
a poor prognosis for recovery. 10. Lesser, R. P., H. Lueders, J. Hahn, and G. Klem. 1981.
• Absence of cortical SEP responses may Early somatosensory potentials evoked by median
nerve stimulation: Intraoperative monitoring. Neurol-
provide confirmatory evidence for brain ogy 31:1519–23.
death. 11. Sonoo, M., T. Shimpo, K. Genba, M. Kunimoto, and T.
Mannen. 1990. Posterior cervical N13 in median nerve
SEP has two components. Electroencephalography
and Clinical Neurophysiology 77:28–38.
SUMMARY 12. Zanette, G., M. Tinazzi, P. Manganotti, C. Bonato, and
A. Polo. 1995. Two distinct cervical N13 potentials are
SEPs recorded with surface electrodes rep- evoked by ulnar nerve stimulation. Electroencephalog-
resent volume-conducted activity arising from raphy and Clinical Neurophysiology 96:114–20.
myelinated peripheral and central axons, 13. Tomberg, C., J. E. Desmedt, I. Ozaki, and P. Noël.
1991. Nasopharyngeal recordings of somatosensory
synapses in central gray matter, and changes evoked potentials document the medullary origin of
in the size and shape of the volume con- the N18 far-field. Electroencephalography and Clini-
ductor. They provide an objective measure of cal Neurophysiology 80:496–503.
function in large-diameter myelinated sensory 14. Restuccia, D. 2000. Anatomic origin of P13 and P14
scalp far-field potentials. Journal of Clinical Neuro-
afferents peripherally and in proprioceptive physiology 17:246–57.
pathways centrally. Changes in amplitude and 15. Urasaki, E., S. Uematsu, and R. P. Lesser. 1993. Short
latency can be used to localize lesions in the latency somatosensory evoked potentials recorded
around the human upper brain-stem. Electroen- 30. Aramideh, M., J. E. Hoogendijk, C. M. Aalfs, et al.
cephalography and Clinical Neurophysiology 88:92– 1992. Somatosensory evoked potentials, sensory nerve
104. potentials and sensory nerve conduction in heredi-
16. Noël, P., I. Ozaki, and J. E. Desmedt. 1996. Origin of tary motor and sensory neuropathy type I. Journal of
N18 and P14 far-fields of median nerve somatosensory Neurology 239:277–83.
evoked potentials studied in patients with a brain-stem 31. Yiannikas, C., and J. C. Walsh. 1983. Somatosensory
lesion. Electroencephalography and Clinical Neuro- evoked responses in the diagnosis of thoracic outlet
physiology 98:167–70. syndrome. Journal of Neurology, Neurosurgery, and
17. Sonoo, M., M. Kobayashi, K. Genba-Shimizu, T. Man- Psychiatry 46:234–40.
nen, and T. Shimizu. 1996. Detailed analysis of 32. Veilleux, M., and J. C. Stevens. 1987. Syringomyelia:
the latencies of median nerve somatosensory evoked Electrophysiologic aspects. Muscle & Nerve 10:
potential components, 1: Selection of the best standard 449–58.
parameters and the establishment of normal values. 33. Komanetsky, R. M., C. B. Novak, S. E. Mack-
Electroencephalography and Clinical Neurophysiol- innon, M. H. Russo, A. M. Padberg, and S.
ogy 100:319–31. Louis. 1996. Somatosensory evoked potentials
18. Sonoo, M. 2000. Anatomic origin and clinical applica- fail to diagnose thoracic outlet syndrome. The
tion of the widespread N18 potential in median nerve Journal of Hand Surgery [American Volume] 21:
somatosensory evoked potentials. Journal of Clinical 662–6.
Neurophysiology 17:258–68. 34. Aminoff, M. J., D. S. Goodin, G. J. Parry, N. M.
19. Yamada, T., T. Muroga, and J. Kimura. 1981. Barbaro, P. R. Weinstein, and M. L. Rosenblum.
Tourniquet-induced ischemia and somatosensory 1985. Electrophysiologic evaluation of lumbosacral
evoked potentials. Neurology 31:1524–9. radiculopathies: Electromyography, late responses,
20. Seyal, M., and A. J. Gabor. 1985. The human pos- and somatosensory evoked potentials. Neurology 35:
terior tibial somatosensory evoked potential: Synapse 1514–18.
dependent and synapse independent spinal compo- 35. Rodriquez, A. A., L. Kanis, A. A. Rodriquez, and D.
nents. Electroencephalography and Clinical Neuro- Lane. 1987. Somatosensory evoked potentials from
physiology 62:323–31. dermatomal stimulation as an indicator of L5 and
21. Halonen, J. P., S. J. Jones, M. A. Edgar, and A. O. S1 radiculopathy. Archives of Physical Medicine and
Ransford. 1989. Conduction properties of epidurally Rehabilitation 68:366–8.
recorded spinal cord potentials following lower limb 36. Walk, D., M. A. Fisher, S. H. Doundoulakis, and
stimulation in man. Electroencephalography and Clin- M. Hemmati. 1992. Somatosensory evoked poten-
ical Neurophysiology 74:161–74. tials in the evaluation of lumbosacral radiculopathy.
22. Seyal, M., L. W. Kraft, and A. J. Gabor. 1987. Cervical Neurology 42:1197–202.
synapse-dependent somatosensory evoked potential 37. Veilleux, M., and J. R. Daube. 1987. The value of
following posterior tibial nerve stimulation. Neurology ulnar somatosensory evoked potentials (SEPs) in cervi-
37:1417–21. cal myelopathy. Electroencephalography and Clinical
23. Yamada, T. 2000. Neuroanatomic substrates of lower Neurophysiology 68:415–23.
extremity somatosensory evoked potentials. Journal of 38. Yiannikas, C., B. T. Shahani, and R. R. Young. 1986.
Clinical Neurophysiology 17:269–79. Short-latency somatosensory-evoked potentials from
24. Baumgärtner, U., H. Vogel, J. Ellrich, J. Gawehn, radial, median, ulnar, and peroneal nerve stimulation
P. Stoeter, and R. D. Treede. 1998. Brain electri- in the assessment of cervical spondylosis. Compari-
cal source analysis of primary cortical components of son with conventional electromyography. Archives of
the tibial nerve somatosensory evoked potential using Neurology 43:1264–71.
regional sources. Electroencephalography and Clinical 39. Yu, Y. L., and S. J. Jones. 1985. Somatosensory
Neurophysiology 108:588–99. evoked potentials in cervical spondylosis. Correlation
25. Aminoff, M. J., and A. A. Eisen. 1998. AAEM min- of median, ulnar and posterior tibial nerve responses
imonograph #19: Somatosensory evoked potentials. with clinical and radiological findings. Brain 108:
Muscle & Nerve 21:277–90. 273–300.
26. Slimp, J. C., D. E. Rubner, M. L. Snowden, and W. 40. Kaneko, K., S. Kawai, T. Taguchi, Y. Fuchigami, T. Ito,
C. Stolov. 1992. Dermatomal somatosensory evoked and H. Morita. 1998. Correlation between spinal cord
potentials: Cervical, thoracic, and lumbosacral levels. compression and abnormal patterns of median nerve
Electroencephalography and Clinical Neurophysiol- somatosensory evoked potentials in compressive cer-
ogy 84:55–70. vical myelopathy: Comparison of surface and epidu-
27. Hume, A. L., B. R. Cant, N. A. Shaw, and J. C. Cowan. rally recorded responses. Journal of the Neurological
1982. Central somatosensory conduction time from Sciences 158:193–202.
10 to 79 years. Electroencephalography and Clinical 41. de Noordhout, A. M., S. Myressiotis, V. Delvaux, J. D.
Neurophysiology 54:49–54. Born, and P. J. Delwaide. 1998. Motor and somatosen-
28. Strenge, H., and J. Hedderich. 1982. Age-dependent sory evoked potentials in cervical spondylotic myelopa-
changes in central somatosensory conduction time. thy. Electroencephalography and Clinical Neurophys-
European Neurology 21:270–6. iology 108:24–31.
29. Yiannikas, C. 1990. Short-latency somatosensory 42. Berthier, E., F. Turjman, and F. Mauguiere. 1996.
evoked potentials in peripheral nerve lesions, plex- Diagnostic utility of somatosensory evoked potentials
opathies, radiculopathies, and spinal cord trauma. In (SEPs) in presurgical assessment of cervical spondy-
Evoked potentials in clinical medicine, ed. K. H. Chi- lotic myelopathy. Neurophysiologie Clinique 26:
appa, 2nd ed., 439–69. New York: Raven Press. 300–10.
43. Turano, G., S. J. Jones, D. H. Miller, G. H. Du Boulay, 59. Goodman, B. P., B. W. Chong, A. C. Patel, G. P.
R. Kakigi, and W. I. McDonald. 1991. Correlation of Fletcher, and B. E. Smith. 2006. Copper deficiency
SEP abnormalities with brain and cervical cord MRI myelopathy resembling B12 deficiency: Partial resolu-
in multiple sclerosis. Brain 114:663–81. tion of MR imaging findings with copper substitution.
44. Guérit, J. M., and A. Monje Argiles. 1988. The sensi- American Journal of Neuroradiology 27:2112–14.
tivity of multimodal evoked potentials in multiple scle- 60. Jones, S. J., M. Baraitser, and A. M. Halliday. 1980.
rosis. A comparison with magnetic resonance imaging Peripheral and central somatosensory nerve conduc-
and cerebrospinal fluid analysis. Electroencephalogra- tion defects in Friedreich’s ataxia. Journal of Neurol-
phy and Clinical Neurophysiology 70:230–8. ogy, Neurosurgery, and Psychiatry 43:495–503.
45. Leocani, L., M. Rovaris, F. M. Boneschi, et al. 2006. 61. Abele, M., K. Burk, F. Andres, et al. 1997. Autosomal
Multimodal evoked potentials to assess the evolution dominant cerebellar ataxia type I. Nerve conduction
of multiple sclerosis: A longitudinal study. Journal of and evoked potential studies in families with SCA1,
Neurology, Neurosurgery, and Psychiatry 77:1030–5. SCA2 and SCA3. Brain 120:2141–8.
46. O’Connor, P., P. Marchetti, L. Liesly, and M. Perera. 62. Perretti, A., L. Santoro, B. Lanzillo, et al. 1996. Auto-
1998. Evoked potential abnormality scores are a use- somal dominant cerebellar ataxia type I: Multimodal
ful measure of disease burden in relapsing–remitting electrophysiological study and comparison between
multiple sclerosis. Annals of Neurology 44:404–7. SCA1 and SCA2 patients. Journal of the Neurological
47. Nuwer, M. R., J. W. Packwood, L. W. Myers, and G. Sciences 142:45–53.
W. Ellison. 1987. Evoked potentials predict the clinical 63. Shibasaki, H., Y. Yamashita, R. Neshige, S. Tobimatsu,
changes in a multiple sclerosis drug study. Neurology and R. Fukui. 1985. Pathogenesis of giant somatosen-
37:1754–61. sory evoked potentials in progressive myoclonic
48. Khatri, B. O., M. P. McQuillen, G. J. Harrington, epilepsy. Brain 108:225–40.
et al. 1984. Chronic progressive multiple sclerosis: 64. Kakigi, R., and H. Shibasaki. 1987. Generator mech-
Double blind controlled study of plasmapheresis in anisms of giant somatosensory evoked potentials in
patients taking immunosuppressant drugs. Neurology cortical reflex myoclonus. Brain 110:1359–73.
34:1273–9. 65. Georgesco, M., A. Salerno, and W. Camu. 1997.
49. Anderson D. C., G. E. Slater, R. Shearman, et al. 1987. Somatosensory evoked potentials elicited by
Evoked potentials to test the treatment of chronic stimulation of lower-limb nerves in amyotrophic
multiple sclerosis. Archives of Neurology 44:1232–6. lateral sclerosis. Electroencephalography and Clinical
50. Smith, P., I. Zeeberg, and O. Sjo. 1986. Evoked Neurophysiology 104:333–42.
potentials in multiple sclerosis before and after high 66. Zakrzewska-Pniewska, B., R. Gasik, A. Kostera-
dose methylprednisolone infusion. European Neurol- Pruszczyk, and B. Emeryk-Szajewska. 1999. Recon-
ogy 23:67–73. siderations about the abnormalities of somatosensory
51. Leocani, L., S. Medaglini, and G. Comi. 2000. Evoked evoked potentials in motor neuron disease. Elec-
potentials in monitoring multiple sclerosis. Neurologi- troencephalography and Clinical Neurophysiology 39:
cal Sciences 21:S889–91. 107–12.
52. Urasaki, E., S. Wada, C. Kadoya, H. Matsuzaki, A. 67. Polo, A., F. Teatini, S. D’Anna, et al. 1996. Sensory
Yokata, and S. Matsuoka. 1988. Absence of spinal N13- involvement in X-linked spino-bulbar muscular atro-
P13 and normal scalp far-field P14 in a patient with phy (Kennedy’s syndrome): An electrophysiological
syringomyelia. Electroencephalography and Clinical study. Journal of Neurology 243:388–92.
Neurophysiology 71:400–4. 68. Carter, B. G., and W. Butt. 2001. Review of the use
53. Restuccia, D., V. Di Lazzaro, M. Valeriani, C. of somatosensory evoked potentials in the prediction
Colosimo, and P. Tonali. 1996. Spinal responses to of outcome after severe brain injury. Critical Care
median and tibial nerve stimulation and magnetic res- Medicine 29:178–86.
onance imaging in intramedullary cord lesions. Neu- 69. Hume, A. L., and B. R. Cant. 1981. Central
rology 46:1706–14. somatosensory conduction after head injury. Annals of
54. Kaplan, P. W., R. J. Tusa, J. Rignani, and H. W. Moser. Neurology 10:411–19.
1997. Somatosensory evoked potentials in adreno- 70. Hutchinson, D. O., R. W. Frith, N. A. Shaw, J. A. Jud-
myeloneuropathy. Neurology 48:1662–7. son, and B. R. Cant. 1991. A comparison between elec-
55. Restuccia, D., V. Di Lazzaro, M. Valeriani, et al. 1997. troencephalography and somatosensory evoked poten-
Neurophysiological abnormalities in adrenoleukodys- tials for outcome prediction following severe head
trophy carriers. Evidence of different degrees of injury. Electroencephalography and Clinical Neuro-
central nervous system involvement. Brain 120: physiology 78:228–33.
1139–48. 71. Moulton, R. J., J. I. Brown, and S. J. Konasiewicz.
56. Jones, S. J., Y. L. Yu, P. Rudge, et al. 1987. Central and 1998. Monitoring severe head injury: A comparison
peripheral SEP defects in neurologically symptomatic of EEG and somatosensory evoked potentials. The
and asymptomatic subjects with low vitamin B12 lev- Canadian Journal of Neurological Sciences 25:S7–11.
els. Journal of the Neurological Sciences 82:55–65. 72. Greenberg, R. P., P. G. Newlon, M. S. Hyatt, R. K.
57. Kumar, N. 2006. Copper deficiency myelopa- Narayan, and D. P. Becker. 1981. Prognostic impli-
thy (human swayback). Mayo Clinic Proceedings cations of early multimodality evoked potentials in
81:1371–84. severely head-injured patients. A prospective study.
58. Kumar, N., J. B. Gross, and J. E. Ahlskog. 2004. Journal of Neurosurgery 55:227–36.
Copper deficiency myelopathy produces a clinical pic- 73. Narayan, R. K., R. P. Greenberg, J. D. Miller, et al.
ture like subacute combined degeneration. Neurology 1981. Improved confidence of outcome prediction
63:33–9. in severe head injury. A comparative analysis of the
clinical examination, multimodality evoked potentials, potentials in pediatric patients. Critical Care Medicine
CT scanning, and intracranial pressure. Journal of 27:412–16.
Neurosurgery 54:751–62. 79. Butinar, D., and A. Gostisa. 1996. Brain stem auditory
74. Newlon, P. G., R. P. Greenberg, M. S. Hyatt, G. evoked potentials and somatosensory evoked poten-
G. Enas, and D. P. Becker. 1982. The dynamics of tials in prediction of posttraumatic coma in children.
neuronal dysfunction and recovery following severe Pflugers Archive 431:R289–90.
head injury assessed with serial multimodality evoked 80. Carter, B. G., A. Taylor, and W. Butt. 1999. Severe
potentials. Journal of Neurosurgery 57:168–77. brain injury in children: Long-term outcome and
75. Chen, R., C. F. Bolton, and B. Young. 1996. Predic- its prediction using somatosensory evoked potentials
tion of outcome in patients with anoxic coma: A clinical (SEPs). Intensive Care Medicine 25:722–8.
and electrophysiologic study. Critical Care Medicine 81. Schwarz, S., S. Schwab, A. Aschoff, and W. Hacke.
24:672–8. 1999. Favorable recovery from bilateral loss of
76. Facco, E., A. U. Behr, M. Munari, et al. 1998. Auditory somatosensory evoked potentials. Critical Care
and somatosensory evoked potentials in coma follow- Medicine 27:182–7.
ing spontaneous cerebral hemorrhage: Early prognosis 82. Sonoo, M., Y. Tsai-Shozawa, M. Aoki, et al. 1999. N18
and outcome. Electroencephalography and Clinical in median somatosensory evoked potentials: A new
Neurophysiology 107:332–8. indicator of medullary function useful for the diagno-
77. Haupt, W. F., and G. Pawllik. 1998. Contribution of sis of brain death. Journal of Neurology, Neurosurgery,
initial median-nerve somatosensory evoked potentials and Psychiatry 67:374–8.
and brain stem auditory evoked potentials to predic- 83. Sonoo, M., H. Hagiwara, Y. Motoyoshi, and T.
tion of clinical outcome in cerebrovascular critical care Shimizu. 1996. Preserved widespread N18 and pro-
patients: A statistical evaluation. Journal of Clinical gressive loss of P13/14 of median nerve SEPs in a
Neurophysiology 15:154–8. patient with unilateral medial medullary syndrome.
78. Ruiz-López, M. J., A. Martínez de Azagra, A. Serrano, Electroencephalography and Clinical Neurophysiol-
and J. Casado-Flores. 1999. Brain death and evoked ogy 100:488–92.
Chapter 19
Brain Stem Auditory Evoked

Potentials in Central Disorders
Jonathan L. Carter
INTRODUCTION Auditory Acuity

AUDITORY ANATOMY AND Stimulus Rate
PHYSIOLOGY Stimulus Intensity
GENERATORS OF THE BRAIN Stimulus Polarity
STEM AUDITORY EVOKED
INTERPRETATION OF BAEPs
POTENTIALS
CLINICAL APPLICATIONS
BRAIN STEM AUDITORY EVOKED
Acoustic Neuroma
POTENTIALS: METHODS
Demyelinating Disease
Stimulation
Intrinsic Brain Stem Lesions
Recording
Coma and Brain Death
FACTORS AFFECTING THE BAEP Intraoperative Monitoring
RESPONSE
SUMMARY
Age and Gender
INTRODUCTION for performing these studies in patients with

neurologic disease is the close correlation
Brain stem auditory evoked potentials (BAEPs) between specific auditory waveforms and
are electrophysiologic studies that assess the structures in the brain stem.
peripheral and central auditory conduction BAEPs are usually performed with a click
pathways. BAEPs may be used to assess stimulus that is delivered monaurally and that
peripheral auditory nerve function, such as activates the peripheral and central auditory
during the assessment for hearing loss in pathways. Auditory stimuli cause sequential
infants (see Chapter 21). BAEPs are also used activity in CN VIII, the cochlear nucleus, the
to assess for lesions or disorders in the cen- superior olivary nucleus, the lateral lemniscus,
tral nervous system (CNS), and they are usu- and the inferior colliculus. Five prominent ver-
ally abnormal in patients with lesions involv- tex positive waveforms, numbered I through
ing the auditory portion of cranial nerve VIII V, are invariably present in normal subjects
(CN VIII), the auditory pathways in the brain after peripheral auditory stimulation. The com-
stem, or both. BAEPs may be performed mon BAEP alterations in patients with brain
in awake and cooperative patients or those stem lesions include increased I–V interpeak
with an altered mental status (e.g., sedation, latency, a low-amplitude or absent wave V, and
general anesthesia, or coma). The rationale an absence of all waveforms.
281
Results of BAEP studies are useful not only provide prognostically important information
in evaluating hearing and specific neurologic about comatose patients. For BAEPs to assess
disorders but also in evaluating the significance peripheral auditory function, the method must
of certain symptoms and signs. BAEPs may be altered.
be useful in evaluating patients with suspected This chapter reviews the method, interpre-
acoustic neuroma, multiple sclerosis (MS), or tation, and clinical applicability of BAEPs in
intrinsic brain stem lesions. The studies also patients with neurologic disease.
Figure 19–1. Neuroanatomy of the brain stem auditory pathways. (From Baehr, M., M. Frotscher, and P. Duus. 2005.
Topical diagnosis in neurology: Anatomy, physiology, signs, symptoms, trans. E. Taub. Stuttgart: Thieme Medical Publishers,
#81. By permission of the publisher.)
Brain Stem Auditory Evoked Potentials in Central Disorders 283
Purpose and Role of BAEPs (in Central I–V

Disorders) III–V
• To assess central auditory conduction I–III
pathways.
• To evaluate patients with suspected acous-
tic neuroma, MS, or intrinsic brain stem
lesions. Acoustic
• To assist in prognostication in comatose nerve
patients.
Pons
Midbrain
Medulla
AUDITORY ANATOMY AND Figure 19–2. BAEP waveforms and anatomical local-
PHYSIOLOGY ization. (From Misulis, K. E., and T. Fakhoury. 2001.
Spehlmann’s evoked potential primer, 3rd ed., 49. Boston:
Butterworth-Heinemann. By permission of the publisher.)
Knowledge of the neuroanatomy of the audi-
tory system is essential to understand the struc- from patients with brain stem lesions.1–3 It is
tures that are sequentially activated during not known whether BAEPs are generated by
BAEPs1 (Fig. 19–1, for color image, see color nuclei, tracts, or a combination of the two. The
plates). The auditory system begins with the following waves are routinely measured during
peripheral auditory apparatus. The cochlea and BAEP testing:
the spiral ganglion must be activated (with
• Wave I represents the distal action poten-
monaural stimulation) before the central audi-
tory pathways can be assessed. The initial cen- tial of CN VIII and appears as a negative
tral structure that is activated is the auditory potential at the ipsilateral ear electrode. If
portion of CN VIII, which enters the brain wave I is absent, central auditory conduc-
stem at the pontomedullary junction. Sequen- tion cannot be assessed reliably. Supple-
tial activation involves the cochlear nucleus in mentary electrodes, for example a needle
caudal pons, the superior olivary complex in electrode in the external auditory canal,
caudal to mid-pons, the lateral lemniscus in may help register this wave if it cannot
mid-pons, and the inferior colliculus in cau- be recorded with a conventional earlobe
dal midbrain. In normal subjects, hearing is electrode.
• Wave II may be generated by either
associated with bilateral activation of the audi-
tory pathway, maximal contralateral to the ear the ipsilateral proximal CN VIII or the
stimulated. BAEPs may be associated with acti- cochlear nucleus.
• Wave III is likely related to activation of
vation of the brain stem pathways that are
involved with sound localization rather than the ipsilateral superior olivary nucleus.
• Wave IV is produced by activation of the
with hearing.
nucleus or axons of the lateral lemniscus.
• Wave V appears to result from activation
of the inferior colliculus.
GENERATORS OF THE BRAIN • Waves VI and VII are presumed to
STEM AUDITORY EVOKED be generated by the medial geniculate
POTENTIALS body and the thalamocortical pathways,
respectively.
Stimulation of the peripheral auditory appa-
Key Points
ratus in normal subjects produces seven ver-
tex positive waveforms, conventionally labeled • Seven vertex positive waves are pro-
I–VII (Figs 19–2 and 19–3).1 Since waves VI duced during auditory nerve stimulation,
and VII are only variably present, they are although only waves I–V are consistently
not useful clinically. Although the specific gen- recorded in normal individuals.
erators of all the waveforms have not been • Wave I represents the action potential of
completely clarified, the anatomical regions CN VIII.
that are activated are known. These sites of • Waves III–V are generated from pathways
waveform generation are based on limited data in the CNS (brain stem).
II IV
Cz-A1 III V
Stimulate
left ear
Cz-A2 I IV
Stimulate V
right ear III
II
0.1 μV
11/second Clicks 2 4 6 8 10
ms
Figure 19–3. Normal BAEPs in a 26-year-old man. Waves I–V, characteristic vertex positive waveforms. (From
Benarroch, E. E., et al. 1999. Medical neurosciences: An approach to anatomy, pathology and physiology by systems and
levels, 4th ed., 513. Philadelphia: Lippincott Williams & Wilkins. By permission of Mayo Foundation for Medical Education
and Research.)
BRAIN STEM AUDITORY level. The optimal stimulus repetition rate for
EVOKED POTENTIALS: identifying waveforms is approximately 10 per
METHODS second. Since binaural stimulation may fail to
reveal abnormality in a patient with a unilat-
BAEPs are performed by stimulating the audi- eral auditory lesion, its use should be avoided.
tory nerve, typically with square wave clicks Therefore, monaural stimulation is used with
and recording the evoked response from the the contralateral ear masked by white noise.
scalp. Every BAEP study should begin with
an assessment of peripheral auditory function Recording
and a bedside test of auditory acuity should be
conducted to determine the hearing threshold. Surface electrodes are placed at the vertex
After assessing peripheral auditory acuity, the (CZ) and on each earlobe (A1 and A2) to
BAEP study is performed. The time required record the auditory waveforms (Fig. 19–4).
to perform BAEPs depends on several clinical Mastoid electrodes are not used routinely
factors, but the study usually can be completed because of increased muscle artifact. The chan-
in 30–45 minutes. nel derivations include ipsilateral ear to vertex
and contralateral ear to vertex. In some cir-
cumstances, an ipsilateral ear to contralateral
Stimulation ear derivation may assist in identifying wave I.
Wave I is identified in the ipsilateral ear deriva-
Conventional audiometric earphones are used tion as a negative peak (near-field potential) at
to deliver an electric square wave “click.” The the ear and as a positive peak (far-field poten-
stimulus that optimally activates the central tial) at the vertex. Wave II may be absent in
auditory system is maximal to the click thresh- normals, and wave IV may fuse with wave V.
old for each ear. The preferred stimulus for The contralateral ear channel may be useful
waveform recognition is 65–70 dB above the in identifying wave II and for distinguishing
click hearing threshold, the click sensation wave IV from wave V (Fig. 19–5). At least two
Figure 19–4. BAEP electrode placement and technique.
averages of 2000–4000 responses are obtained BAEPs is interpreting the study when there
from each ear. Additional trials may be neces- is a poorly formed or absent wave I. Several
sary to recognize waveforms. The optimal filter maneuvers may help to elicit a wave I if it is not
bandpass is 100–3000 Hz, and the bandpass identified with the standard setup, including
should be held constant for clinical BAEP stud- the following:
ies. Averaging is performed on 10 ms of data
after auditory stimulation. 1. Increasing the stimulus intensity if waves
The presence of a discrete, reproducible are low amplitude and poorly defined.
wave I and the demonstration of reproducibility 2. Decreasing the stimulus intensity if there
for all subsequent waveforms are critical for are too many or poorly defined waves.
rational interpretation of the study. One of 3. Changing the click polarity from rarefac-
the major technical challenges in performing tion to condensation.
Figure 19–5. Normal BAEP tracing, illustrating how the contralateral ear reference aids in identification of waves II, IV,
and V.
4. Slowing the stimulus rate. Age and Gender

5. Using an external canal supplementary
electrode (tiptrode). The age of the patient may affect wave-
6. Decreasing muscle artifact, using seda- form morphology and latency. BAEPs can be
tion with chloral hydrate or diazepam if recorded even in premature infants, but the
patients are unable to relax or if excessive absolute and interpeak latencies are longer
muscle artifact is present. than in older patients (Fig. 19–6).6 By 2 years of
age, the latencies are about the same as the nor-
Recent advances in BAEPs have focused on mal values of adult subjects. Persons older than
improved methodology. The three main areas 60 years have a statistically significant increase
of emphasis have been (1) increasing wave
definition, (2) decreasing recording time, and I–V IPL
(3) increasing the objectivity of detecting and Age
(ms)
identifying waves. Methods that use non- I V
(weeks)
conventional averaging formulas, steady state
responses, improved signal-to-noise ratio cal- 1
8.46
culations, and template correlation analysis
have been developed.4 These newer tech-
niques may have application for studies involv-
2
ing patients whose cooperation is minimal, for 7.74
example, in newborns and children. Probably,
equipment in which some of these techniques 3
are incorporated will become more available 7.00
commercially.
4
Key Points
6.33
• BAEPs are evoked with electrical square
wave clicks. 5
• Stimulus intensity should be 65–70 dB 5.88
above the click hearing threshold for opti-
mal waveform recognition. 6
• Monaural stimulation with contralateral 5.71
masking noise provides optimal recogni-
tion of unilateral auditory pathway lesions. 7
• Failure to identify wave I can be improved 5.49
with techniques such as changing stimu-
lus intensity, changing click polarity from 8
rarefaction to condensation clicks, slow- 5.16
ing the stimulus rate, using tiptrodes, or
decreasing muscle artifact. 9
• The contralateral ear channel recording 5.04
may help in identification of wave II and
in distinguishing wave IV from wave V. Stim AS 0.1 μV
75 dB HL
5/second
R click
2 4 6 8 10 12 14 ms
FACTORS AFFECTING THE BAEP
Figure 19–6. Effect of age (weeks) on wave I and wave
RESPONSE V absolute latencies in an infant of 28-week gestation. IPL,
interpeak latency; Stim AS, stimulation of left ear; HL,
Several physiologic and technical variables that hearing level. (From Stockard, J. E., and B. F. Westmore-
may alter BAEPs include age, sex, and audi- land. 1981. Technical considerations in the recording and
interpretation of the brain stem auditory evoked potiential
tory acuity.5 Factors related to the stimulus, for neonatal neurologic diagnosis. The American Journal of
including stimulus repetition rate, intensity, EEG Technology 21:31–54. By permission of the American
and polarity, also may affect BAEPs. Society of Electroneurodiagnostic Technologists.)
in BAEP latencies compared with those of Key Points

younger subjects. The BAEP interpeak laten- • BAEP latencies are longer in infants
cies are significantly shorter in women than
under the age of 2 and in adults over 60,
in men.
and are longer in men than women.
• Peripheral hearing loss may cause absence
of wave I, delayed absolute latencies of all
Auditory Acuity waveforms, or distorted or absent wave-
forms beyond wave I.
For patients with hearing loss, a higher stim- • Stimulus rates greater than 10 Hz may
ulus intensity is required to activate the cen-
increase the BAEP latencies and decrease
tral auditory pathways. Significant peripheral
amplitudes.
hearing loss may not allow brain stem audi- • Stimulus intensities less than 65–70 dB
tory conduction to be assessed. The absence of
above hearing threshold may increase
waves II–V in patients with a normal wave I
absolute and interpeak latencies, dec-
or an increased wave I–V interpeak latency
rease amplitude, and alter waveform
cannot be explained on the basis of hearing
morphology.
loss alone. Common BAEP findings in patients • Changing polarity of the click may
with hearing loss include a delay in the abso-
affect the amplitude and morphology of
lute latencies of all waveforms, absence of only
wave I.
wave I with delayed waves II–V, or absence of
all waveforms.2
INTERPRETATION OF BAEPs
Stimulus Rate
BAEP variables evaluated in patients with
Stimulation rates greater than 10 per second suspected neurologic disease include mea-
may be associated with a significant increase in surement of absolute waveform latencies and
absolute and interpeak latencies and a decrease interpeak latencies (I–III, III–V, and I–V) and
in waveform amplitude. determination of wave V/I amplitude ratio.7
Normative data obtained by similar methods,
preferably within the same laboratory, should
Stimulus Intensity be available to determine latency and ampli-
tude criteria for abnormal BAEPs. Right and
Decreasing stimulus intensity affects the left ear evoked potential studies should be
BAEPs much like hearing loss does. Stimu- compared only by using identical stimulus
lus intensities less than 65–70 dB above the variables.
sensation level increase absolute and interpeak The most common BAEP abnormality in
latencies, decrease waveform amplitude, and patients with CNS disease is a prolonged I–V
alter waveform morphology. interpeak latency (Fig. 19–7). Other alter-
ations include the absence of all waveforms,
a decreased V/I amplitude ratio, and preser-
Stimulus Polarity vation of wave I with poorly formed waves
II–V (Figs. 19–8, 19–9, and 19–10). The I–III
The polarity of the click produces movement and III–V interpeak latencies may be useful
of the earphone diaphragm away from the in determining the anatomical localization of
tympanic membrane (rarefaction) or toward auditory dysfunction. In patients with a pro-
the tympanic membrane (condensation). The longed I–III interpeak latency and a normal
former polarity may be preferred because of III–V interpeak latency, the auditory dysfunc-
an increase in wave I amplitude. Occasion- tion is assumed to be located between the distal
ally, condensation may produce a more obvious part of CN VIII (near the cochlea) and the
wave I. A mixture of the two polarities (alter- superior olivary nucleus, ipsilateral to the ear
nating) is not used routinely because of alter- stimulated. In patients with a prolonged III–V
ations in waveform morphology and interpeak interpeak latency and a normal I–III interpeak
latencies. latency, the auditory conduction defect likely
I V
1.76 6.12 F 24
Rt monaural 4.36
stimulation
Σ 4096
Σ 2048 × 2
I V
1.73 5.45
0.2 μV
Lt monaural 3.72
stimulation
Σ 4096
Σ 2048 × 2
70 dBSL
11/second
R click 2 4 6 8 10
ms
Figure 19–7. BAEPs in a patient with MS are abnormal because of prolonged I–V interpeak latency. The patient had no
symptoms or signs of brain stem disease, and the neurologic examination findings were unremarkable.
is located between the superior olivary nucleus CLINICAL APPLICATIONS

and the inferior colliculus, ipsilateral to the ear
stimulated (Fig. 19–11). The use of the BAEPs provides an objec-
tive physiologic measure that complements the
findings of the clinical history and examination
Key Points and neuroimaging. Categories of clinical prob-
lems for which BAEPs can be used include the
• Absolute and interpeak latencies (waves
following:
I–III, III–V, and I–V) are used in the
interpretation of the BAEP study. 1. Confirmation of brain stem abnormality
• The most common abnormality with CNS if the symptoms and signs of brain stem
disease is a prolonged I–V interpeak dysfunction are equivocal.
latency. 2. Confirmation of brain stem abnormality
• Determination of the region of prolonga- in patients known to have diffuse or mul-
tion using the I–III and III–V interpeak tifocal CNS disease.
latencies assists in determining anatomical 3. Screening for brain stem dysfunction
localization of the lesion. in patients who have symptoms that
, Age: 74 yrs
2.19 2.48
2/7/78
Pre-op
IV
I III IV
Δ IV–V
= –0.36 ms
2/20/78
Post-op
2.20 2.12
2 4 6 8 10 ms 2 4 6 8 10 ms
Rt monaural stimulation Lt monaural stimulation
(95 dB HL, masking A.S.)
Figure 19–8. Preoperative and postoperative BAEPs in a 74-year-old woman with a right acoustic neuroma associated
with brain stem compression. The tumor was resected. No response was observed after stimulation of the right (Rt) ear,
either preoperatively or postoperatively. Postoperatively, stimulation of the left (Lt) ear showed significant shortening of
the III–V interpeak latency related to resection of the lesion. (From Stockard, J. J., and F. W. Sharbrough. 1980. Unique
contributions of short-latency auditory and somatosensory evoked potentials to neurologic diagnosis. Progress in Clinical
Neurophysiology 7:231–63. By permission of S. Karger, AG.)
usually refer to a peripheral cranial nerve been determined. The most common neuro-
or sensory organ, for example hearing logic indication for BAEPs is evaluation for
loss, dizziness, peripheral facial weakness,
diplopia, and peripheral jaw weakness. III
I
The potential advantages of BAEPs over neu- IV

roimaging studies are lower cost, less discom- II V
fort for patients, and the ability to perform
the studies in patients who have a contraindi-
cation to magnetic resonance imaging (MRI).
Also, sequential BAEP studies are easier to
perform. A disadvantage of BAEPs is the lack
of specificity and sensitivity if the brain stem 0.1 μV
lesion does not involve the central auditory
pathways. Clinical studies have indicated that
BAEPs are complementary to MRI studies
for certain central auditory abnormalities, for
example acoustic neuromas.2 2 6 10 ms
BAEP abnormalities have been observed in
several specific neurologic disorders, including
cerebellopontine angle tumors, demyelinating
disease, brain stem tumors, brain stem infarcts : Age 39—unsteady gait
and hemorrhages, coma, and leukodystro- Figure 19–9. BAEPs in a 39-year-old woman with MS
phies.5 The sensitivity and specificity of show the combined abnormalities of prolonged I–V
BAEPs for these neurologic diseases have interpeak latency and decreased V/I amplitude ratio.
Stim. AS 49 dBSL Stim. AD 70 dBSL

2.61 ms
Cz-A1
1.69 ms
Cz-A2
1 μV
2 4 6 8 10 2 4 6 8 10
ms ms
Figure 19–10. BAEPs in a 30-year-old woman with MS reveal only wave I bilaterally. Waves II–V are absent. Stim. AS,
stimulation of left ear; Stim. AD, stimulation of right ear.
suspected acoustic neuroma or MS. The ratio- diagnostic yield in patients with normal BAEPs
nale for BAEP studies in these suspected dis- and suspected acoustic neuroma. Chiappa2
orders is to demonstrate an electrophysiologic has suggested that neuroimaging studies are
alteration indicative of a CNS abnormality and unnecessary in most patients with suspected
to provide information about the anatomical acoustic neuroma and normal BAEPs (if the
localization of the lesion. BAEPs may be use- studies are performed correctly). Other cere-
ful as a prognostic indicator in patients with bellopontine angle tumors, for example menin-
coma.8 Also, BAEPs have been used to monitor giomas, may not produce BAEP abnormalities
response to therapy.9 until the tumor is large and involves CN VIII.
Thus, the diagnostic yield of BAEPs for these
tumors early in the course of disease is low, and
Acoustic Neuroma neuroimaging is a more important neurodiag-
nostic technique.
BAEPs are a reliable indicator of the presence
of cerebellopontine angle tumors affecting CN Key Points
VIII (see Fig. 19–8).10 Auditory conduction
abnormalities almost invariably are found in • BAEPs are useful in the diagnosis of
patients with acoustic neuroma, even in those acoustic neuromas. The most sensitive
who are asymptomatic.11 BAEPs may be abnor- abnormalities are an absolute or asym-
mal when the findings of other audiometric metric prolongation of the wave I–III
studies and even neuroimaging studies are nor- interpeak latency.
mal. According to Chiappa,2 “BAEPs are the • The characteristic BAEP changes are pro-
most sensitive screening test when an acous- longed I–V and I–III interpeak latencies
tic neuroma is suspected.” The characteristic ipsilateral to the tumor.
BAEP changes are prolonged I–V and I–III • BAEPs may be abnormal when the find-
interpeak latencies ipsilateral to the tumor. ings of other audiometric studies and even
The absolute or asymmetrical prolongation of neuroimaging studies are normal.
the I–III interpeak latency may be the most • Normal BAEP results in a patient with
sensitive BAEP variable. Patients with auto- symptoms suggestive of acoustic neuroma,
somal dominant neurofibromatosis in whom such as dizziness and hearing loss, argue
bilateral acoustic neuromas develop may have strongly against the diagnosis.
normal BAEPs if the tumors are asymptomatic,
small, and confined to the intracanalicular
region. Normal BAEP results in a patient Demyelinating Disease
with symptoms suggestive of acoustic neu-
roma, such as dizziness and hearing loss, argue The frequency of BAEP abnormalities in
strongly against the diagnosis. MRI has a low patients with suspected MS is related directly
to the likelihood the person has the disorder Intrinsic Brain Stem Lesions
and the presence of clinical evidence for
brain stem disease.12 Abnormal BAEP results BAEPs are abnormal in most patients with
are common in patients with clinically def- brain stem tumors, for example pontine glioma
inite MS (Figs. 19–7, 19–9, and 19–10). In (Figs 19–11 and 19–12).5 The findings are sim-
a study of 60 patients with clinically defi- ilar to those in patients with other central audi-
nite MS and symptoms or signs of brain stem tory conduction defects. Brain stem tumors are
lesions, 34 (57%) had abnormal BAEPs. Of 33 often associated with bilateral BAEP abnor-
patients with brain stem disease and possible malities (maximal ipsilateral to the lesion).
MS, 7 (21%) had central auditory conduction Extra-axial tumors may not be associated with
defects.2 Importantly, BAEPs may be abnor- BAEP abnormalities unless there is direct
mal in patients with suspected MS who do compression on and disruption of the brain
not have evidence of brain stem lesions. Of stem. Ependymomas of the fourth ventricle
patients without symptoms or signs of brain and cerebellar tumors may also be associated
stem disease related to MS, 20%–50% may with central auditory conduction defects.
have abnormal BAEPs. However, the diag-
nostic yield of BAEPs in MS is much lower
IV–V
than with visual evoked potentials (VEPs) or
somatosensory evoked potentials (SEPs).
BAEP results in patients with demyelinating III
disease include a unilateral or bilateral pro-
longed I–V interpeak latency and a decreased I
V/I amplitude ratio (Figs. 19–7, 19–9, and
19–10).5 Characteristically, patients with bilat- VI VII
eral central auditory conduction defects do
not have auditory symptoms or abnormal click
Normal II
thresholds. The usefulness of BAEPs in these
cases includes identifying unsuspected brain
stem lesions in patients with an anatomically
unrelated disorder such as optic neuritis.
BAEPs may also provide confirmatory evi-
dence of a CNS alteration when neurologic
evaluation does not suggest the diagnosis of
demyelinating disease. BAEPs may implicate
brain stem disease in patients with vague,
ill-defined, nonspecific symptoms. Potentially,
BAEPs may be used to monitor the response
to treatment of patients with MS. Ultimately, Brain stem
MS is a clinical diagnosis, and the results of glioma
BAEP studies must be interpreted carefully
in conjunction with the rest of the neurologic
evaluation.
2.5 ms
Key Points
• Abnormal BAEP results are common in
patients with clinically definite MS.
• BAEPs may be abnormal in patients with
2 6 10
suspected MS who do not have evidence ms
of brain stem lesions.
• BAEP results in patients with demyelinat- Figure 19–11. BAEPs from a normal subject and a
patient with a brain stem glioma. The latter study shows
ing disease include a unilateral or bilateral prolonged III–V interpeak latency, indicating a central
prolonged I–V interpeak latency and a auditory conduction defect between the caudal pons and
decreased V/I amplitude ratio. caudal midbrain.
: Age 66 Key Points

Brainstem tumor
0.1 μV
2.71 2.90
• BAEPs are abnormal in most patients with
brain stem tumors.
11/7/77 IV/V:1 • Extra-axial tumors may not be associated
CZ – M1 = 0.71
Gait ataxia
with BAEP abnormalities unless there is
direct compression on and disruption of
I III IV V the brain stem.
12/22/77 IV/V:1
–
P irradiation; = 1.78
improved
Coma and Brain Death
2.52 2.28
Stim AS 60 dBSL
10/sec clicks
2 4 6 8 10 ms
BAEPs may be normal or abnormal in a
comatose patient, depending on the underlying
Figure 19–12. BAEPs in a 66-year-old woman before cause and the presence of a brain stem lesion.
and after (p–) radiation therapy for a brain stem tumor.
Note the significant shortening of III–V interpeak latency
BAEPs may provide prognostically important
after treatment. (From Stockard J. J., J. E. Pope-Stockard, information about the outcome.8 Patients in
and F. W. Sharbrough. 1992. Brainstem auditory evoked whom BAEPs are absent (if peripheral audi-
potentials in neurology: Methodology, interpretation, and tory dysfunction can be excluded) are unlikely
clinical application. In Electrodiagnosis in clinical neurol- to survive. Lesions confined to the cerebral
ogy, ed. M. J. Aminoff, 3rd ed., 402. New York: Churchill
Livingstone. By permission of the publisher.) hemispheres are not associated with abnormal
BAEPs unless the upper brain stem is function-
ally disrupted. BAEPs may be useful in assess-
The results of BAEP studies are variable ing the integrity of brain stem structures when
in brain stem strokes, that is infarcts and the findings on neurologic examination are
hemorrhages (Fig. 19–13). Although abnormal unreliable; for example, after treatment with
BAEPs are found in these patients, BAEPs a high dose of barbiturates or with the patient
are normal if the stroke spares the brain stem under general anesthesia. The brain-dead per-
auditory pathways, for example infarction of son invariably has abnormal BAEPs, with the
the posterior inferior cerebellar artery (lat- characteristic findings of the bilateral absence
eral medullary syndrome). Brain stem tran- of all waveforms or the presence of wave I and
sient ischemic attacks usually do not produce the absence of waves II–V bilaterally.
abnormal BAEPs.2
Key Points
V • BAEPs may provide prognostically impor-
III tant information about outcome in coma-
I II IV tose patients; patients with absent BAEPs
VI VII Normal short-latency are unlikely to survive.
auditory evoked potentials • The brain-dead person invariably has
from normal brain stem abnormal BAEPs, with the characteris-
tic findings of the bilateral absence of all
Cz + waveforms or the presence of wave I and
the absence of waves II–V bilaterally.
0.1 μV
I
II
Intraoperative Monitoring
BAEPs are useful for monitoring auditory
pathway integrity during neurosurgical proce-
2 4 6 8 10 dures, since they are resistant to the effects of
Figure 19–13. Top, BAEPs in a normal subject and,
drugs or anesthesia. Compression or vascular
bottom, in a patient who died of brain stem hemorrhage. insult to the eighth nerve or brain stem during
Note the absence of waves III–V in the latter study. surgery will alter the response.13–15
SUMMARY 4. Hwang, J. H., J. C. Chao, H. C. Ho, and S. H. Hsiao.

2008. Effects of sex, age, and hearing asymmetry
BAEPs are performed primarily in patients on the interaural differences of auditory brainstem
responses. Audiology & Neuro-Otology 13(1):29–33.
with suspected neurologic disorders to deter- 5. Markand, O. N. 1994. Brain stem auditory evoked
mine whether there is evidence of a brain stem potentials. Journal of Clinical Neurophysiology
lesion. BAEPs are highly sensitive to audi- 11:319–42.
tory conduction defects, but the findings are 6. Boettcher, F. A. 2002. Presbyacusis and the auditory
brainstem response. Journal of Speech Language &
not pathologically specific. BAEPs provide data Hearing Research 45(6):1249–61.
that are highly reproducible and objective and 7. Hall, J. W. III, and K. A. Rupp. 1997. Auditory brain
lend themselves to sequential studies for com- stem response: Recent developments in recording and
parison. BAEPs are noninvasive and can be analysis. Advances in Oto-rhino-laryngology 53:21–45.
performed not only in the clinical neurophys- 8. Guerit, J. M. 1992. Evoked potentials: A safe brain-
death confirmatory tool? The European Journal of
iology laboratory but also in a hospital room or Medicine 1:233–43.
the intensive care unit. Patient cooperation is 9. Nuwer, M. R., J. W. Packwood, L. W. Myers, and
not critical, and BAEP waveforms are resistant G. W. Ellison. 1987. Evoked potentials predict the
to the effects of drugs or anesthesia. Impor- clinical changes in a multiple sclerosis drug study.
tant factors that need to be considered for 10. Matthies, C., and M. Samii. 1997. Management
accurate interpretation of BAEPs include the of vestibular schwannomas (acoustic neuromas): The
patient’s age, sex, and auditory acuity. The diag- value of neurophysiology for evaluation and predic-
nostic yield of BAEPs has been confirmed in tion of auditory function in 420 cases. Neurosurgery
patients with acoustic neuromas, MS, or intra- 40:919–29.
11. Tandon, O. P. 1998. Average evoked potentials—
axial brain stem lesions involving the auditory Clinical applications of short latency responses. Indian
pathways. BAEPs appear to be complemen- Journal of Physiology and Pharmacology 42:172–88.
tary to structural neuroimaging studies such 12. Drislane, F. W. 1994. Use of evoked potentials in the
as MRI. diagnosis and follow-up of multiple sclerosis. Clinical
Neuroscience 2:196–201.
13. Kumar, A., A. Bhattacharya, and N. Makhija. 2000.
Evoked potential monitoring in anaesthesia and anal-
gesia. Anaesthesia 55:225–41.
REFERENCES 14. Bischoff, B., J. Romstock, R. Fahlbusch, M. Buch-
felder, and C. Strauss. 2008. Intraoperative brain-
1. Moller, A. R. 1994. Auditory neurophysiology. Journal stem auditory evoked potential pattern and perioper-
of Clinical Neurophysiology 11:284–308. ative vasoactive treatment for hearing preservation in
2. Chiappa, K. H. 1997. Evoked potentials in clini- vestibular schwannoma surgery. Journal of Neurology,
cal medicine, 3rd ed. Philadelphia: Lippincott-Raven, Neurosurgery, & Psychiatry 79(2): 170–5.
31–130. 15. Raftopoulos, C., B. Abu Serieh, T. Duprez, M. A. Doc-
3. Nuwer, M. R. 1998. Fundamentals of evoked poten- quier, and J. M. Guerit. 2005. Microsurgical results
tials and common clinical applications today. Elec- with large vestibular schwannomas with preservation
troencephalography and Clinical Neurophysiology of facial and cochlear nerve function as the primary
106:142–8. aim. Acta Neurochirurgica 147(7):697–706.
Chapter 20
Audiogram, Acoustic Reflexes, and

Evoked Otoacoustic Emissions
Christopher D. Bauch and Wayne O. Olsen
INTRODUCTION EVOKED OTOACOUSTIC

AUDIOGRAM EMISSIONS
Speech-Recognition Thresholds Neonatal Screening
Word Recognition Auditory Neuropathy/
Dys-synchrony
ACOUSTIC REFLEX Pseudohypacusis
CN VIII vs. Cochlear Findings
Disorders of CN VII APPLICATIONS
Disorders of Brain Stem SUMMARY
INTRODUCTION 1. Conductive—abnormality of the ear canal,

tympanic membrane, or middle ear ossi-
The auditory system can be assessed with a cles, or a combination of these.
variety of acoustic stimuli presented via stan- 2. Sensorineural—disorder of the cochlea or
dardized procedures, and these are routinely cranial nerve (CN) VIII.
used in the evaluation of patients for whom 3. Mixed—a combination of conductive and
there is concern about hearing or balance. sensorineural disorders.
Hearing tests performed with pure-tone and
speech stimuli are used to assess hearing func- Acoustic reflex tests assess contraction of the
tion in patients for whom there is concern stapedius muscle and require no voluntary
about hearing or balance. These tests can help behavioral response from the patient. They
determine whether there are related balance are used primarily to help differentiate sen-
and hearing problems, can document difficul- sory (cochlear) lesions from neural (CN VIII)
ties in communication attributable to hearing lesions and to evaluate a portion of CN VII.
disorders, and can help establish a diagnosis. Evoked otoacoustic emissions (EOAEs)
On the basis of the patterns of results from require no voluntary behavioral response,
these tests, hearing loss can be categorized as assess preneural function in the cochlea, and
follows: are useful screening tests for hearing in
295
newborn infants, for diagnosis of auditory assess thresholds of hearing (just barely audi-
neuropathy/dys-synchrony, and for document- ble) for air-conducted and bone-conducted
ing pseudohypacusis, that is, feigned or exag- stimuli. These tests are akin to the Schwabach
gerated hearing loss. and Rinne tuning fork tests. However, the pre-
Pure-tone and speech audiometric testing sentation of electronically generated signals
constitute the basic hearing evaluation of per- through standard transducers allows testing
sons who have a balance or hearing prob- over a greater frequency and intensity range
lem or both. Acoustic reflex tests often are and with far greater precision than is possi-
administered to patients who have tinnitus, ble with tuning forks. Children as young as
unilateral or asymmetric sensorineural hear- 6 months can be conditioned to respond to
ing loss, vestibular disorders, or facial paralysis pure-tone stimuli at or near threshold levels.
(or a combination of these) to help determine School-age and older children and adults need
whether the lesion involves the end organ or only to be instructed to provide a behavioral
the acoustic nerve portion of CN VIII or to response on hearing the designated signals.
help define the site of involvement of CN The responses of the patient are plotted on
VII. EOAEs are used to test cochlear func- a standardized chart called an audiogram and
tion in newborn infants, young children under- compared with internationally established ref-
going hearing evaluations, and older children erence levels for normal hearing.
or adults suspected of feigning or exaggerat- A conventional audiogram format is shown
ing hearing loss during routine pure-tone and in Figure 20–1, with intensity in decibels of
speech testing. hearing level (HL) (the American National
Standards Institute1 ) on the ordinate and fre-
Purpose and Role of Audiogram, Acoustic quency (125–8000 Hz) on the abscissa. The “0
Reflexes, and EOAEs decibel HL” line is an internationally accepted
reference representing the average hearing
• Audiologic tests can help determine sensitivity for young people with normal hear-
whether there are related balance and ing. The normal range indicated on the right
hearing problems, can document diffi- side of the audiogram extends to 25 dB HL,
culties in communication attributable to because persons with hearing thresholds in this
hearing disorders, and can help establish range, at least for frequencies of 500–4000 Hz,
a diagnosis. generally do not report difficulty hearing and
• The audiogram establishes the site of understanding conversational speech in quiet
pathology as conductive, sensorineural, surroundings. People whose hearing thresh-
or mixed. olds are
• Acoustic reflexes assess middle ear func-
tion and CN VII function, help differen- 1. In the 26–45 dB HL range have mild
tiate cochlear vs. CN VIII lesions, and, in hearing loss and difficulty hearing soft or
some instances, the integrity of the lower distant speech.
brain stem. 2. In the 46–65 dB HL range have difficulty
• EOAEs assess preneural function in the hearing speech at normal conversational
cochlea, require no behavioral response, levels and are considered to have moder-
and are useful hearing screening tests for ate hearing loss.
newborn infants, young children under- 3. In the 66–85 dB HL range have severe
going hearing evaluations, and for docu- hearing loss, indicating difficulty hearing
menting pseudohypacusis for older chil- even loud speech.
dren and adults suspected of feigning or 4. Greater than 85 dB HL have profound
exaggerating hearing loss during routine hearing loss.
pure-tone and speech testing.
The term deaf is reserved for people in group
4, and hearing impaired and hard of hearing
AUDIOGRAM are used for those in groups 1–3.
The threshold data shown in Figure 20–1
Basic audiologic tests use pure tones delivered reveal mild hearing loss in the left ear (marked
by standard earphones and bone vibrators to by “X” in Fig. 20–1), and in the right ear
Audiogram, Acoustic Reflexes, and Evoked Otoacoustic Emissions 297
Figure 20–1. Audiogram showing pure-tone and speech test results for sensorineural hearing loss (right ear) and
conductive hearing loss (left ear). Degree of hearing loss on right ordinate. ANSI, American National Standards Institute.
(marked by “O”), hearing sensitivity that ranges Word Recognition

from normal for the lower frequencies (125–
250 Hz) to mild hearing loss for the middle The ability to understand lists of monosylla-
frequencies (500–3000 Hz), and moderate to bles (one-syllable words such as “thin,” “sack,”
severe hearing loss for the higher frequen- and “vote,” presented singly) at clearly audi-
cies. The hearing loss for the right ear is ble levels is shown by the percentage correct
sensorineural, as shown by the interweaving word recognition score. In the example shown
air-conduction and bone-conduction thresh- in Figure 20–1, lists of monosyllables were
olds (marked by “O” and “[” respectively in presented at a sensation level of 25 dB, that
Fig. 20–1). The hearing loss in the left ear is, 25 dB above the speech-recognition thresh-
is conductive, as indicated by the separation olds. The score of 76% reveals less than per-
of the air-conduction and bone-conduction fect understanding of the test items through
thresholds (marked by “X” and “]” respectively the right ear. The score for the left ear is
in Fig. 20–1). perfect (100%), as expected for conductive
hearing loss. These results demonstrate that
speech loud enough to be heard easily was
Speech-Recognition Thresholds understood perfectly by the left ear (conduc-
tive hearing loss) but not by the right ear
The audiogram in Figure 20–1 also shows (sensorineural hearing loss). Scores of 90%–
speech-recognition thresholds that are based 100% indicate that people should have little
on correct responses to 50% of spondaic words difficulty understanding speech loud enough to
(spondees are words with two syllables of be heard easily in quiet environments. Correct
equal stress when spoken, such as “airplane” responses to 70%–88% of the monosyllables in
and “baseball”) at the hearing levels indicated a given list suggest occasional difficulty; 60%–
(40 dB HL). This measurement reflects the 68%, definite difficulty; 40%–58%, marked
level at which approximately 50% of spoken difficulty; and less than 40%, extreme diffi-
meaningful sentences would be understood. culty in understanding speech. Persons who
The levels in Figure 20–1 indicate mild hearing have CN VIII lesions often have consider-
loss for speech, as do pure-tone thresholds in able difficulty comprehending speech in the
the 500–2000 Hz frequency range. affected ear.2
Key Points to understand 50% of two-syllable words,

and reflect the level at which approximately
• Basic audiologic tests use pure tones deliv-
50% of meaningful sentences would be
ered by standard earphones and bone understood.
vibrators to assess thresholds of hearing • Word recognition results indicate how
(just barely audible) for air-conducted and well the patient understands speech at
bone-conducted stimuli. clearly audible levels.
• Pure-tone test results plotted on a stan- • Persons who have CN VIII lesions often
dardized audiogram format provide data have difficulty comprehending speech in
regarding the patient’s hearing status rel- the affected ear.
ative to international standards and also
delineate whether the hearing is conduc-
tive, sensorineural, or mixed. ACOUSTIC REFLEX
• Persons with hearing thresholds up to
25 dB HL, at least for frequencies of 500– Intense stimulation of either ear causes con-
4000 Hz, generally do not report difficulty traction of the stapedius muscle in both ears.
hearing and understanding conversational This protective response, called the acoustic
speech in quiet surroundings. reflex, is measured easily and quickly with a
• Persons whose hearing thresholds are in device called an immittance unit (Fig. 20–2).
the 46–65 dB HL range have difficulty The system microphone monitors the level of
hearing speech at normal conversational a low-frequency tone (226 Hz) maintained in
levels and are considered to have moder- the space between the probe tip sealed in the
ate hearing loss. ear canal and the tympanic membrane, while
• Persons with hearing thresholds greater intense tones (500, 1000, 2000, or 4000 Hz) are
than 85 dB HL are considered deaf. presented to the opposite ear or, in some sit-
• Speech-recognition thresholds reveal the uations, to the same ear as the 226-Hz probe
intensity level necessary for the patient tone. Contraction of the stapedius muscle,
Contralateral
Oscillator Off
500; 1,000; Attenuator lpsilateral
2,000; 4,000 Hz On HL
Probe
of immittance Earphone
Oscillator
226 Hz unit
Loudspeaker
Manometer
Air
pump
Reference Microphone
voltage
Balance
Amplifier
meter
Bridge
circuit
Figure 20–2. Block diagram of immittance unit showing the setting for eliciting contralateral acoustic reflexes (stimulus
presented through earphone). HL, hearing level.
which is attached to the neck of the stapes, CN VIII vs. Cochlear Findings
stiffens the middle ear system, thereby alter-
ing the level of the 226-Hz tone maintained The absence of acoustic reflexes or response
between the probe tip and the tympanic only to very intense tones in an ear with sen-
membrane. This change is the acoustic reflex sorineural hearing loss no worse than severe
response measured by the immittance unit. in degree makes one suspect neural (CN VIII)
Measurement of acoustic reflexes requires an involvement on the side of the stimulated ear.3
intact tympanic membrane, mobile middle Similarly, acoustic reflex decay—that is, dimin-
ear ossicles (no conductive hearing loss), ished amplitude of the acoustic reflex response
hearing adequate to allow sufficient stim- to less than half within 5 seconds to a 500-Hz or
ulation of the ear with at least one of 1000-Hz tone delivered 10 dB above the acous-
the above-mentioned tones, intact CN VII tic reflex threshold—suggests a lesion of CN
and CN VIII, intact reflex arc in the brain VIII. Elicitation of the acoustic reflexes by nor-
stem, and stapedius muscle attachment to mal levels of stimulation and the absence of
the stapes. Because of the complexity of reflex decay indicate that the middle and inner
this system, various response patterns emerge ears are normal or, in the case of sensorineural
(Table 20–1). In conjunction with the case his- hearing loss, indicate sensory (cochlear) abnor-
tory and other audiologic test results, acous- mality (Table 20–1). The sensitivity and speci-
tic reflex testing provides valuable diagnostic ficity of acoustic reflex and reflex decay tests
information. for identification of CN VIII lesions are 85%;
Table 20–1 Audiogram and Acoustic Reflex Findings for Various Conditions
Audiogram Acoustic Reflexes
Condition Normal Conductive Sensorineural Normal Abnormal
hearing hearing loss hearing loss response response
Normal X X
Conductive disorders
Cerumen plug X X
Thickened tympanic X X
membrane
Perforated tympanic X X
membrane
Otitis media fluid X X
Ossicular discontinuity X X
Otosclerosis stapes X X
fixation
Sensorineural loss
Cochlea X X
CN VIII X X
Facial paralysis
Proximal to stapedial X X
branch CN VII
Distal to stapedial X X
branch CN VII
CN, cranial nerve. (Modified from Keating, L. W., and W. O. Olsen. 1978. Practical considerations and applications of
middle-ear impedance measurements. In Audiological assessment, ed. D. E. Rose, 2nd ed., 336–67. Englewood Cliffs, NJ:
Prentice-Hall. By permission of the author.)
Superior Superior
olivary olivary
complex complex
CN VIII Ventral Ventral CN VIII

Cochlea cochlear cochlear Cochlea
nucleus nucleus
Facial Facial
motor motor
nucleus nucleus
CN VII CN VII
Stapedius Stapedius
Figure 20–3. Contralateral and ipsilateral acoustic reflex arcs. Crossing tracts for contralateral reflexes are shaded. CN,
cranial nerve. (From Wiley, T. L., and M. G. Block. 1984. Acoustic and nonacoustic reflex patterns in audiologic diagnosis.
In The acoustic reflex: Basic principles and clinical applications, ed. S. Silman, 387–411. Orlando, FL: Academic Press. By
permission of the publisher.)
that is, this test combination correctly identi- information regarding the mobility of the
fies lesions of CN VIII and correctly rules out middle ear system.
such lesions 85% of the time. • Measurement of acoustic reflexes requires
an intact tympanic membrane, mobile
middle ear ossicles, hearing adequate to
Disorders of CN VII allow sufficient stimulation of the ear with
at least one of the above-mentioned tones,
The presence of a normal stapedius muscle intact CN VII and CN VIII, reflex arc
contraction on the same side as facial paraly- in the brain stem, and stapedius muscle
sis reveals that the CN VII lesion is distal to the attachment to the stapes.
branch that innervates the stapedius muscle. In • Acoustic reflexes require intact CNs VIII
contrast, the absence of a reflex response on and VII and lower brain stem, and
the same side as facial paralysis indicates that help differentiate cochlear vs. CN VIII
the involvement of CN VII is proximal to the lesions.
stapedius branch of the nerve (Table 20–1). • Acoustic reflexes can assist in localizing
the site of involvement in instances of CN
VII or lower brain stem pathology.
• The absence of acoustic reflexes or
Disorders of Brain Stem
response only to very intense tones in
The absence of acoustic reflexes with contralat- an ear with sensorineural hearing loss no
eral stimulation (e.g., stimulating the right ear worse than severe in degree makes one
and measuring the acoustic reflex in the left ear suspect neural (CN VIII) involvement on
and vice versa) but their occurrence with ipsi- the side of the stimulated ear.
• Elicitation of the acoustic reflexes by
lateral stimulation (i.e., stimulating and mea-
suring the response in the same ear) indicates normal levels of stimulation and the
a brain stem lesion that interrupts the crossing absence of reflex decay indicate that
acoustic reflex tracts (Fig. 20–3). the middle and inner ears are nor-
mal or, in the case of sensorineural
hearing loss, indicate sensory (cochlear)
Key Points
abnormality.
• Acoustic reflex tests measure involuntary • The presence of a normal stapedius mus-
contraction of the stapedius muscle in cle contraction on the same side as facial
response to intense stimulation and yield paralysis reveals that the CN VII lesion
is distal to the branch that innervates the level in the ear canal, test time, and other
stapedius muscle. variables. The graph in Figure 20–5 shows the
• The absence of acoustic reflexes with amplitude of the distortion products generated
contralateral stimulation but their pres- within the cochlea (line graph near center on
ence with ipsilateral stimulation indicates left) in response to two tones presented simul-
a brain stem lesion that interrupts the taneously, the noise level in the ear canal, fre-
crossing acoustic reflex tracts. quency separation, level of the stimulus tones,
and test time.
Robust EOAEs, such as those shown in
Figures 20–4 and 20–5, indicate good function
EVOKED OTOACOUSTIC of the cochlear outer hair cells and are gener-
EMISSIONS ally associated with normal hearing sensitivity,
25 dB HL or better for frequencies of 1000–
EOAEs reflect the response of electromotile 6000 Hz.4, 5 Low-frequency physiologic noise in
activity within the cochlea in response to exter- the ear canal, and occasionally low-frequency
nal sound stimuli.4 This miniscule activity can environmental noise, precludes measurement
be measured in the ear canal with a sen- of otoacoustic emissions for frequencies less
sitive microphone sealed in the ear canal. than 1000 Hz.6 Nevertheless, observation of
The output of the microphone is averaged EOAEs for the 1000–4000 Hz range provides
to reduce the inherent physiologic and envi- important information, because it suggests
ronmental noise in the ear canal. Transient normal or near-normal hearing sensitivity for
evoked otoacoustic emissions (TEOAEs) are the frequencies most important for hearing
measurements of the active cochlear response and understanding speech. EOAEs are rarely
to clicks. Distortion product otoacoustic emis- observed at a given test frequency when sen-
sions (DPOAEs) reflect the interaction within sorineural hearing loss is 30 dB HL or greater
the cochlea to stimulation with two pure tones at that frequency or when conductive hearing
simultaneously. Displays of normal TEOAEs loss blocks transmission of the low-level otoa-
and DPOAEs are shown in Figures 20–4 coustic emissions from the cochlea back to the
and 20–5, respectively. The different segments microphone in the ear canal.4
in Figure 20–4 show the waveform, stability,
level, and spectrum of the click stimulus as well Neonatal Screening
as the waveform, reproducibility, level, and
spectrum of the response from the cochlea, Because EOAE tests can be completed quickly
signal-to-noise ratio of the response, the noise and do not require a voluntary response and
Figure 20–4. Display of TEOAE measurement showing test variables and response.
Figure 20–5. Display of DPOAE showing test variables and response.
because elicitation of a response suggests with varying degrees and configuration of

normal middle ear and cochlear function, they hearing loss, and often very poor speech or
have been used as hearing screening tests in word recognition, particularly in background
neonatal intensive care units and well-baby noise.8 The normal electromotile activity of
nurseries. The absence of a response raises the outer hair cells is revealed by the EOAE
the possibility of a significant hearing loss that responses, but the poor behavioral, acoustic
warrants further audiologic and medical eval- reflex, and BAEP responses indicate abnormal
uation. EOAEs are often used to help con- neural connections at the inner hair cells of the
firm behavioral observations made during an cochlea, a defective synaptic juncture between
audiologic evaluation of a young child. inner hair cells and the cochlear branch of CN
VIII, or a defective CN VIII itself.8 This entity
was delineated only recently with the discovery
of EOAEs.
Auditory Neuropathy/
Dys-synchrony
An unusual auditory disorder, termed auditory Pseudohypacusis
neuropathy/dys-synchrony, has been reported
in patients ranging in age from infancy to For patients who are not cooperative or who
adulthood.7, 8 The hallmark audiologic find- are suspected of feigning or exaggerating a
ings of this disorder are normal EOAEs in hearing loss greater than 30 dB HL, EOAE
the presence of absent acoustic reflexes and tests provide a fast and objective assessment
absent or severely abnormal brain stem audi- of the peripheral portion of the auditory sys-
tory evoked potentials (BAEPs) (see Chap- tem. If behavioral responses indicate a hear-
ter 21).9 Additional test findings may include ing loss that is mild or greater in degree for
poor or inconsistent responses to pure tones frequencies between 1000 and 6000 Hz (or
no behavioral responses are obtained), but administered to assist in determining whether

EOAEs are observed at various test frequen- the localization of a sensorineural hearing loss
cies, there is reason to suspect pseudohypa- is cochlear or involving CN VIII. If the audi-
cusis and the need for additional testing to ologist questions the accuracy or validity of the
establish hearing sensitivity more precisely. pure-tone and speech test results, otoacoustic
emissions tests may help identify pseudohy-
Key Points pacusis, or possibly auditory neuropathy/dys-
synchrony. The patterns of abnormalities in
• EOAEs measure electromotile activity of
different disorders are shown in Table 20–1.
the outer hair cells in the cochlea in
response to sound stimulation.
• These emissions most often are obtained
only when the middle ear mobility is nor- SUMMARY
mal and when hearing sensitivity is within
normal or near-normal limits (approxi- Audiologic testing in the form of pure-tone
mately 30 dB HL). air-conduction and bone-conduction audio-
• Otoacoustic emissions require no volun- grams provides diagnostic information about
tary response and as such are useful for the type of hearing loss (conductive, sen-
hearing screening in newborn infants, for sorineural, or mixed) and the degree of hearing
assessment of hearing in young children, loss and attendant communication difficulties.
for evaluation of auditory neuropathy/dys- The addition of speech tests that use spe-
synchrony, and for evaluation of persons cific types of speech stimuli directly assesses
suspected of feigning or exaggerating their the patient’s ability to hear and to understand
hearing loss (pseudohypacusis). speech. Acoustic reflex and reflex decay tests
• Auditory neuropathy/dys-synchrony is a are used to evaluate the integrity of a com-
rare disorder due to either abnormal neu- plicated neural network involving not only the
ral connections at the inner hair cells of auditory tracts to and through the brain stem
the cochlea, a defective synaptic junc- but also decussating pathways in the brain stem
ture between inner hair cells and the and the course of CN VII to the innervation
cochlear branch of CN VIII, or a defec- of the stapedius muscle. EOAE tests provide
tive CN VIII, and is characterized by nor- an objective measurement of the peripheral
mal EOAEs in the presence of absent hearing system from the external ear through
acoustic reflexes and absent or severely the cochlear outer hair cells. They are use-
abnormal BAEPs. ful screening tests for hearing in infants, in
patients suspected of auditory neuropathy/dys-
synchrony, and in patients suspected to have
pseudohypacusis, that is, feigned or exagger-
APPLICATIONS ated hearing loss.
Patients referred for audiologic evaluations

are routinely queried regarding their con-
cerns about hearing and balance and history REFERENCES
of such problems. Such information assists in
1. American National Standards Institute. 1996. Ameri-
the selection of subsequent appropriate audi- can National Standard Specification for Audiometers
ologic tests. The audiologic evaluation usu- (ANSI S3.6-1996). New York: The Society.
ally begins with pure-tone threshold tests (air 2. Van Dijk, J. E., J. Duijndam, and K. Graamans. 2000.
conduction and bone conduction) followed by Acoustic neuroma: Deterioration of speech discrimi-
speech-recognition and word-recognition test- nation related to thresholds in pure-tone audiometry.
Acta Oto-Laryngologica 120:627–32.
ing. These results yield basic information about 3. Bauch, C. D., W. O. Olsen, and S. G. Harner. 1990.
the patient’s hearing and communication status Preoperative and postoperative auditory brain-stem
as well as the type of hearing loss (conduc- response results for patients with eighth-nerve tumors.
tive, sensorineural, or mixed). Depending on Archives of Otolaryngology—Head & Neck Surgery
116:1026–9.
the specific diagnostic concerns and patterns 4. Robinette, M. S., and T. J. Glattke, eds. 2002. Otoa-
of findings on routine audiologic testing, other coustic emissions: Clinical applications, 2nd ed. New
studies, such as acoustic reflex tests, may be York: Thieme.
5. Lonsbury-Martin, B. L., and G. K. Martin. 2007. Otoa- 8. Hood, L. J. 2007. Auditory neuropathy and dys-
coustic emissions. In Auditory evoked potentials: Basic synchrony. In Auditory evoked potentials: Basic prin-
principles and clinical application, ed. R. F. Burkhard, ciples and clinical application, ed. R. F. Burkhard,
M. Don, and J. J. Eggermont. Baltimore: Lippincott M. Don, and J. J. Eggermont. Baltimore: Lippincott
Williams & Wilkins. Williams and Wilkins.
6. Headley, G. M., D. E. Campbell, and J. S. 9. Berlin, C. I., L. J. Hood, T. Morlet, et al. 2005. Absent
Gravel. 2000. Effect of neonatal test environment or elevated middle ear muscle reflexes in the presence
on recording transient-evoked otoacoustic emissions. of normal otoacoustic emissions: A universal finding
Pediatrics 105:1279–85. in 136 cases of auditory neuropathy/dys-synchrony.
7. Starr, A., T. W. Picton, and Y. Sininger. 1996. Auditory Journal of the American Academy of Audiology 16:
neuropathy. Brain 119:741–53. 546–53.
Chapter 21
Brain Stem Auditory Evoked

Potentials in Peripheral Acoustic
Disorders
Christopher D. Bauch
INTRODUCTION Interaural Latency

STIMULI Differences
ELECTRODES Interpeak Intervals
INTERPRETATION
Absolute Latencies APPLICATIONS
SUMMARY
INTRODUCTION BAEP findings. The utility of BAEPs in the

evaluation of central nervous system disorders
Brain stem auditory evoked potentials (BAEPs) is discussed in Chapter 19. However, of partic-
testing (also called brain stem auditory evoked ular interest here is the very small wave I, the
response [BAER]) is a useful technique for whole nerve action potential, generated in CN
the otoneurologic assessment of patients with VIII. Wave I is a critical component for neu-
complaints of dizziness, hearing loss, or tin- rodiagnostic assessment, which often requires
nitus. The BAEP evaluation records neuro- interpeak interval determination. Other waves,
electric potentials from cranial nerve (CN) especially waves III and V, are typically more
VIII and the ascending brain stem pathways robust and may be observed even though wave
that are elicited as a response to click stim- I often cannot be identified.
uli. When evaluating patients with complaints
of dizziness, hearing loss, or tinnitus, concomi- Purpose and Role of BAEPs in Peripheral
tant audiology testing may be necessary when Acoustic Disorders
considering BAEP evaluation since conductive
hearing losses resulting from abnormalities of • BAEPs can be used to evaluate CN VIII
the ear canal, tympanic membrane, or middle and ascending brain stem pathways.
ear as well as sensorineural disorders (lesions • The test helps to differentiate cochlear
of the cochlea or CN VIII) can affect BAEP from retrocochlear pathology in patients
waveform morphology and latencies. There- with complaints of hearing loss, tin-
fore, the potential effects of hearing losses nitus, dizziness, unsteadiness, or facial
must be considered in the interpretation of weakness.
305
STIMULI
Clicks are the most effective stimuli in BAEP
assessment since their short duration (50–
100 μs) and abrupt onset disperse acoustic
energy and provide good synchronization of
neural discharges across a broad frequency
range. However, the importance of high-
frequency hearing sensitivity is accentuated by
the spectral characteristics of the earphone
and by the response characteristics of the
ear canal and middle ear, resulting in greater
excitation in the frequencies above 1000 Hz, Figure 21–1. Etymotic ER-3A transducer, sound tube,
that is, 2000–4000 Hz range. Because this ear canal electrode (TIPtrode), and electrode lead. (From
region is stimulated maximally by clicks, rou- Nicolet Biomedical Instruments, Inc., Madison, WI.)
tine pure-tone assessment is recommended
before BAEP evaluation. Hearing losses, par-
ticularly in the 2000–4000 Hz frequency range,
can affect BAEP results.1, 2 Behavioral thresh- In direct comparison with mastoid elec-
olds for clicks are not an adequate screen for trodes, the ear canal electrode improves wave-I
hearing, because the click’s spectral spread of amplitude by nearly 100% for patients with
energy across low- and high-frequency regions normal hearing and 41%–127% for those with
can yield relatively good thresholds despite mild-to-severe hearing losses. In a large sam-
significant hearing loss in the 2000–4000 Hz ple of hearing-impaired patients, wave I was
range. identified easily 96% of the time with the ear
canal electrode, compared with 70% of the
time with mastoid electrodes.3 The primary
Key Points advantages of the ear canal electrode are that
• Stimuli are usually brief (50–100 μs) clicks wave-I amplitude is improved for all degrees of
that disperse acoustic energy and provide hearing loss and the foam material is compress-
good synchronization of neural discharges. ible, fits comfortably, and prevents collapse of
• Because hearing losses can affect BAEP the ear canals. Its disadvantage is that it can be
results, pure-tone audiometry is recom- used only once.
mended prior to BAEP evaluation.
Key Points
ELECTRODES • Ear canal electrodes enhance wave-I
amplitude and thereby allow additional
BAEPs are recorded with surface electrodes, analyses of BAEP measurements.
typically placed at or near the vertex and the
ears. Although conventional mastoid or earlobe
electrodes usually allow recording of waves I,
III, and V from patients with normal hearing INTERPRETATION
sensitivity, wave I is difficult or impossible to
identify in patients with mild, moderate, or Three basic measurements are often made in
severe cochlear hearing loss. An ear canal elec- the typical evaluation of BAEP waveforms:
trode can enhance wave-I amplitude. The elec- (1) absolute latencies, (2) wave-V interaural
trode is a disposable, soft, foam plug wrapped latency differences, and (3) interpeak inter-
in a thin layer of conducting foil. It couples vals. BAEP waveform amplitude may be an
to a transducer through flexible silicon tubing. unreliable criterion for clinical testing because
Such an electrode serves dual roles as a record- of marked variations among normal sub-
ing electrode and a stimulus delivery system jects and are therefore not routinely used in
(Fig. 21–1). interpretation.1, 4, 5
BAEPs in Peripheral Acoustic Disorders 307
Absolute Latencies • A 0.4-ms shift can be expected for each

10 dB of conductive hearing loss.
Absolute latencies of the BAEP waves may • Age and temperature usually do not affect
be influenced by peripheral conductive or the interpretation of these results.
cochlear hearing loss. Conductive hearing loss • Sedative effects on the BAEP are mini-
reduces the effective stimulus reaching the mal with the response never being totally
cochlea and causes absolute latency delays eliminated by anesthetics.
dependent on the degree of conductive impair-
ment. On average, a 0.4-ms shift can be
expected for each 10 dB of conductive hear- Interaural Latency Differences
ing loss.
For cochlear disorders, the hearing thresh- Another measure, interaural latency differ-
olds for 2000, 3000, and 4000 Hz are impor- ences, compares wave-V latencies at the two
tant. As hearing thresholds for these frequen- ears of the patient. The advantage of this mea-
cies become poorer, the latencies of waves I, sure is that the patient serves as his or her
III, and V increase systematically. Presumably, own control. Normal variability for interaural
this is because of the time delay of the traveling latency differences is 0.4 ms. Larger wave-V
wave in the cochlea reaching more responsive latency differences between ears are consid-
apical (lower frequency) areas and also because ered indicative of CN VIII involvement. How-
of the decrease in effective intensity stimulat- ever, the degree of hearing loss in the 2000–
ing the defective cochlea. These factors, and 4000 Hz range can also influence the validity
the fact that the cochlea produces more syn- of such comparisons.6 When wave-V latency
chronous responses at its high-frequency basal differences between ears exceed the predeter-
end, lead to latencies that depend on the mined criterion, the examiner must determine
integrity of high-frequency hearing. Clinical what influence, if any, the hearing loss has
experience has shown that when the average on the results. Adjustments in wave-V latency
hearing thresholds for 2000, 3000, and 4000 Hz based on various levels of high-frequency hear-
are equal to or greater than a 60-dB hearing ing loss have been advocated, but the appli-
loss, at least 60% of the absolute latencies for cation of these corrections is often misleading
waves I, III, and V are abnormal. Although and confusing.
absolute latencies may be delayed slightly for
elderly patients, such delays are quite vari- Key Points
able and age is usually not considered in the
interpretation of these results. Further, while • Interaural latency differences are an
body temperature can be a factor in BAEP important BAEP measurement.
latencies for seriously ill patients with hyper- • Wave-V latency differences between ears
thermia or hypothermia, it is usually not nec- can be influenced by hearing loss.
essary to document temperature for otherwise • Normal variability for interaural latency
healthy patients. Sedative effects on the BAEP differences is 0.4 ms.
are minimal, with the response never being
totally eliminated by anesthetics. Knowledge of
these tendencies is important in the interpre- Interpeak Intervals
tation of absolute wave-V latencies for patients
with a suspected CN VIII abnormality. Wave Interpeak intervals reflect the time inter-
latency delays for patients with similar cochlear val from one neural generator to another.
hearing losses are often indistinguishable from The primary interpeak intervals (I–III, III–V,
those of patients with lesions of CN VIII. and I–V) separate a delayed wave-V absolute
latency into its peripheral (I–III) and central
(III–V) components. Normal I–III and III–V
Key Points intervals are each approximately 2 ms, which
provide an overall I–V interval of approxi-
• Latencies of all BAEP waves may be mately 4 ms. An advantage of measuring the
delayed by both conductive and sen- interpeak intervals is that they usually are
sorineural hearing losses. not affected by moderate-to-severe levels of
cochlear or conductive hearing loss. A pro- nontumor group who have a similar degree of
longed I–V interval (longer than 4.54 ms) sug- cochlear hearing loss.
gests a retrocochlear lesion, whereas conduc- Interaural latency differences that exceed
tive and cochlear hearing losses usually have the 0.4-ms criterion identify more than 90% of
normal intervals (4.54 ms or less). The main the patients with CN VIII tumors. If the crite-
disadvantage to using the interpeak intervals rion is decreased to 0.3 ms, the rate of tumor
is that wave I cannot always be identified detection increases only slightly and the num-
if peripheral hearing loss is mild, moderate, ber of patients with cochlear hearing loss that
or severe. In these cases, the examiner must exceeds the 0.3-ms criterion is substantial. The
rely on the absolute latencies or interaural 0.4-ms criterion for interaural latency differ-
latency difference measures for interpretation. ences appears to be a reasonable compromise.6
When all three measures—absolute latency, Interpeak intervals have also been compared
interaural latency difference, and interpeak between tumor and nontumor groups matched
intervals—are used collectively, the sensitivity for hearing loss. The mean I–III interval for
of BAEP for CN VIII lesions is more than the tumor group exceeds that of the nontu-
90%,7–10 and the specificity for cochlear hearing mor group by approximately 0.6 ms, whereas
loss is nearly 90%.7 the mean III–V interval is similar for both
groups. The mean overall I–V interval is larger
Key Points by nearly a whole millisecond for the tumor
group. Only rarely does the I–V interpeak
• Interpretation of results is based on abso- interval for patients in the nontumor group
lute latencies, interaural latency differ- exceed 4.54 ms.
ences, and interpeak intervals. BAEP waveforms for a person with normal
• Absolute latencies are often influenced hearing (A) and for patients with a tumor of CN
by peripheral (conductive, cochlear) dis- VIII (B and C) are shown in Figure 21–2. The
orders, that is, the greater the degree of normal tracing (A) is well defined and depicts
hearing loss, the greater the latency delay. waves I, III, and V at the appropriate latencies.
• Interaural latency differences for wave V The lower tracings show abnormal I–III and
allow for the patient to serve as his or her I–V interpeak intervals (B) or the absence of
own control. waves following wave I (C).
• Interaural latency differences greater than Various BAEP latency indices and their sen-
0.4 ms are often considered suggestive of sitivity and specificity for CN VIII lesions are
CN VIII disorders. shown in Figure 21–3. These results were
• Interpeak intervals are increased by
lesions of CN VIII, the pons, and medulla.
• Interpeak intervals are not affected by
moderate-to-severe levels of cochlear or
conductive hearing loss.
APPLICATIONS
Absolute latencies of waves I, III, and V
have been compared between patients with
CN VIII tumors (tumor group) and patients
with cochlear hearing loss (nontumor group)
matched for pure-tone audiometric configu-
rations. Mean wave-I absolute latencies are
usually similar between tumor and nontumor
groups, but mean latencies for waves III and V
are prolonged by as much as 1 ms for the tumor
group. The range of latencies for waves I, III, Figure 21–2. BAEP recordings showing, A, normal
and V is also considerably larger for patients waveforms (I, III, V) and, B and C, abnormal waveforms
in the tumor group than for patients in the of patients with a CN VIII tumor.
BAEPs in Peripheral Acoustic Disorders 309
Figure 21–3. Percentage of abnormal (delayed or absent) BAEPs for 75 patients without tumor (nontumor) and 75
patients with CN VIII tumor (tumor) matched for hearing loss. I, III, V, BAEP waves; I–III, III–V, I–V, interpeak
intervals; ILD, wave-V interaural latency difference. (From Bauch, C. D., W. O. Olsen, and A. F. Pool. 1996. ABR
indices: Sensitivity, specificity, and tumor size. American Journal of Audiology 5:97–104. By permission of the American
Speech–Language–Hearing Association.)
obtained from 75 patients with confirmed CN synchronous neural activity due to a tumor will
VIII tumors who were matched audiometri- result in a reduction of the BAEP amplitude.
cally with 75 patients with cochlear hearing Although very time-consuming, preliminary
loss.7 The highest sensitivity for this group of studies with the stacked ABR amplitude tech-
CN VIII tumors was 92% when using abnormal nique have reported high sensitivity (95%)
wave-V interaural latency difference (greater for small tumors.2, 11 Further studies with the
than 0.4 ms) or abnormal I–V interpeak inter- stacked ABR are underway.
val. The specificity with these same criteria
was 88% (false-positive rate of 12%). Absolute Key Points
latency measures for waves III and V are also
sensitive for retrocochlear disorders, but they • Applications of latency measures help sep-
have an unacceptably high false-positive rate arate cochlear from retrocochlear disor-
(25% and 37%, respectively) because of the ders.
influence of cochlear hearing loss. • Overall sensitivity of BAEP is 92% for
Tumor size influences the sensitivity of tradi- patients with a CN VIII tumor.
tional BAEP latency measurement for patients • Interaural latency differences >0.4 ms
in the tumor and nontumor groups (Fig. 21–4). identify approximately 90% of patients
In a study that compared tumor size, five with CN VIII tumors without an exces-
BAEP indices had a sensitivity of 100% if the sively high false-positive rate for cochlear
tumor was larger than 2 cm. However, if the hearing losses.
tumor was 1 cm or smaller, the best sensitivity • Interpeak intervals >4.54 ms for waves
was 82%. I–V strongly suggest retrocochlear involve-
A recent BAEP method having potentially ment.
higher sensitivity to smaller tumors (<1 cm) • Absolute latency measures often indicate
is referred to as the stacked ABR (auditory retrocochlear pathology for patients hav-
brain stem response).11 Rather than relying on ing cochlear hearing losses.
various latency measurements obtained with • Traditional BAEP measurements success-
standard BAEP testing, the stacked ABR pro- fully indentify CN VIII tumors larger than
cedure uses special amplitude measurements. 2.0 cm, but are less sensitive for CN VIII
The underlying theory is that any reduction in tumors that are 1 cm or smaller.
Figure 21–4. Percentage of abnormal (delayed or absent) BAEPs for 75 patients with CN VIII tumor as a function of
tumor size. I, III, V, BAEP waves; I–III, III–V, I–V, interpeak intervals; ILD, wave-V interaural latency difference. (From
Bauch, C. D., W. O. Olsen, and A. F. Pool. 1996. ABR indices: Sensitivity, specificity, and tumor size. American Journal of
Audiology 5:97–104. By permission of the American Speech–Language–Hearing Association.)
SUMMARY 3. Bauch, C. D., and W. O. Olsen. 1990. Compari-

son of ABR amplitudes with TIPtrode and mastoid
Hearing sensitivity in the 2000–4000 Hz range electrodes. Ear and Hearing 11:463–7.
4. Jiang, Z. D., L. Zhang, Y. Y. Wu, and X. Y. Liu. 1993.
is important to BAEP assessment. Absolute Brain stem auditory evoked responses from birth to
latencies and interaural latency differences are adulthood: Development of wave amplitude. Hearing
often affected by increasing degrees of hearing Research 68:35–41.
loss in this frequency range, whereas interpeak 5. Burkard, R. F., and M. Don. 2007. The auditory brain-
stem response. In Auditory evoked potentials, basic
intervals are relatively stable measures, even principles and clinical application, ed. R. F. Burkard,
for patients with moderate-to-severe degrees M. Don, and J. J. Eggermont. Baltimore: Lippincott
of peripheral hearing loss. However, the reduc- Williams and Wilkins.
tion in amplitude or the absence of a measur- 6. Bauch, C. D., and W. O. Olsen. 1989. Wave V inter-
able wave I associated with peripheral hearing aural latency differences as a function of asymmetry in
2,000–4,000 Hz hearing sensitivity. American Journal
losses often makes it difficult or impossible to of Otology 10:389–92.
measure I–III or I–V intervals. Overall sensitiv- 7. Bauch, C. D., W. O. Olsen, and A. F. Pool. 1996.
ity of BAEP is 92% for patients with a CN VIII ABR indices: Sensitivity, specificity, and tumor size.
tumor. The false-positive rate for patients with American Journal of Audiology 5:97–104.
8. El-Kashlan, H. K., D. Eisenmann, and P. R. Kileny.
cochlear hearing loss is 12%. Tumor size influ- 2000. Auditory brain stem response in small acoustic
ences BAEP test results: the sensitivity is 100% neuromas. Ear and Hearing 21:257–62.
for CN VIII tumors larger than 2 cm, but it is 9. Haapaniemi, J. J., E. T. Laurikainen, R. Johansson,
only 82% for CN VIII tumors 1 cm or smaller. T. Rinne, and M. Varpula. 2000. Audiovestibu-
lar findings and location of an acoustic neuroma.
European Archives of Oto-Rhino-Laryngology 257:
REFERENCES 237–41.
10. Schmidt, R. J., R. T. Sataloff, J. Newman, J. R. Spiegel,
1. Hall, J. W. 1992. Handbook of auditory evoked and D. L. Myers. 2001. The sensitivity of auditory
responses. Boston: Allyn and Bacon. brain stem response testing for the diagnosis of acous-
2. Musiek, F. E., J. B. Shinn, and R. E. Jirsa. 2007. tic neuromas. Archives of Otolaryngology Head &
The auditory brainstem response in auditory nerve Neck Surgery 127:19–22.
and brainstem dysfunction. In Auditory evoked poten- 11. Don, M., B. Kwong, C. Tanaka, D. Brackmann, and
tials: Basic principles and clinical application, ed. R. Nelson. 2005. The Stacked ABR: A sensitive and
R. F. Burkard, M. Don, and J. J. Eggermont. Balti- specific screening tool for detecting small acoustic
more: Lippincott Williams and Wilkins. tumors. Audiology and Neurotology 10(5):274–90.
Chapter 22
Visual Evoked Potentials

Jonathan L. Carter
INTRODUCTION Patient cooperation

VISUAL SYSTEM ANATOMY AND Sedation and anesthesia
PHYSIOLOGY INTERPRETATION OF VEPs
VISUAL EVOKED POTENTIALS: LOCALIZATION OF VISUAL
METHODS SYSTEM LESIONS
Stimulation Anterior Visual Pathway
Recording (Prechiasmatic or Chiasmatic) Lesions
Visual Evoked Potentials in
FACTORS AFFECTING THE VEP Multiple Sclerosis
RESPONSE Other Anterior Visual Pathway Lesions
Visual acuity Posterior Visual Pathway
Pupillary size (Retrochiasmatic) Lesions
Age and gender
SUMMARY
INTRODUCTION or P100 wave, is generated by striate and

peristriate occipital cortex after visual stim-
Visual evoked potentials (VEPs) are noninva- ulation. The most common transient VEP
sive studies that measure the evoked responses abnormality in patients with anterior visual
to visual stimuli and assess the visual con- pathway lesions is prolonged latency of the
duction pathways through the optic nerves P100 wave.
and brain. VEPs allow a quantitative deter- Monocular P100-wave alterations in latency
mination of visual function and are highly have a higher diagnostic yield than physical
sensitive to lesions of the optic nerve and examination findings in patients with optic neu-
anterior chiasm but relatively insensitive to ritis. Full-field visual stimulation usually does
ophthalmologic disorders. They are usually not demonstrate abnormality in patients with
performed in cooperative patients with good unilateral retrochiasmatic lesions. The results
visual acuity. of VEP studies in cases of bilateral optic nerve
Pattern-reversal VEPs are performed by lesions may be indistinguishable from those
using a shift of a checkerboard pattern without of bilateral retrochiasmatic lesions. Alterations
changing luminance. Monocular visual stim- in test methods can assist with delineating
ulation is always preferred. In normal sub- retrochiasmatic lesions.
jects, the visual stimulus evokes a promi- This chapter reviews the method, interpre-
nent waveform with positive polarity in the tation, and clinical applicability of transient
posterior head region at a mean latency of full-field pattern-reversal VEPs in patients with
approximately 100 ms. This potential, the P1 neurologic disease.
311
Purpose and Role of VEPs VISUAL EVOKED POTENTIALS:

• To provide a quantitative determination of METHODS
visual function through the visual conduc-
tion pathways. Visual evoked potentials are performed by
• To assess for clinical or subclinical lesions stimulating the visual fields, usually with a
in the optic nerve or anterior chiasm, such checkerboard visual stimulus, and recording
as when evaluating a patient with multiple the evoked response using surface recording
sclerosis (MS). electrodes over the occipital lobe.
VISUAL SYSTEM ANATOMY AND

Stimulation
PHYSIOLOGY
Pattern-reversal VEP studies generally are per-
The visual system functions at several levels, formed with a shift of a checkerboard pattern
beginning with the retina and terminat- (black and white). No change in luminance
ing in several regions of the cerebral cortex1 (total light output) occurs with this form of
(Fig. 22–1, for color image, see color plates). stimulation.2 Studies that change luminance,
Each eye projects to both occipital lobes for example pattern-flash or strobe light, pro-
through the decussation of the axons from duce more variable results in normal sub-
the nasal half of each retina. Important struc- jects and are not as sensitive for detecting
tures involved in visual conduction include the abnormalities in visual conduction. Flash VEPs
macula, optic nerve, optic chiasm, optic tract, may be appropriate for intraoperative mon-
lateral geniculate body in the thalamus, and itoring of visual function.3 Monocular test-
thalamocortical pathways. The macula at the ing is preferred because binocular stimula-
posterior pole of the retina is specialized for tion may mask a unilateral visual conduction
high-acuity central vision. The primary visual abnormality. The patient is seated at a fixed
system projects to striate and peristriate areas distance, 70–100 cm, from the screen (usually
of the occipital cortex (Brodmann areas 17, 18, a television monitor) and is asked to focus
and 19). The occipital cortex projects to the on the center of the screen. In certain sit-
midtemporal cortex and the posterior parietal uations, the technician may need to verify
cortex. Cells in the visual cortex are most sen- that the patient is looking at the screen. The
sitive to movement and to edges. The retina patient should not be sedated before being
topographically transmits visual information to examined.
the occipital cortex. The macula projects to the The size of the checks used in the checker-
occipital poles, and more peripheral regions board pattern may affect the amplitude and
of the retina project to medial calcarine cor- latency of the P100 wave.4 The checks are mea-
tex. Different features of a visual stimulus sured by the degree of visual angle (minutes of
activate specific neurons in the visual system. arc [ ]). Checks that are between 28 and 31 are
For example, certain neuronal groups in the associated with stimulation of the central retina
retina and lateral geniculate body are involved and are usually satisfactory.5 Larger check sizes
primarily in detecting visual motion or color. may be necessary in patients with decreased
Neurons in the visual cortex also appear to visual acuity. A decrease in check size to 10
demonstrate these unique electrophysiologic in people with normal acuity is associated with
properties. an increase in the amplitude and latency of
Key Points the P100 wave. The fovea is most sensitive
to smaller checks and makes the largest con-
• The macula is specialized for high-acuity tribution to the P100 amplitude.6 However,
central vision and projects to the occipital smaller check sizes are inherently sensitive
poles. to ophthalmologic disorders, including poor
• The visual system projects to the stri- visual acuity, and are not used routinely in per-
ate and peristriate areas of the occipital forming VEPs. VEPs are routinely performed
cortex. with full-field stimulation. Hemifield stimula-
• Cells in the visual cortex are most sensitive tion can be performed in an attempt to localize
to movement and edges. retrochiasmatic lesions. A normal response to
A Optic nerve
3rd neuron
Ganglion
cells
Light
2nd neuron
Bipolar
cells
Impulses
1st neuron
Rods and
cones
Pigment
epithelium
Fovea Periphery
Visual cortex
Optic nerve
Optic tract
4th neuron
Lateral
Optic radiation
geniculate
body
Superior striate area

Area 17
Areas 19 and 18
Calcarine sulcus
Inferior striate area
Figure 22–1. Neuroanatomy of central visual pathways. (From Baehr, M., M. Frotscher, and P. Duus. 2005. Topical
diagnosis in neurology: Anatomy, physiology, signs, symptoms, trans. E. Taub. Stuttgart: Thieme Medical Publishers. #81.
313
hemifield stimulation shows paradoxical local-

ization, which refers to an ipsilateral distribu-
tion of the P100 wave to the hemifield being
stimulated.
The checkerboard pattern is reversed (black
to white to black) at a rate of 1 or 2 per second.
An increase of the stimulus rate to 4 per sec-
ond or greater may prolong the P100 latency.
Steady-state VEPs obtained with stimulus rates
of 8–10 per second are technically more dif-
ficult to perform and are not commonly used
to evaluate patients with suspected neurologic
disease.5
Recording
Electrodes are placed at Cz, Oz, A1, and Fz
(Fig. 22–2). Other acceptable electrode posi-
tions include midoccipital (5 cm above the
inion), right and left occipital (5 cm lateral
to the midoccipital electrode), and midfrontal
(12 cm above the nasion) regions. With full-
field stimulation, the P100 waveform is maxi-
mal in the midoccipital region but may be well
recorded between the inion and the vertex of
the head.
The low-frequency filter may range between
0.2 and 1.0 Hz; the high-frequency filter should
be 200–300 Hz. A sweep length of 200–250 ms Figure 22–2. Electrode placement for visual evoked
is used, and 100–200 responses are averaged. potential recordings.
Increasing the number of responses may pro-
duce a more favorable signal-to-noise ratio, but • Faster stimulus rates (>4 Hz) may pro-
the subject may find it difficult to maintain long the P100 latency.
fixation for a longer time. At least two trials • The size of the checks used in the checker-
should be performed before the P100 latency board pattern may affect the amplitude
is identified. The trials should be reproducible. and latency of the P100 wave.
The American Electroencephalographic Soci- • Larger check sizes may be needed in
ety guidelines recommend a check size of patients with decreased visual acuity.
approximately 30 .7 • A decrease in check size is associated with
The procedure for performing a VEP test an increase in the amplitude and latency
should be explained to the patient before the of the P100 wave.
study is initiated. Visual acuity and pupillary • Monocular testing is preferred because
size should be determined in each eye of binocular stimulation may mask a unilat-
the patient before the study is performed. If eral visual conduction abnormality.
appropriate, the patient should wear his or
her eyeglasses or contact lenses for the study.
Mydriatic drops should not be used before the
procedure. FACTORS AFFECTING THE VEP
RESPONSE
Key Points
A normal transient VEP to a pattern-reversal
• Pattern-reversal VEP studies are per- checkerboard is a positive midoccipital peak
formed with a shift of a checkerboard that occurs at a mean latency of 100 ms
pattern at a rate of 1–2 Hz. (Fig. 22–3).8 The waveform consists of three
Visual Evoked Potentials 315
separate phases: an initial negative deflection before the P100 latency becomes abnormal
(N1 or N75), a prominent positive deflection (Fig. 22–4). P100 latency is not prolonged with
(P1 or P100), and a later negative deflec- visual acuity of 20/200 or worse if large checks
tion (N2 or N145). The numbers used for the (greater than 35 ) are used. Therefore, subtle
waveform designation refer to the approximate changes in visual acuity, for example 20/40, do
latency (in milliseconds) in the normal popu- not explain significant prolongations of P100
lation. The amplitude and latency of the N1 latency.
and N2 waveforms are too variable in normal
subjects to be useful in interpreting VEPs in
patients with neurologic diseases. Pupillary size
A number of physiological and technical fac-
tors may affect the response recorded with Patients who have an asymmetry in pupil-
VEPs. Visual acuity, pupillary size, age, and lary diameter may have interocular differ-
sex may alter the P100 waveform in normal ences in P100 latency. Therefore, patients
subjects.5 should not have their pupils dilated before
undergoing VEP studies. A miotic pupil
may reduce luminance and prolong the
Visual acuity latency and decrease the amplitude of P100.
A pharmacologically dilated pupil may inc-
In the absence of an alteration in luminance, rease the luminance and produce distorted
visual acuity must be decreased to 20/200 waveforms.
F 25, ® eye
Cz-A1+2
Pz-A1+2
Oz-A1+2
N1
N2
Oz-Cz
5 μV
∑ 256 × 3
P1
50 100 150 200

ms
Figure 22–3. Full-field visual evoked potential after stimulation of the right eye of a 25-year-old woman. P1 wave-
form is maximal in the posterior head region (Oz electrode); P1 amplitude and latency are normal. (From Stockard, J.
J., J. F. Hughes, and F. W. Sharbrough. 1979. Visual evoked potentials to electronic pattern reversal: Latency variations
with gender, age, and technical factors. American Journal of EEG Technology 19:171–204. By permission of the American
Society of Electroneurodiagnostic Technologists.)
F 25
Oz-A1+2
Acuity corrected
to 20/20 87.8
Oz-Cz
Remove 87.9
contact
lens
Oz-A1+2
Acuity = 20/400
112.8
Oz-Cz
3 μV 113.7
Oz 50 100 150 200
ms
Figure 22–4. Full-field VEP obtained with and without a contact lens. Top, P1 amplitude and latency are normal. Bottom,
With reduction of visual acuity to 20/400, P1 morphology is distorted and latency is prolonged.
Age and gender Sedation and anesthesia

Age is a significant variable in determining nor- Because sedation and anesthesia abolish VEPs,
mal P100 latency. The latency increases with these studies are not useful for intraopera-
age and may be more marked in those older tive neurophysiologic monitoring. Flash VEPs
than 60 years. Also, women have a shorter P100 can be used in such situations, but the infor-
latency than men. This factor has to be consid- mation gathered is more qualitative than
ered when deciding on normative data for P100 quantitative.3, 5
latency.
Key Points
Patient cooperation • Technical parameters must be standard-
ized and normal values adjusted for both
If a normal subject chooses not to look directly age and gender before interpreting VEPs.
at the screen used for visual stimulation, • Factors such as visual acuity, pupillary
the P100 waveform may be distorted. Patient size, patient inattention, or failure to fix-
cooperation may be extremely important in ate on the screen may produce erroneous
a person with psychiatric disease and in very VEP results.
young and old people. Patients with oculomo- • Visual acuity must be decreased to
tor disorders (i.e., pendular nystagmus) may 20/200 before the P100 latency becomes
not be able to voluntarily fixate on the screen. abnormal.
• Patients with an asymmetry in pupillary pathway and the lack of specificity must be
diameter may have interocular differences emphasized.
in P100 latency. VEP studies provide an objective physio-
• The VEP may be altered in patients with logic measure that complements the results
oculomotor disorders or in those who are obtained for the clinical history and exami-
unable to cooperate and cannot voluntar- nation and from neuroimaging. Categories of
ily fixate on the screen. clinical problems to which VEP studies can be
• Sedation and anesthesia can abolish applied include the following:
VEPs.
1. Confirmation of a visual system abnor-
mality in the presence of current equiv-
INTERPRETATION OF VEPs ocal visual symptoms and signs.
2. Confirmation of a visual system abnor-
The interpretation of VEPs in patients with mality in the presence of known or sus-
suspected neurologic disease begins with the pected diffuse or multifocal central ner-
identification of the amplitude and latency of vous system disease.
the P100 wave. The results of VEP studies 3. Confirmation of a visual system abnor-
in normal subjects should be available in the mality when functional recovery has
laboratory to determine whether an absolute occurred after a past visual system insult.
P100 latency and the interocular difference The classic example is finding a P100
in latency are abnormal (Table 22–1). Each latency delay after a patient has recovered
evoked potential laboratory preferably should from an episode of optic neuritis in the
have its own normative data. An acceptable past.
alternative is to use published normal val- 4. Producing evidence for the nature of
ues obtained at a reference laboratory. Before the pathologic process.9 Demyelinating
VEP studies are performed, however, at least disease (e.g., multiple sclerosis) usually
20 normal subjects should be examined with produces significant P100 latency delays,
methods similar to those of the reference lab- with relative preservation of amplitude.
oratory. P100 latencies and interocular dif- Compressive or ischemic lesions often
ferences in latencies greater than the mean show amplitude loss, with relative preser-
plus three standard deviations are often used vation of latency. VEP changes in degen-
to identify abnormal studies. Absolute ampli- erative disease are more nonspecific, and
tude determinations are not particularly use- small changes in latency and amplitude
ful when interpreting a VEP study. An inte- are seen.
rocular difference in amplitude greater than 5. Localization of visual system lesions (this
50% may be considered abnormal if the asym- is considered below under the section
metry cannot be explained by technical fac- Localization of Visual System Lesions).
tors.6 However, amplitude abnormalities usu-
ally occur with latency abnormalities as well.
Certain lesions in the visual pathway may dis- Key Points
tort amplitude more than latency. In reporting • The P1 or P100 wave is the most repro-
VEP studies, the anatomical localization (or ducible and clinically useful waveform in
the lack thereof) of the lesion in the visual normal subjects or in those with neuro-
logic disease.
• Abnormalities in VEP latencies are much
Table 22–1 Normative P1 Latency
more important diagnostically than abnor-
Values Used at Mayo Clinic
malities of VEP amplitude.
• VEPs may provide objective evidence of a
LATENCY (ms)
current or previous lesion in visual path-
Age, year Females Males ways.
• While VEP changes are nonspecific, the
Less than 60 <115 <120
60 or older <120 <125 finding of prolonged latencies suggests a
demyelinating process.
LOCALIZATION OF VISUAL Visual Evoked Potentials in

SYSTEM LESIONS Multiple Sclerosis
Anterior Visual Pathway The most common neurologic disease asso-

ciated with a unilateral P100 abnormality is
(Prechiasmatic or Chiasmatic) demyelinating disease. An alteration of VEPs
Lesions may be identified in patients with a history of
optic neuritis who characteristically have no
Transient full-field VEPs are highly sensitive abnormality on physical examination. The sen-
to anterior visual conduction lesions. Unilateral sitivity of VEPs may be superior to that of mag-
P100 abnormalities indicate a visual con- netic resonance imaging (MRI) of the optic
duction defect anterior to the optic chiasm nerves in patients with optic nerve lesions due
(Fig. 22–5). An abnormal interocular difference to demyelinating disease, even when triple-
in P100 latency when both P100 values dose gadolinium with delayed imaging is used
are normal suggests an optic nerve lesion to image the optic nerve.10 VEPs should be
on the side of the increased value. Bilat- used to complement other neurodiagnostic
eral increased P100 latency values can be studies and should be correlated with the
found with bilateral optic nerve lesions, a clinical presentation before the diagnosis of
chiasmatic lesion, or bilateral retrochiasmatic demyelinating disease is made. VEPs may be
lesions (Fig. 22–6). However, if the interocular useful diagnostically in demonstrating a lesion
difference is abnormal when both P100 laten- in the optic nerve in patients with suspected
cies are prolonged, bilateral retrochiasmatic MS, who have disease localized to the cerebral
lesions are less likely. It should be emphasized hemispheres or spinal cord. VEPs are abnor-
that VEP abnormalities are nonspecific, and mal in 85% of patients with clinically definite
do not identify the pathology underlying these MS, 58% of those with probable MS, and 37%
changes. of those with possible MS.3 VEPs will detect
clinically unsuspected lesions in approximately
37%–41% of patients with MS. The most com-
mon VEP abnormality in patients with optic
Age: 42 Yrs
neuritis is an ipsilateral P100 latency prolon-
Right eye gation (Figs. 22–5 and 22–7). This may be
A2-Oz shown by an abnormality in the absolute P100
latency or with a prolonged interocular differ-
ence. The amplitude of the P100 wave may
118
be normal even when the latency is markedly
Left eye prolonged, especially after recovery from acute
A1-Oz optic neuritis. Virtually all patients with clini-
cally demonstrated optic neuritis have unilat-
5 μV
140 eral or bilateral abnormalities in VEPs.
In acute optic neuritis with severe alter-
50 100 150 200
ation in visual acuity, a P100 wave may not
ms be recorded. VEP amplitudes usually recover
within 3 months after an episode of acute
Figure 22–5. Full-field VEP in a 42-year-old man with
MS and optic neuritis in the left eye. P1 latency and optic neuritis and parallel the recovery in visual
amplitude are normal with stimulation of the unaffected acuity. The VEP may remain abnormal even
right eye. Absolute P1 latency is prolonged and the inte- years after the optic neuritis has resolved;
rocular difference is abnormal with stimulation of the however, improvement may also occur over
left eye. Note that the amplitude in the lower tracing
is preserved. This study suggests an anterior conduction
time. One series found that 17 out of 60
defect on the left. (From Benarroch E. E., B. F. West- eyes “improved” by at least 10 ms in latency
moreland, J. R. Daube, T. J. Reagan, and B. A. Sandok. with serial VEPs over 2 years.11 Another series
1999. Medical neurosciences: An approach to anatomy, found normal P100 latencies in only 3% of
pathology, and physiology by systems and levels, 4th ed., patients 3 months after an episode of acute
592. Philadelphia: Lippincott Williams & Wilkins. By per-
mission of Mayo Foundation for Medical Education and optic neuritis. After 2 years, however, the
Research.) P100 latencies were normal in 10% of affected
F 20
CZ-A1+2
PZ-A1+2
OZ-A1+2
OZ-CZ
OZ –
5 μV 50 100 150 200 ms 50 100 150 200 ms

Left monocular stimulation Right monocular stimulation
OZ +
Figure 22–6. P1 latencies are prolonged bilaterally, maximal on the left, in a patient who subsequently was shown to
have demyelinating disease. (From Stockard, J. J., J. F. Hughes, and F. W. Sharbrough. 1979. Visual evoked potentials to
electronic pattern reversal: Latency variations with gender, age, and technical factors. American Journal of EEG Technology
19:171–204. By permission of the American Society of Electroneurodiagnostic Technologists.)
Figure 22–7. Full-field VEP in a normal subject (left) and in a patient with an anterior visual conduction defect (right).
Note that the P1 latency is prolonged in the patient, with preservation of P1 amplitude.
eyes.12 Possible mechanisms for this improve- Patients may also experience gradual prolonga-
ment in latency include ion channel redis- tion of P100 latencies over time in the absence
tribution in the demyelinated optic nerve, of new clinical attacks of optic neuritis.13
partial remyelination of the optic nerve, or This situation is suggestive of a chronic pro-
cortical reorganization in the occipital cortex. gressive optic neuropathy and resembles the
progressive worsening in other functional sys- • A “delayed but well-preserved” P100

tems that occurs during the progressive phases waveform is defined as a “positive” VEP
of MS. Patients with optic neuritis may have in the McDonald Criteria for diagnosis of
a “bifid” P100 waveform, also known as a “W” MS, and may be particularly useful in the
waveform, where there is a small negative peak diagnosis of primary progressive MS.
in the middle of the large positive deflection.
While this finding can be seen in normal indi-
viduals, it appears to be more common in MS Other Anterior Visual Pathway
according to some authors.14 Lesions
VEPs have been incorporated into the
McDonald Criteria for the diagnosis of MS Tumors compressing the optic nerve and optic
(2001) and the revised McDonald Criteria chiasm or occurring within the optic nerve may
(2005).15, 16 A “positive” VEP is defined as a be associated with a unilateral P100 abnor-
“delayed but well-preserved” waveform. VEPs mality.3, 17 P100 latency may be prolonged; but
are most useful in the McDonald Criteria more commonly, the amplitude is decreased
in the diagnosis of primary progressive MS, disproportionately to the change in latency.
where they form one of the criteria for dissem- The morphology of the VEP may be markedly
ination of lesions in space, providing evidence distorted, and occasionally the P100 wave
for a second objective lesion in the central may not be recorded. Neoplasms associated
nervous system (Fig. 22–8). with optic nerve compression include optic
Key Points nerve gliomas, meningiomas, craniopharyn-
giomas, and pituitary tumors. Giant aneurysms
• Optic neuritis due to MS is the most com- may produce a similar optic nerve lesion.
mon cause of VEP abnormalities, usually Improvement in P100 waveforms may occur
prolonged P100 latency. after surgical removal of the tumor.
• VEP latencies may remain abnormal for Other anterior visual pathway lesions that
years after an episode of optic neuritis, may be associated with an abnormality in full-
or may “improve” or “worsen” over time field VEPs include anterior ischemic optic neu-
without further clinical attacks of optic ropathy, toxic (drug-induced) amblyopia, glau-
neuritis. coma, and Leber’s optic atrophy.2 The results
Figure 22–8. VEPs in a 60-year-old female who presented with a slowly progressive myelopathy over the preceding 2
years. Unilateral prolongation of P100 latency on the right gave evidence for dissemination of lesions in space, confirming
a diagnosis of primary progressive MS.
of the electrophysiologic studies must be corre- and interpretation of partial-field studies). The
lated with the clinical presentation to confirm clinical applicability of partial-field VEPs is
these diagnoses. uncertain because of developments in quanti-
tative visual perimetry and neuroimaging.
Patients with cortical blindness associated
Posterior Visual Pathway with various pathologic processes have been
(Retrochiasmatic) Lesions studied with transient VEPs.19 Importantly,
full-field VEPs have been reported to be nor-
The recording of full-field VEPs from the mal in patients with blindness and with neu-
midoccipital region usually does not show any roimaging and pathologic changes confined to
P100 abnormality in patients with unilateral the visual cortex.4 The sensitivity of VEPs in
posterior visual conduction defects. Bilateral patients with cortical blindness depends on
P100 abnormalities are seen in retrochiasmatic the anatomy of the cortical lesion and the
lesions, but this VEP result is nonlocalizing and method of the study. Lesions involving only
nonspecific. MRI has been shown to be more Brodmann area 17 (bilaterally) may be asso-
useful than full-field transient VEPs in eval- ciated with visual loss and normal VEPs. The
uating patients with retrochiasmatic lesions.18 use of smaller check sizes is important to iden-
Full-field VEPs may be normal even in patients tify changes in VEPs. Patients evaluated with
with abnormal neuroimaging findings retrochi- normal size checks, for example 27 , may have
asmatically or visual field defects or both. normal VEPs, but checks less than 20 usually
The diagnostic yield of VEPs is increased reveal an alteration.4 Normal VEPs obtained
with partial-field stimulation in patients with with large checks in patients with suspected
posterior visual conduction defects.5 Partial- cortical blindness should not be considered
field studies are not commonly performed evidence for functional visual loss. A normal
and require a modified method. They require P100 latency and amplitude in a blind person
the additional placement of lateral tempo- are highly unusual except for those with visual
ral electrodes (Fig. 22–9) (see Chiappa5 for cortex disease. Normal findings on a VEP study
a more complete discussion of the method virtually exclude an optic nerve or anterior
5 μV
ref
100 ms
Figure 22–9. Partial-field VEP after stimulation of the right hemifield. A P1 waveform maximal on the right is present
in the posterior head region. (From Stockard, J. J., J. F. Hughes, and F. W. Sharbrough. 1979. Visual evoked potentials to
electronic pattern reversal: Latency variations with gender, age, and technical factors. American Journal of EEG Technology
19:171–204. By permission of the American Society of Electroneurodiagnostic Technologists.)
chiasm lesion as the cause of visual loss.6 As 2. Tandon, O. P. 1998. Average evoked potentials—
noted above, with small checks, a significant Clinical applications of short latency responses. Indian
percentage of patients with retrochiasmatic Journal of Physiology and Pharmacology 42: 172–88.
3. Jones, N. S. 1997. Visual evoked potentials in endo-
lesions have changes in VEPs. However, in scopic and anterior skull base surgery: A review. Jour-
most patients with cortical blindness, the neu- nal of Laryngology Otology 111:513–6.
roimaging findings indicate the anatomy and 4. Celesia, G. G. 1993. Visual evoked potentials and
pathology of the lesion. electroretinograms. In Electroencephalography: Basic
principles, clinical applications, and related fields, ed.
E. Niedermeyer, and F. Lopes da Silva, 3rd ed.,
Key Points 911–36. Baltimore: Lippincott Williams & Wilkins.
5. Chiappa, K. H., ed. 1997. Evoked potentials in clinical
• VEPs do not show abnormalities in medicine, 3rd ed. Philadelphia: Lippincott-Raven.
patients with unilateral posterior visual 6. Epstein, C. M. 1990. Visual evoked potentials. In Cur-
conduction defects. rent practice of clinical electroencephalography, ed.
• Bilateral P100 abnormalities are nonlocal- D. D. Daly, and T. A. Pedley, 2nd ed., 593–623. New
York: Raven Press.
izing and can be seen in retrochiasmatic 7. American Electroencephalographic Society. 1994.
lesions or bilateral prechiasmatic lesions. Guideline nine: Guidelines on evoked potentials. Jour-
• VEPs may be normal in patients with cor- nal of Clinical Neurophysiology 11:40–73.
tical blindness, especially in lesions involv- 8. Aminoff, M. J., and D. S. Goodin. 1994. Visual
ing Brodmann area 17 bilaterally. evoked potentials. Journal of Clinical Neurophysiology
11:493–9.
9. Nuwer, M. R. 1998. Fundamentals of evoked poten-
tials and common clinical applications today. Elec-
SUMMARY troencephalography and Clinical Neurophysiology
106:142–8.
10. Acar, G., S. Ozajbas, H. Cakmakci, et al. 2004. Visual
The role of VEPs in evaluating patients with evoked potential is superior to triple dose magnetic
neurologic disease has evolved in an era of resonance imaging in the diagnosis of optic nerve
advanced neuroimaging techniques. MRI is involvement. International Journal of Neuroscience
clearly superior in sensitivity and specificity 114:1025–33.
11. Fuhr, P., A. Borggrefe-Chappuis, C. Schindler, and
to VEPs in detecting retrochiasmatic lesions. L. Kappos. 2001. Visual and motor evoked potentials
However, in patients with lesions involving the in the course of multiple sclerosis. Brain 124: 2162–8.
optic nerve and anterior chiasm, VEPs have 12. Brusa, A., S. J. Jones, and G. T. Plant. 2001. Long-
several important advantages: (1) VEPs are term remyelination after optic neuritis: A 2-year
objective and reproducible and may demon- visual evoked potential and psychophysical serial study.
Brain 124:468–79.
strate a functional abnormality that is not evi- 13. Jones, S. J., and A. Brusa. 2003. Neurophysiological
dent on physical examination or with neu- evidence for long-term repair of MS lesions: Implica-
roimaging studies; (2) VEP abnormalities may tions for axon protection. Journal of the Neurological
persist over time even when there is clini- Sciences 2006:193–8.
14. Rousseff, R. T., P. Tzvetanov, and M. A. Rousseva.
cal resolution of visual symptoms; (3) VEPs 2005. The bifid visual evoked potential—normal vari-
may be a more reliable indicator of disease ant or a sign of demyelination? Clinical Neurology and
than MRI (MRI may reveal nonspecific abnor- Neurosurgery 107:113–16.
malities that do not represent a pathologic 15. McDonald, W. I., A. Compston, G. Edan, et al. 2001.
process, such as nonspecific white matter sig- Recommended diagnostic criteria for multiple sclero-
sis: Guidelines from the International Panel on the
nal changes in the cerebral hemispheres); (4) Diagnosis of Multiple Sclerosis. Annals of Neurology
VEPs may be more sensitive than MRI for 50:121–7.
detecting abnormalities in optic nerves; and (5) 16. Polman, C. H., S. C. Reingold, G. Edan, et al. 2005.
VEP studies are less expensive than MRI stud- Diagnostic criteria for multiple sclerosis: 2005 revi-
sions to the “McDonald Criteria”. Annals of Neurology
ies and can be used in situations where MRI 58:840–6.
studies are contraindicated (i.e., pacemakers, 17. Ng, Y. T., and K. N. North. 2001.Visual-evoked poten-
aneurysm clips, etc). tials in the assessment of optic gliomas. Pediatric
18. Drislane, F. W. 1994. Use of evoked potentials in the
diagnosis and follow-up of multiple sclerosis. Clinical
REFERENCES Neuroscience 2:196–201.
19. Stockard, J. J., J. F. Hughes, and F. W. Sharbrough.
1. Celesia, G. G., N. S. Peachey, M. Brigell, and P. J. 1979. Visual evoked potentials to electronic pattern
DeMarco Jr. 1996. Visual evoked potentials: Recent reversal: Latency variations with gender, age, and tech-
advances. Electroencephalography and Clinical Neu- nical factors. American Journal of EEG Technology
rophysiology 46(Suppl):3–14. 19:171–204.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART C
Motor Pathways
Weakness, fatigue, loss of strength, and loss electrodes for stimulation. Surface electrical
of power are among the major symptoms of stimulation is adequate for stimulation of most
neurologic disease that can be assessed with peripheral motor nerves.
neurophysiologic testing. Strength and move- The recording of compound muscle action
ment are under the control of the motor sys- potentials described in Chapter 23 assesses
tem, which includes the central mechanisms motor nerve function in peripheral neuromus-
for integrating motor activity and the output cular disorders. Repetitive activation of com-
pathways. Reflexes and other central motor pound muscle action potentials, described in
control systems are discussed in Part E of this Chapter 24, assesses the function of the neu-
section. The electrophysiologic assessment of romuscular junction. Central stimulation of
peripheral motor pathways is reviewed in this motor pathways at the spinal cord or cortical
part and Part D. As with the sensory pathways, level evokes compound muscle action poten-
the most direct assessment of the motor path- tials, called motor evoked potentials, which
ways can be obtained with stimulation along is described in Chapter 25. The distinction
the motor pathway and measurement of the between the terms compound muscle action
response evoked by the stimulation. These potential and motor evoked potential is made
measurements can include the threshold for on the basis of the site of stimulation. Stim-
activation, the conduction time or velocity (or ulation of motor nerve fibers anywhere along
both) between the points of stimulation and their course after they leave the spinal cord
recording, and the size and shape of the evoked produces a response in the muscle called a
response. compound muscle action potential. Stimula-
Compound muscle action potentials rec- tion along the motor pathways in the spinal
orded directly from a muscle are measured cord or at the cortical level produces an iden-
for each assessment of the motor pathways tical muscle response called a motor evoked
whether activated centrally or peripherally. potential.
The method of application, the strength, and The use of motor evoked potentials for
the type of stimulus vary with the site along monitoring central motor function during
the motor pathway being stimulated. Stimula- surgery has been expanded recently. Com-
tion at the cortical level requires high-intensity pound muscle action potentials continue to be
electric or magnetic stimuli to produce use- the mainstay for providing insight into periph-
ful responses. Deep-lying motor nerves, such eral neuromuscular disease involving motor
as the spinal nerves, may require needle fibers.
325
Chapter 23
Compound Muscle Action

Potentials
James C. Watson and Jasper R. Daube
INTRODUCTION Temperature
GENERAL CLINICAL APPLICATIONS Age
RECORDING CMAPs
Type of Recording Electrode CMAP CHANGES IN DISEASE
Location of Recording Electrode Pathophysiology
Mechanisms of Conduction in
STIMULATION Myelinated Fibers
Type of Stimulating Electrode Mechanisms of Slow Conduction
Position of Stimulating Electrode in Disease
CMAP MEASUREMENTS
FINDINGS IN PERIPHERAL
Amplitude and Area
NERVE DISORDERS
Duration
Findings in Focal Lesions
Latency
Findings in Diffuse Peripheral
Conduction Velocity
Nerve Damage and Peripheral
Potential Errors in CMAP Measurements
Neuropathies
Normal Values in CMAP Recordings
Findings in Specific Focal
F WAVES Mononeuropathies
Methods Brachial Plexus Lesions
Radiculopathies
AXON REFLEXES (A WAVES)
PHYSIOLOGIC VARIABLES SUMMARY
AFFECTING THE CMAP
INTRODUCTION pathway is sufficient to activate some or all the

muscle fibers in that muscle. The CMAP is the
A compound muscle action potential (CMAP) summated activity of the synchronously acti-
is the action potential recorded from muscle vated muscle fibers in the muscle innervated
when stimulation anywhere along the motor by the axons and motor units represented in
327
that muscle. Therefore, a CMAP provides a lower motor neuron disease, neuromuscular
physiologic assessment of (1) the descending junction disease, and myopathies. NCS can
motor axons in the pathway below the level also assist when the weakness may be caused
of stimulation, (2) the neuromuscular junc- by hysteria, malingering, or upper motor neu-
tion, and (3) the muscle fibers activated by the ron disease. In these situations, the CMAP is
stimulus. Because disease of the axons, neuro- normal.
muscular junctions, or muscle fibers can alter CMAP recordings can go beyond confirm-
the CMAP, CMAP recording can be used to ing the presence of disease and the definition
assess disease at each of these locations. CMAP of severity by identifying disease that may not
recordings are least useful for assessing muscle be apparent clinically. For example, in patients
disease because the potentials are not altered with clinical evidence of a mononeuropathy,
until the disease is either severe or late in its CMAP recording may show signs of multiple
course, when marked atrophy and loss of mus- mononeuropathies or widespread peripheral
cle tissue occur. CMAP assessment for disease nerve damage that may not be apparent clin-
of the neuromuscular junction is discussed in ically. In patients with inherited neuropathies,
Chapter 24. CMAPs are also recorded with motor NCS can identify the process early in the
motor evoked potentials to assess central motor disease or when there is mild involvement and
pathways (see Chapter 25). The major appli- no clinical evidence of neuropathy. Motor NCS
cation of CMAP recording is in motor nerve can also identify disease early in its evolution,
conduction studies (NCS). for example, diabetes mellitus, when a mild
Motor NCS and CMAP recordings are peripheral neuropathy may not yet be apparent
equivalent. This chapter focuses on several clinically.1 In patients with an atypical distri-
aspects of CMAP recording as part of motor bution of deficits, the presence of anomalous
NCS and their applications. The chapter innervation can be traced. This is particularly
begins with a review of the techniques of useful for Martin-Gruber anastomosis (median
stimulation and recording, including technical to ulnar) in the forearm, Riche–Cannieu anas-
problems. The next section discusses modifi- tomosis (ulnar to median) in the hand, the
cations of the techniques of stimulation and accessory branch of the superficial peroneal
recording to obtain F-wave latencies and is fol- nerve in the leg, and crossed innervation after
lowed by a general discussion of the approach reinnervation.
to selecting motor NCS and CMAP recording A less common application of CMAP record-
for different clinical entities. ing is to identify and measure transient loss
of function in primary muscle disease such
as periodic paralysis. The recordings can also
be used to study abnormal reflex responses
GENERAL CLINICAL in upper motor neuron lesions. In selected
APPLICATIONS patients who have primary muscle disease, a
study of the mechanical twitch and its rela-
Recording of CMAPs in motor NCS is used tionship to electric events may be useful as
for several purposes in assessing neuromus- part of a CMAP recording. A recent report has
cular disease. CMAPs are particularly useful suggested that measurement of CMAP with
in providing objective measurements of the sequential incrementing stimulus intensities
extent and type of weakness. If the weakness can identify multiple different neuromuscular
is caused by a peripheral neuromuscular dis- disorders.2
ease, motor NCS can identify and localize the
sites of damage, whether from compression,
Purpose and Role of Motor NCS
ischemia, or other focal lesion. These studies
can also characterize the type of abnormality as • Provides objective assessment of the
a conduction block with neurapraxia or as slow- motor nerve, neuromuscular junction, and
ing of conduction at a localized area. They can muscle fiber without patient participation.
identify the changes associated with Wallerian • Identifies subclinical disease.
degeneration and regeneration in the motor • Defines pathophysiologic process (e.g.,
nerve. Measurement of CMAPs can assist in demyelinating vs. axonal).
distinguishing peripheral nerve disease from • Localizes focal disease.
Compound Muscle Action Potentials 329
• Assesses proximal conduction (e.g., F electrodes, and intramuscular electrodes each

waves). have advantages and disadvantages. Metal
• Defines severity of a peripheral disease. electrodes, stick-on electrodes, and saline
• Defines the extent of a neurogenic injury electrodes also have different electrical char-
(e.g., length-dependent peripheral neu- acteristics. These are summarized in
ropathy vs. mononeuritis multiplex). Table 23–1.
• Provides prognostic information (e.g., Normal values for CMAP recording depend
facial CMAP amplitude in Bell’s palsy). in part on the type of electrode used for record-
• Follows a disease to assess progression or ing; therefore, these values are most reliable
response to treatment. when the electrodes are identical. Most com-
• Estimates the number of axons in a nerve. monly, a 5–10 mm small surface electrode is
used because of the ease of application and
the availability of well-defined normal values.
Larger electrodes provide a better depiction
of CMAPs obtained from large muscles or
RECORDING CMAPs multiple muscles with a common innervation;
they also have better reproducibility.4 Gener-
Type of Recording Electrode ally, the use of these larger electrodes is limited
to laboratories in which normal values have
The electrodes that record CMAPs can alter been developed specifically for them. Also, in
the size, shape, and, to a lesser extent, the some laboratories, large electrodes are used to
latency of the response.3 Large surface elec- measure the number of motor units in a muscle
trodes, small surface electrodes, subcutaneous (see Chapter 29).
Table 23–1 Types of CMAP Recording Electrodes

Electrode size/type Uses Advantages Disadvantages
Small (5–10 mm) Most muscles Ease of use

surface electrode
Routine motor NCS Well-defined normal
values
Large surface Large muscles (anterior Ease of use
electrode tibialis, biceps)
Closely approximated Better reproducibility
muscles of similar
innervation
Subcutaneous Intraoperative monitoring Larger CMAP Discomfort
electrode (can be placed, secured as amplitude Infection, bleeding
surgical field defined risk
preoperatively) Higher impedance,
noise
Intramuscular Intraoperative monitoring Precision Discomfort
electrode Infection, bleeding
Deep muscles risk
Irregular waveform
Confirming surface CMAP (see text) with
response is from muscle variable amplitude
of interest and not and area makes
volume-conducted unsuitable for
comparison with
normal controls
Subcutaneous needle electrodes have the be difficult to measure because of irregular

advantage of being placed closer to muscle tis- initiation of the CMAP.
sue; therefore, they sometimes record higher
amplitude CMAPs. For some muscles, these
electrodes are easier to apply. The disadvan- Location of Recording Electrode
tages are those of any invasive technique,
including greater discomfort for the patient CMAP recording is made with an active elec-
and the (extremely low) risk of infection or trode (G1) and a reference electrode (G2)
bleeding. The subcutaneous needle recording whose locations are critical for the size, shape,
electrode occasionally has higher impedance, and latency of the CMAP. Normal values must
resulting in greater noise, shock artifact, or be determined with specific recording elec-
both. trode locations. The amplitude and area of the
Intramuscular needle or wire electrodes CMAP decrease with the distance of the active
have the advantage of recording from well- electrode from the muscle.4 The size and con-
defined small areas of muscle, thereby bet- figuration of the CMAP vary with placement of
ter isolating the CMAP of individual muscles. the active recording electrode over a muscle,
Intramuscular recordings are able to record as shown in Figure 23–1.5 The active elec-
small potentials, particularly of deep muscles trode (G1) is optimally placed over the end
that may not be recordable on the skin surface. plate region of the muscle, where the poten-
However, the configuration of the potential tial recorded is a well-defined waveform with
varies markedly with the precise location of the an initial negativity, a sharp inflection, and
recording electrode, which may shift during maximal amplitude. If multiple muscles are
the movement produced by the stimulation. activated or the end plate region is not well
Thus, amplitude and area measurements with localized, as in some disease, the CMAPs are
intramuscular recordings are not sufficiently more complex. Large electrodes will average
reliable to be useful clinically. Latencies may the differences between electrode locations
Figure 23–1. Amplitude and configuration changes in ulnar (upper) and median (lower) CMAPs in two normal subjects.
CMAPs were recorded from small electrodes in multiple locations in 839 grids over the thenar and hypothenar muscles to
show the variation in size and shape with electrode location. Note the double peaks, marked changes in potential over short
distances, and the differences between subjects. These variations make CMAPs highly susceptible to small differences in
electrode placement, especially with small recording electrodes. Note also the difference in pattern of distribution between
normal subjects. (From van Dijk, J. G., I. van Benton, C. G. Kramer, and D. F. Stegeman. 1999. CMAP amplitude cartog-
raphy of muscles innervated by the median, ulnar, peroneal, and tibial nerves. Muscle & Nerve 22:378–89. By permission
of John Wiley & Sons.)
to reduce the variation that can occur with or electrically silent).6–8 The contribution
different placements of small electrodes, and from the G2 electrode to CMAP morphol-
hence have better reproducibility than small ogy appears to influence the ulnar and tibial
electrodes.4 However, large electrodes are too CMAPs most significantly.9 The bifid appear-
large and impractical for the most common ance of the ulnar CMAP can be explained by
motor nerve conduction study recording sites the contributions of the large negative peaks
(intrinsic hand and foot muscles). from the G1 and an active G2 electrode.
A CMAP can be recorded with the active Because of this, it is particularly important
electrode far from the muscle, but it is maximal that the placement of both the G1 and the
when located directly over the muscle generat- G2 electrodes be placed at the same location
ing it. If the electrode is either off the motor with the same protocols that were used when
end plate or located at some distance from the normal values were obtained. In CMAPs
the muscle that is generating the CMAP, the with a large contribution from the G2 electrode
potential is predominantly or initially positive (ulnar and tibial), the CMAP is less sensitive to
in polarity and much smaller, with a signifi- precise positioning of the active G1 electrode.9
cantly slower rise time to the negative peak. Maximal amplitude is generally obtained with
(Fig. 23–2). The presence of an initial posi- the reference electrode over the tendon of the
tivity on a CMAP at all sites of stimulation is, muscle being recorded, optimally at the junc-
therefore, evidence that the active electrode tion of the tendon with the muscle. Because
is not over the end plate region of the mus- muscles vary in size, in motor NCS, the refer-
cle generating the CMAP and may be entirely ence electrode should not be at a fixed distance
off the muscle. The slope or rate of rise of from the active electrode (an important dif-
the positive-to-negative peak of the CMAP is a ference from sensory NCS). CMAPs recorded
rough gauge of the distance between the active with fixed interelectrode distances often have a
recording electrode and the muscle generating shorter duration, lower amplitude, and smaller
the CMAP. area, and, occasionally, a different configura-
The location of the reference electrode, tion; therefore, they should not be used.10
sometimes referred to as the inactive, G2,
or terminal-two electrode, also has an effect Key Points
on the amplitude and configuration, as well • Recording electrode type and location in a
as on the onset latency of the CMAP (i.e., laboratory should be consistent and have
the G2 electrode is not in reality “inactive” well-defined normal values. Different
A B C
SUBMAXIMAL MAXIMAL SUPRAMAXIMAL
1. Ulnar
2. Between
ulnar
and
median
3. Median
Figure 23–2. Summation of CMAPs recorded from thenar muscles with stimulation of the median and ulnar nerves.
Rows 1 and 3 show the CMAPs obtained with isolated stimulation of each of the two nerves. Note the initial positivity with
ulnar stimulation that results from recording CMAPs from ulnar-innervated thenar muscles at a distance from the thenar
recording electrode (placed over the median-innervated abductor pollicis brevis). Row 2 shows the effect of simultaneous
stimulation of both median and ulnar nerves, with summation of the potentials recorded in rows 1 and 3.
types of recording electrodes should have Needle electrodes. Needle electrodes, 1–2 cm,
normal values and should be used only for that are entirely uninsulated or a longer needle
defined clinical purposes. that is insulated except for 1–2 mm at its tip can
• Active recording electrodes (G1) should also be used to stimulate motor nerves. These
be placed at the site of maximal amplitude electrodes are particularly useful for stimulat-
with no positivity. ing deep nerves such as the median and ulnar
• G1 electrode is mispositioned off the mus- nerves in the forearm and the tibial and sci-
cle end plate if there is an initial positivity. atic nerves in the leg. This type of stimulus
• Inactive electrodes (G2) should be placed is often less painful, despite the needle stick,
over the tendon (G1–G2 fixed distance than surface stimulation that requires higher
placements are not reliable). currents or excessive surface pressure to elicit a
supramaximal response (i.e., obese patients or
in tender areas such as the femoral triangle).
STIMULATION However, needle electrodes are more difficult
to move when attempting to find the optimal
CMAP recording requires stimulating a nerve location for nerve stimulation. A stimulating
at site(s) along its length. Stimuli can be needle for long periods (e.g., repetitive stim-
applied in several ways. The stimulation tech- ulation) may be displaced with limb movement
nique used to activate a nerve affects the values and loss of the supramaximal response.
obtained.11 Magnetic stimulation. Magnetic stimulation
can activate some but not all peripheral nerves
and is seldom used for neuromuscular elec-
Type of Stimulating Electrode trodiagnosis. The site of onset of the initia-
tion of the action potential cannot be precisely
Electrical stimulation is applied through a
defined with magnetic stimulation. Despite the
cathode (negative) and an anode (positive) that
advantage of minimal discomfort with mag-
may vary in size and shape. Electrodes over and
netic stimulation, especially with deep nerves,
in parallel with the nerve evoke the most repro-
the inability to assure maximal stimulation and
ducible responses with the lowest stimulus
to accurately calculate velocities precludes its
intensity. The advantages and disadvantages of
use for routine NCS.
different types of stimulating electrodes must
be understood to select the optimal electrode
for each motor nerve.
Handheld electrical surface stimulator. The Position of Stimulating Electrode
commonly used handheld surface stimula-
tor allows the electrodes to be moved eas- Cathode–anode relationship. Depolarization of
ily in search of the nerve. Smooth, rounded, motor axons occurs at the cathode. The anode
5–10 cm electrodes mounted on the end of hyperpolarizes the nerve and may block con-
curved, removable stimulating poles permit duction of an action potential through the area
rapid change of the anode and cathode posi- of hyperpolarization (“anodal block”). Activa-
tions. This stimulator is more convenient for tion of a motor axon requires areas of both
stimulating nerves that may require pressing depolarization and hyperpolarization along the
on the overlying skin so the electrode is closer length of the axon, with current flow through
to the nerve and for rotating the position of the axon between the two locations.11 There-
the anode to reduce shock artifact. This type of fore, the optimal position of stimulating elec-
electrode is optimal for standard motor NCS. trodes is for the cathode to be as close as
possible to the nerve between the anode and
Flat disc electrodes. When stimuli have to be the recording site so that the activated action
applied for longer periods, as in testing peri- potential does not traverse the area of hyper-
odic paralysis and measuring motor unit num- polarization at the anode. The optimal location
ber estimates, flat disk electrodes taped on the of the anode is longitudinally along the course
skin over the nerve or a pair of electrodes of the axon away from the recording electrode.
mounted in a bar with the electrode protruding Ideally, the anode and cathode are adjacent to
from the bar allows more stable positioning of the nerve and only a few millimeters apart so
the electrode. that all current flow is directed through the
nerve being tested and not into surrounding become necessary to locate the anode perpen-
muscle, another nerve, or other tissue. dicularly to the nerve as the anode is rotated
The ideal position of surface stimulating to find a position of minimal shock artifact.
electrodes is along the length of the nerve, with Occasionally, the anode may need to be rotated
the cathode closest to the recording electrode. excessively to a position where it sits closer to
The anode and cathode must be farther apart the G1-recording electrode than the cathode to
than for needle electrode stimulation. If the eliminate the shock artifact.
anode and cathode are too close, current flow
passes directly between them without entering Localizing the nerve with stimulator (sliding).
the tissue to the depth of the nerve. Thus, acti- The location of most nerves can be iden-
vation of all motor axons may not occur despite tified reasonably well from anatomical land-
the use of high voltage and the passage of a marks for each nerve. However, it must always
large current. For most motor nerves, a dis- be remembered that the exact location of a
tance of 3–5 cm between the anode and the nerve can vary significantly among normal sub-
cathode is sufficient for adequate current to jects. The most striking example is the per-
penetrate the tissues to the depth of the motor oneal nerve at the ankle; its position can vary
axons. For nerves that are very deep in the tis- from 0.5 to 4 cm lateral to the tibia. There-
sue, a greater distance between the anode and fore, when attempting to stimulate a motor
the cathode may be necessary. This increases nerve, the nerve must be localized to minimize
the depth and diameter of the depolarizing stimulus intensity for lessened patient discom-
stimuli, increasing the risk of inadvertent stim- fort and to decrease the likelihood of current
ulation of other nerves and muscles and of spread to other nerves. Placing the stimulat-
stimulating the nerve of interest at some dis- ing electrodes at the location judged to be over
tance from the intended site of depolarization the nerve and then obtaining an initial low-
at the cathode (making measurements for dis- amplitude CMAP best accomplishes this. The
tal latencies and conduction velocities prone stimulating electrode is then moved medially
to error). This is known as the virtual cathode or laterally perpendicularly to the nerve with-
effect. out changing the stimulus intensity. If the sub-
sequent CMAPs have increasing amplitude,
Shock artifact from stimulator location. The the electrode is being moved closer to the
anode may also be placed perpendicular to the nerve. However, if the amplitude decreases,
course of the nerve and lateral from it. The the electrode is being moved away from the
anode may need to be on the opposite side nerve. The electrode continues to be moved
of the limb, for example, to activate the tib- until the maximal amplitude is obtained with
ial nerve in an obese patient. A perpendicular the original stimulus intensity. This is known
location requires a higher current intensity to as sliding. The voltage is then increased until
obtain depolarization, increasing the possibil- the CMAP does not increase further with
ity that adjacent nerves will be stimulated. The a 25%–30% increase in applied voltage or
most common need for the lateral position is current.
when the stimulating and recording electrodes
are placed so close that a prominent shock arti- Needle stimulator position. Needle electrodes
fact occurs in the recording. The shock artifact can be placed immediately adjacent to the
occurs because the current flow from the stim- nerve, but this may require considerable prob-
ulating electrode spreads through the tissue ing in the tissue. The optimal location of a
directly to the recording electrode and charges needle electrode can be obtained by repeated
the capacitance of the intervening tissue, which stimulation to identify the region of minimum
then discharges over 2–20 ms, with a waveform threshold. When the anode and cathode are
superimposed on the CMAP (or sensory nerve both immediately adjacent to the nerve, stim-
action potential, SNAP). This occurs especially uli of less than 2 mA are adequate for activating
when the distance between the stimulating and all the motor axons. An anode at some dis-
recording electrodes is short, such as the tibial tance from the nerve, either on the surface or
or sural nerve stimulation at the ankle, mixed elsewhere in the tissue, may be used with the
motor and sensory median and ulnar palmar needle cathode near the nerve. A distant anode
responses, and the facial nerve at the angle can result in a somewhat higher threshold for
of the mandible. In these situations, it may activation, a greater risk of current spreading to
the surrounding nerves, and a less accurate site wave (M for motor). It is characterized by
of stimulation. These disadvantages are gener- several specific measurements, each of which
ally outweighed by the advantage of not having reflects the physiologic activity occurring in the
to probe the tissue with the anode to find the muscle or nerve.
optimal location near the nerve. The invasive
nature of needle stimulation and the time it
takes to achieve optimal location of the stim- Amplitude and Area
ulating electrode have made it less accepted
than surface stimulation, unless a deep, focal The most valuable measurement is the size
conduction block is likely. of the CMAP, measured as either the ampli-
tude or area. Both of these variables reflect
In normal subjects, these techniques allow the total number of muscle fibers that con-
supramaximal or full amplitude CMAPs to be tribute to the potential. In most laboratories,
obtained with stimulus intensities less than amplitude is measured from the baseline to the
20 mA (100 V) in the arm and less than 40 mA peak. Recall that for CMAPs, an initial posi-
(200 V) in the leg. In obese subjects or in cases tivity generally indicates that the G1 electrode
of particularly deep nerves and in patients with is not placed appropriately over the muscle
peripheral nerve disease, a greater intensity end plate and that the G1 electrode should be
of current may be needed to activate motor moved. However, with certain disease states,
nerves. The intensity of a stimulus applied to anomalous innervation (Martin-Gruber Anas-
a motor nerve is defined by total current flow, tomosis), or excessive stimulation stimulating
which is a function of the intensity of the adjacent nerves and adding a volume con-
applied voltage, the resistance to current flow, ducted distal response to the CMAP waveform,
and the duration of the stimulus. Pulses of 0.1– one may not be able to eliminate this initial
0.2 ms are usually adequate for stimulation of positivity and in that situation the CMAP is
motor nerves, but longer durations of up to measured from the positive peak to the neg-
1 ms may be necessary for deep or diseased ative peak. The area of the CMAP is related
nerves. most directly to the number of muscle fibers
or motor units that contribute to the CMAP.
Key Points Changes in the CMAP amplitude and area are
also measured as part of repetitive stimula-
• Depolarization of the nerve occurs at the tion during assessment of neuromuscular junc-
cathode. tion disorders or in disorders such as periodic
• Optimal stimulation occurs with cath- paralysis.
ode and anode over and parallel to the
nerve with the cathode closest to the G1-
recording electrode. Duration
• Cathode and anode should be kept as
close together as possible to prevent stim- The duration of the response of the CMAP is a
ulation distally along the nerve or of function of the duration of the action potential
nearby nerves. of individual muscle fibers within the muscle
• Shock artifact is best eliminated by rotat- as well as the synchrony of firing of the mus-
ing the anode around the cathode as much cle fibers contributing to the potential. A loss
as needed while watching the artifact con- of synchrony results in longer duration and
figuration change. lower amplitude. Prolongation of the action
• “Sliding” the stimulating electrode per- potentials of individual muscle fibers, as occurs
pendicular to the nerve localizes a nerve in critical illness myopathy, will also result in
medial to lateral. longer duration CMAPs (Fig. 23–3).
CMAP MEASUREMENTS Latency

The CMAP recorded over any muscle in The latency of the CMAP is best measured
response to nerve stimulation is called the M from the time of stimulation to the onset of
Elbow
7.2 16.4
A1
Wrist
200 μV
A2
2.8 17.2
Knee
A1
11.8 17.9
500 μV
Ankle
A2
4.2 15.1
Figure 23–3. Ulnar (upper two traces) and peroneal (lower two traces) motor nerve conduction studies in severe critical
illness myopathy. A1, stimulation at elbow (ulnar) and knee (peroneal); A2, stimulation at wrist (ulnar) and ankle (peroneal).
Conduction velocities are normal. CMAP durations are markedly prolonged with the typical configuration of critical illness
myopathy without dispersion of the type seen in demyelinating neuropathies.
the initial negativity. The latency defines the represent the time (milliseconds) from the
time it takes the action potential to travel from stimulus to the onset point of the negative wave
the stimulation site to the recording site and of the CMAP. The reproducibility of latency
depends mainly on the conduction time in the measurements can be enhanced by automated
peripheral axons. A small amount of time is measurement at a fixed voltage above base-
needed to traverse the neuromuscular junc- line (200 μV/cm is often recommended).11
tion. If the electrodes are not over the end Attempts have been made to correct for slow-
plates, latency also includes the time for con- ing in the nerve terminal and at the neuromus-
duction along the muscle fiber to the recording cular junction, a measurement called residual
electrode. In this case, the CMAP initially is latency. This method has been reported to be
positive rather than negative, with the elapsed of value in diagnosing early carpal tunnel syn-
time to reach the end plate being the latency drome (CTS). Residual latency is calculated
of the initial positive deflection (Figs. 23–1 and with the formula:
23–2). Initial positive deflections may also be
caused by the recording of a CMAP of a dis- RL = DML − (distal distance /
tant muscle, for example, a contribution from conduction velocity),
the anterior compartment muscles with stimu-
lation of the peroneal nerve at the knee when where RL is residual latency and DML is distal
recording from the extensor digitorum brevis motor latency.
(EDB). This initial positivity should not be
measured.
Distal latency is the onset of a CMAP Conduction Velocity
at the most distal site of stimulation and is
best measured as an absolute value. Distal The difference in CMAP latency with stimula-
latency measurements should be made from tion at two points along a nerve is a function
the CMAP at the distal stimulation site and of the distance between the two points and
8.5 ms
5 mV
2 ms
2.7 ms
335 mm
CV (m/second) = conduction distance/conduction time (CT)

CT = proximal latency–distal latency
CT = 8.5–2.7 = 5.8 ms
CV = 335 mm/5.8 ms = 58 mm/ms = 58 m/second
Figure 23–4. Calculation of conduction velocity from latency and distance measurements on standard motor NCS. The
upper waveform is the response from stimulation at the elbow and the lower waveform from stimulation at the wrist.
The calculation for the CV is demonstrated in the box.
the rate of conduction of the action potentials is the latency of the proximal stimulation site
in that nerve between the two points. Divid- minus the latency of the distal stimulation site
ing the distance between the two points by (in milliseconds) (Fig. 23–4).
the difference in CMAP latencies measures the
conduction velocity (CV) of the nerve fibers.
Because the latency measurements are made Potential Errors in CMAP
to the initial negativity, the conduction velocity Measurements
measurement is that of the fastest conduct-
ing fibers. Paired stimulation techniques, in Several potential sources of error must be kept
which the action potentials in the fast con- in mind in measuring CMAPs during NCS.
ducting fibers are obliterated by collision, have The most common one is incorrect measure-
been used to measure conduction velocity in ment of the distance between the two points
slower conducting axons. However, the addi- of stimulation, which may be caused by (1)
tional clinical data provided by paired stimu- distortion of the skin when applying the stim-
lation are not sufficiently useful clinically to ulating electrodes or when making the mea-
make it a standard procedure.12 surement, (2) nonstandard position of the body
The conduction velocity is measured using during the measurement, such as having the
the following formula: elbow extended rather than flexed during ulnar
NCS, (3) erroneous polarity of the stimulating
CV (meters/second) = conduction distance electrode, and (4) simultaneous stimulation of
(millimeters) / conduction time (milliseconds), adjacent nerves (Fig. 23–2). Sources of error
in latency measurements include (1) failure to
where the conduction distance is the distance note the sweep speed correctly, (2) a poorly
between the two stimulation sites along the defined shock artifact that interferes with the
nerve (in millimeters) and the conduction time take-off of the CMAP, (3) incorrect electrode
Action potentials from hypothenar muscles

following stimulation of right ulnar nerve at elbow
0 1 2 3 4 5
A
10 mv.
–
C
+
Figure 23–5. The size and configuration of CMAPs evoked by ulnar nerve stimulation vary with location of the
hypothenar recording electrodes. Top, Location of the active recording electrode with the reference electrode on the fifth
digit. Bottom, The corresponding CMAPs. (From Carpendale, M. T. F. 1956. Conduction Time in the Terminal Portion of
the Motor Fibers of the Ulnar, Median and Peroneal Nerves in Healthy Subjects and in Patients With Neuropathy. Thesis,
Mayo Graduate School of Medicine [University of Minnesota], Rochester. By permission of Mayo Foundation.)
location, resulting in an initial positivity or a Normal Values in CMAP

poorly defined onset of the negative CMAP Recordings
(Fig. 23–5A and 23–5B), and (4) failure to
select the same point on the inflection of the For clinical reports, values in CMAP record-
CMAP for the measurement at two points of ings should include the actual value measured
stimulation. When a CMAP is recorded at two and the appropriate associated normal values.
sites of stimulation, it should be very similar at Ideally, normal values are those collected in
both sites unless disease or anomalous inner- the same laboratory using the same techniques,
vation is present. If the two responses are not including careful attention to sources of error.13
similar, technical or physiologic errors must be This is not always possible, and values collected
excluded before the difference is attributed in large studies can serve if the techniques are
to localized disease. Technical errors can be similar to those of the user.14, 15 First, normal
caused by submaximal stimulation at one loca- values are corrected as needed for the physio-
tion or excessive stimulation with activation of logic variables described below. The resulting
an adjacent nerve in another location. Also, normal values have been presented in various
excessive stimulation at one site may shorten ways, but often without a full understanding of
the latency because of current spread along the their complexity and possible errors resulting
nerve. from their skewed distribution.16, 17 Because
no single value can truly identify a conduc-

tion value as normal or abnormal, it is best
to use normal deviates or percentile values.
Normal deviates define the value’s extent of
deviation from normal, and percentiles define
the proportion of a normal population that
has this value.18 Composite NCS scores appear
to be more reproducible and, in some neu-
ropathies, more sensitive and better correlated A B
5 500
with neurologic impairment than individual mV μV
20 ms 20 ms
NCS attributes.19
Figure 23–6. CMAP recorded from abductor hallucis
Key Points muscle with tibial nerve stimulation at, A, low and, B, high
amplification. F waves are depicted much better for mea-
• CMAP amplitude depends on the number surement at higher amplification. (From Daube, J. R. 1992.
Nerve conduction studies. In Electrodiagnosis in Clinical
and synchrony of firing of axons. Neurology, ed. M. J. Aminoff, 3rd ed., 293. New York:
• CMAP area depends on the number of Churchill Livingstone. By permission of WB Saunders
active axons. Company.)
• Distal latency is the time from the stimu-
lus artifact to the onset of the CMAP.
• Measurement errors and low tempera-
tures are the most common causes of the directly evoked CMAP (Fig. 23–6). Addi-
unexpected conduction velocity abnor- tionally, which motor units are activated varies
malities. from stimulus to stimulus, so each F wave may
have a different morphology (representing the
summated response of the myofibers activated
by whichever motor unit(s) was activated). F-
F WAVES wave latencies vary with each stimulus, because
axons with different conduction velocities are
F waves are small CMAPs recorded from the activated from stimulus to stimulus. As the
muscle fibers of a single or small number of site of stimulation is moved proximally on a
motor units which are activated by antidromic limb, F-wave latency decreases (because the
action potentials that travel centrally along distance the action potential travels decreases)
motor axons to anterior horn cells. Thus, the and M-wave latency increases until the two
criteria for identifying F waves are responses potentials merge, usually with stimulation at
that are variable in latency, amplitude, and con- the elbow or just proximal to it (Fig. 23–7).
figuration, but occur grouped within a consis- Hence, F waves are routinely elicited at the
tent range of latencies. However, in disorders distal stimulation site so that they are clearly
in which there is a loss of significant num- distinguishable from the M wave. The decrease
bers of motor units and only a small number in latency of F waves with more proximal stim-
of motor units remain, elicited F waves may ulation is an important test to ensure that the
have the same morphology and be mistaken responses are in fact late responses, if there
as A waves. Consequently, the latency of an F is a question. The F-wave latency varies with
wave includes the time required for the action the distance from the spinal cord to the site
potential to travel antidromically from the site of stimulation, with the distance to the muscle,
of stimulation to the spinal cord and the time to and with the conduction velocity of the motor
travel orthodromically from the spinal cord to fibers21 (Fig. 23–8).
the muscle. Because F waves travel over long
segments of nerve, they are among the most
sensitive measures of diffuse nerve disease, as
well as a measure of proximal conduction.20 Methods
In most muscles, only a small proportion of
the motor units are activated antidromically Recording. F waves can be elicited by stimulat-
by any one supramaximal stimulus. Therefore, ing any nerve, but they are more prominent in
F waves are much lower in amplitude than some nerves, for example, in the tibial nerve
Figure 23–7. CMAP (M waves) (thin arrows) and F waves (thick arrows) recorded from hypothenar muscles with ulnar
nerve stimulation at elbow and wrist. With proximal stimulation (at the elbow), the M-wave latency increases, whereas the
F-wave latency decreases.
Figure 23–8. Variation of F-wave latency with distance in normal subjects for, A, arm conduction studies and, B, leg
conduction studies.
when recording from foot muscles (hence, F for the routine motor NCS and then imme-
for foot). Recording electrodes for F waves diately rotating the anode perpendicular to
are placed over the muscle in the standard the nerve so that there is no anodal block of
locations used for motor NCS. the antidromic volley necessary to elicit the F
waves. F waves should be recorded with only
Stimulation. Stimulation applied to the median, supramaximal stimulation; otherwise, they may
ulnar, tibial, or peroneal nerve at the wrist or be confused with H reflexes. Higher amplifi-
ankle evokes an F wave that is separated clearly cation is needed than for standard NCS; gains
from the M wave. The cathode should be prox- of 200 or 500 mV/cm are usually adequate.
imal to the anode, and the stimulus should be The longer latencies of F waves require slower
supramaximal to ensure antidromic activation sweep speeds than needed for standard NCS.
of all the axons.22 This is most easily accom- The rate of stimulation does not affect the F
plished after obtaining the supramaximal waves, but minimal muscle contraction may
CMAP response at the distal stimulation site enhance them. However, such contraction can
Figure 23–9. F-wave recordings made, A, with the muscle at rest and, B, with muscle contraction. Reliable measurements
are not possible with poor relaxation.
make it more difficult to recognize F waves velocity (Fig. 23–10). The most convenient and
(Fig. 23–9). Jendrassik or other distracting readily applied method is to compare F-wave
maneuvers such as contraction and relaxation latency with normal values corrected for dis-
of muscles in another limb or in the jaw may tance. However, because F-wave latency varies
enhance F waves without obscuring them. with distance, the absolute latencies depend
A series of stimuli is applied until a min- on limb length. Measurements of limb length
imum of eight to ten F waves have been should be made as described for each nerve
obtained.23 Too few F waves will result in an whenever F waves are recorded.
inadequate sample for reliable measurement Another method is to compare the actual
of the variables. In some nerves, particularly F-wave latencies with an estimated F-wave
the peroneal, F waves may be too infrequent latency, F estimate (Fest ), based on the distance
for an adequate number to be obtained for and conduction velocity in the distal segment
reliable measurements. Therefore, for the per- using the following formula (Fig. 23–10D):
oneal nerve, no elicitable F waves may be a
normal variant. Fest = [(2 × distance)/conduction velocity]
Measurements. The F-wave latency is mea- + distal latency
sured to the earliest reproducible potential in
the series recorded. The latency of each of the F-wave latencies should be within the normal
F waves can be measured and the values plot- range of F-wave estimates. If they are shorter,
ted as a histogram that gives the dispersion proximal conduction is faster than distal con-
(chrono-dispersion) of the F latencies, but this duction. If they are longer, proximal conduc-
is time-consuming and adds little additional tion is slower than distal conduction. Thus, F
value clinically24 (Fig. 23–10). Different labo- waves can distinguish diffuse peripheral nerve
ratories use different distance measurements. disorders from those that are primarily distal
Normal values must be recorded using the and those that are primarily proximal. F-wave
same techniques. In the Mayo EMG labora- latencies obtained in normal subjects 18–88
tory, we have found that arm measurements years old are listed in Table 23–2.
made from the site of stimulation at the wrist In measuring F-wave latencies, it is partic-
(cathode) to the sternoclavicular joint and leg ularly important to pay attention to potential
measurements from the cathode to the xiphoid errors.25 A poorly relaxed muscle may pro-
process are most useful. duce deflections throughout the sweep, mak-
Several methods have been suggested for ing it difficult to identify F waves (Fig. 23–9).
assessing F waves, including comparing the Late components or satellite potentials of a
latency with normal values corrected for age dispersed compound action potential may be
and distance, calculating the conduction veloc- identified incorrectly as F waves. Satellite
ity in the central segment, and calculating potentials can be recognized by their constant
a central latency and comparing it with an location and configuration, in contrast to the
estimated latency based on known conduction variable F waves. Also, the latency of satellite
Figure 23–10. Calculated values for ulnar F-wave latency based on recordings from 96 normal subjects. A, All values are
derived from the F-wave latencies, by the calculations shown with each histogram. B, Central latency estimates the time
from elbow stimulation to return of the F-wave response to the elbow location. C, Conduction velocity is the velocity of
the F waves over the length of the nerve from the wrist to the spinal cord. D, Estimated latency for the F wave is based
on peripheral conduction and the distance from the wrist to the sternal notch. E, Latency difference compares estimated
latency with measured F-wave latency. Proximal slowing alone results in positive differences; distal slowing alone results in
negative differences.
Table 23–2 F-Wave Latency in Normal Subjects 18–88 Years Old

Mean Range Distance Contralateral
(ms) (ms) (cm) Difference
Ulnar/ADM 26.6 21–32 50–76 0–3

Median/APB 26.4 22–31 57–73 0–3
Tibial/AHB 48.6 41–57 106–125 0–4
Peroneal/EDB 47.4 38–57 102–128 0–4
ADM, abductor digiti minimi; AHB, abductor hallucis brevis; APB, abductor pollicis brevis; EDB,
extensor digitorum brevis.
potentials increases with more proximal stimu- ◦ Only about 1% of motor axons are
lation, whereas F-wave latencies decrease. activated with a supramaximal sti-
mulus.
Key Points ◦ Different axons are activated with
• F waves are of lower amplitude than the sequential stimuli.
CMAP and vary in latency and morphol- • Both F-wave and A-wave latencies are
ogy because shorter with more proximal stimulation.
◦ F waves vary in configuration with a single stimulus intensity. The two types of
sequential stimuli (unless there are too A waves are axon reflexes and indirect dis-
few present, that the same one recurs). charges.26 Both decrease in latency with more
◦ A waves have a constant configura- proximal stimulation (Fig. 23–11, small arrow).
tion and occur with each stimulus Indirect discharges are the identical backfir-
(unless the stimulus is near threshold ing activation at a proximal location on an axon
for A-wave activation). that can be blocked by paired stimuli, as can
• F-wave recording varies with a number of F waves (Fig. 23–12). Axon reflexes are poten-
factors: tials that invade a proximal branch of an axon
◦ May be absent with a normal peroneal and can become more or less frequent with a
nerve. change in stimulus intensity. The morphology
◦ Distracting maneuvers like the Jendras- of an axon reflex does not change with repeat
sik enhance F-wave occurrence. stimulation. In contrast, F waves are responses
◦ F-wave measurements are not possible
with poor relaxation.
• Comparison of F-wave latency to the F
estimate localizes the slowing in a nerve:
◦ F-wave latency (the same)—conduction
is the same along the length of the
nerve.
◦ F-wave latency (longer)—slowing is
greater in the proximal segment of the
nerve.
◦ F-wave latency (shorter)—slowing is
greater in the distal segment of the
nerve.
AXON REFLEXES (A WAVES)

Other late responses, A waves, can resem-
ble F waves, but they are more persistent at
Figure 23–12. Diagram showing the effects of

antidromic action potentials on two forms of A waves (late
responses). Top, A single axon with proximal hyperex-
citability is reactivated by an antidromic action potential
to give a late response; the reactivated late response
is blocked by a paired stimulus. Bottom, Theoretical
Figure 23–11. Superimposed abductor hallucis CMAPs outcome if the late responses were activated ephaptically.
recorded in response to tibial nerve stimulation. Late None of the patients with these late responses showed
responses are (1) F waves (thick, second arrow) that the pattern of ephaptic activation. e, point of ephaptic
are variable in latency, amplitude, and configuration and activation; O, point of action potential backfiring; S ,
(2) stable axon reflexes (thin, first arrow). Both have single stimulus; S , second of a pair of stimuli. (From
shorter latency with more proximal stimulation. (From Magistris, M. R., and G. Roth. 1992. Motor axon reflex
Daube, J. R. 1992. Nerve conduction studies. In Electro- and indirect double discharge: Ephaptic transmission?
diagnosis in clinical neurology, ed. M. J. Aminoff, 3rd ed., A reappraisal. Electroencephalography and Clinical
294. New York: Churchill Livingstone. By permission of Neurophysiology 85:124–30. By permission of Elsevier
WB Saunders Company.) Scientific Publishers.)
that are variable in latency, amplitude, and patients should always be performed in the
configuration but occur grouped within a con- same temperature range in which the normal
sistent range of latencies. However, in disor- values were determined.
ders in which there is a loss of significant num- Similar to the worsening of conduction and,
bers of motor units and only a small number hence, worsening of neurologic function in
of motor units remain, elicited F waves may central demyelinating disorders, such as mul-
have the same morphology and be mistaken as tiple sclerosis, with increasing temperature
A waves. (Uhthoff’s phenomenon), increasing tempera-
ture has been shown to be important in the
Key Points function of peripheral nerves and the electro-
physiologic assessment of them. This is particu-
• Axon reflexes are potentials that invade larly important in demyelinating neuropathies.
a proximal branch of an axon and can In normal nerves, increasing temperature low-
become more or less frequent with a ers the amplitude, increases conduction veloc-
change in stimulus intensity. ity, and reduces temporal dispersion. Increas-
• The two types of A waves are axon reflexes ing temperature increases how quickly sodium
and indirect discharges. channels open (and close) which shortens the
action potential rise time and thus decreases
the internodal conduction time (increas-
PHYSIOLOGIC VARIABLES ing conduction velocity). This phenomenon
decreases the sodium influx. Coupled with
AFFECTING THE CMAP increased current leak about the areas of
demyelination and hence decreasing available
Temperature sodium current at the next node, in demyeli-
nating states at higher temperatures there may
In normal subjects, CMAPs vary with sev- not be enough sodium current to propagate
eral factors, which need to be controlled. an action potential. This results in conduc-
The temperature of the limb is the most sig- tion block.28 Hence, increasing temperature
nificant factor; a temperature decrease pro- for NCS increases the degree of demonstrable
duces a 2.0 m/second slowing per degree centi- conduction block (while decreasing temporal
grade and increases both amplitude and area.27 dispersion) in demyelinating neuropathies28, 29
Temperature is a greater cause of variation and compressive mononeuropathies (median
in measurements of conduction velocity than neuropathy at the wrist30 and ulnar neuropathy
errors in measurements of latency or distance. at the elbow31 ). Therefore, conduction block
Between 22◦ C and 38◦ C, conduction veloc- may be missed when the limb being examined
ity is related approximately linearly to the is too cool (e.g., this may be problematic and
temperature, increasing about 50% when the confounding when assessing for possible acute
temperature is increased 10◦ C (Q10 = 1.5). inflammatory demyelinating polyradiculoneu-
Thus, a nerve with a conduction velocity of ropathy (AIDP; Guillain–Barré Syndrome) in
60 m/second at 36◦ C conducts at 40 m/second a patient in the intensive care unit). Focal
at 26◦ C (i.e., a decrease of 2 m/second per cooling has been shown to decrease the
degree centigrade). The change per degree demonstrable conduction block and (inter-
centigrade is proportionally less for nerves that estingly from a clinical implication stand-
have a lower conduction velocity. Calculations point) improve clinical function in patients
can be made to correct for a cool limb, but it with peroneal neuropathy at the fibular
is more reliable and effective to warm a cool head.32
limb before doing the nerve conduction study.
Immersing the limb in a water bath at 40◦ C
for 5 minutes is best. The temperature of the Age
arm measured on the surface over the hand
should be at least 32◦ C; the temperature of the Age must also be considered in determin-
leg measured anterior to the lateral malleolus ing the significance of prolonged latencies,
should be at least 30◦ C. More distal sites have slow conduction velocities, and low amplitudes
lower temperatures. CMAP measurements in of compound action potentials. Conduction
velocity slows progressively between 20 and 30 study in which the normal values were deter-
years of age, and by age 80, it is approximately mined, using the same methods and making a
10 m/second slower.33 Conduction velocities percentile comparison.
are slower in children younger than 3 years and The reproducibility of CMAP recordings
in people older than 65 years. CMAP record- must also be considered both in identifying
ings show no significant differences between abnormality and in comparing values over time
men and women. Height and body size, for when a patient’s condition is being monitored.
example, finger circumference for median sen- These range from 5% for F-wave latencies in
sory values, are also contributing factors to the arm to 15% for CMAP amplitudes in the
normal values. Conduction velocity slows with foot.11 Reliable motor NCS require vigilance in
axonal length. The effect is particularly note- recognizing the many pitfalls possible.37
worthy in persons taller than 6 ft in whom
normal values are significantly slower than in Key Points
shorter subjects.34 Ideally, the normal value for • Lower temperature slows conduction
a patient should be adjusted for temperature, velocity and increases CMAP amplitude
age, and height. and area.
The range of normal values is wide, making • Increasing temperature increases the
the measurement of a single value less reliable degree of demonstrable conduction block.
in identifying mild disease. For example, the • Conduction velocities are slower in
range of normal peroneal/EDB CMAP ampli- ◦ Children <3 years old and adults >65
tudes from 2 to 12 mV means that a patient who years old.
has a baseline amplitude of 10 mV may lose ◦ Longer axons and hence in adults >6 ft
80% of the response before the value is outside tall.
the normal range. It is thus critical to compare • There is no significant difference in
to the other side (recognizing some poten- CMAP responses between men and
tial pitfalls—see below) for unilateral processes women.
where there is a high clinical suspicion, but • Side-to-side CMAP amplitude differences
NCS are still in the normal range. This is par- of greater than 50% identify pathology
ticularly true in younger patients. Generally, with “normal” amplitude responses.
a greater than 50% side-to-side difference in
amplitude is considered a significant asymme-
try. The range of normal values for conduction
velocity, latency, and F-wave latency is nar- CMAP CHANGES IN DISEASE
rower and, thus, somewhat better for identify-
ing mild changes in disease. However, in each Pathophysiology
case, percentile or normal deviation measure-
ments are better for detecting mild disease.18 The techniques routinely used to study nerve
Combinations of variables may improve the conduction test large-diameter afferent fibers
recognition scores.35 and alpha motor fibers. The nerve action
There are also significant differences in the potential from a mixed nerve is predominantly
normal values for amplitudes and rates of from large afferent fibers. Components result-
conduction between different nerves, partic- ing from activation of small myelinated (delta)
ularly between the upper and lower extrem- fibers and C fibers cannot be identified. Hence,
ities. In unilateral disorders, comparisons of normal NCS and electromyography do not
values obtained in the affected limb with those exclude a small fiber neuropathy or process.
obtained in the opposite limb can be help- Special techniques of measuring distribution of
ful, but there may be large side-to-side dif- conduction in the activated axons have gener-
ferences in normals (amplitude, 20%–70%; ally not been accepted.38–40
latency, 30%–40%; conduction velocity, 20%– The CMAP recorded from peripheral nerves
30%; F-wave latency, 10%).36 Therefore, the is the action potential that results from acti-
significance of any value is best evaluated by vation of these large myelinated motor fibers.
comparing it with values obtained in the same Changes in the motor fibers, as occurs in cer-
nerve in a limb at the same temperature of sub- tain diseases, may affect the conduction time
jects of the same age who participated in the along the fibers.
Mechanisms of Conduction in in a more profound increase in capac-

Myelinated Fibers itance and decrease of resistance across
the internodal membrane. In large-diameter
Conduction in myelinated fibers is saltatory. fibers, conduction may be blocked. In smaller
The action potential jumps from one node of diameter fibers, conduction may become con-
Ranvier to the next, with the action potential of tinuous, as in unmyelinated fibers, instead of
one node providing the current that excites the saltatory.41
subsequent node. Conduction velocity is deter- A decrease in the diameter of fibers occurs
mined by the time required for one node to with axonal atrophy or compression. This has
excite the next. Thus, if the distance between been best demonstrated in ultrasound42–44 and
two nodes (internodal distance) is 1 mm and MRI studies45 for CTS. It has been consis-
the nodal conduction time is 20 μs, the conduc- tently shown that there are compressive signs
tion velocity is 50 m/second. The time required and flattening at the site of compression dis-
for one node to excite the next node is deter- tally in the carpal tunnel and swelling with
mined by several factors: an increased cross sectional area at the most
proximal aspect of the carpal tunnel. It is the
1. The faster the rate of rise of the action demonstration of the enlarged nerve proximal
potential at node 1, the more rapidly node to the compression that is most reproducible in
2 will be activated. ultrasonography evaluation and hence is what
2. The smaller the amount of current is measured. (These ultrasound findings have
required to neutralize the charge held been demonstrated to be of similar sensitivity
by the membrane capacitance of node 2 (89%) and specificity (98%) to electrophysio-
and to depolarize the nodal membrane logic diagnosis of CTS with emerging evidence
to threshold, the more rapidly an action that it may also correlate with electrophysi-
potential will appear at node 2. Large ologic measures of severity.)42–44, 46 Decreased
myelinated fibers have a lower membrane diameter increases the resistance to flow of
capacitance than unmyelinated fibers and current from node 1 to node 2. Reduction
hence have a faster conduction velocity. in current flow increases the time required
3. The more current that is lost in neutral- to excite node 2. Simultaneous reduction in
izing the charge across the axonal mem- internodal membrane capacitance because of
brane in the internode and by leakage reduced membrane area does not compensate
through the myelin, the longer it will take for the higher resistance.
to activate the next node. Large myeli- With the loss of large-diameter fibers
nated fibers lose less charge and hence because of conduction block or degeneration,
have a faster conduction velocity. conduction measurements reflect conduction
4. The higher the resistance to current flow in smaller diameter, more slowly conducting
in the axoplasm from node 1 to node 2, fibers instead of the slowing of conduction
the longer it will take to activate node 2. in larger fibers. This may account for the
Large myelinated fibers have lower resis- slow conduction observed in segments proxi-
tance than unmyelinated fibers and hence mal to a focal lesion instead of reflecting an
have faster conduction velocity. extension of the lesion proximal to the site of
compression.
With decreased myelin thickness, partic-
ularly during remyelination, the number of
Mechanisms of Slow Conduction myelin lamellae is small in proportion to fiber
in Disease diameter. The capacitance and conductance of
the internodal membrane are high, the loss of
Paranodal demyelination increases the capac- current through the internode is more than
itance of the internodal membrane. More normal, and a longer time is required to excite
current is needed to neutralize the charge the next node. Other possible factors in the
across the internodal membrane and less slowing of conduction are altered characteris-
is available to discharge node 2. Thus, it tics of the nodal membrane, which affect the
takes longer to initiate an action potential generation of the action potential. No such
at node 2. Segmental demyelination results factor has been identified in focal lesions.
Other effects of demyelination should also after axonal disruption, or axonal degeneration,
be kept in mind. Demyelination increases the as in dying-back neuropathies.
refractory period, decreases the ability of the Large-diameter myelinated fibers are the
fiber to conduct impulses at high frequency, nerve fibers that are most sensitive to dam-
and increases the susceptibility to blocking of age by localized pressure. The largest ones are
conduction with increasing temperature (see the afferent fibers that mediate touch pres-
explanation in the section on Physiologic Vari- sure, vibration, and proprioception. In a mixed
ables Affecting the CMAP). nerve, these fibers generally have larger diam-
eters than alpha motor fibers, as evidenced
Key Points by their 10%–15% faster conduction velocity.
In a chronic compression lesion, measurement
• NCS test large-diameter afferent fibers of conduction velocity in the sensory fibers
and alpha motor fibers. often demonstrates an abnormality before it is
• Normal NCS and EMG do not exclude a evident in motor fibers.
small fiber neuropathy.
• Saltatory conduction (“jumping”) of the Conduction block. Conduction block is identi-
action potential along a myelinated nerve fiable most clearly in individual axons at a site
provides faster conduction velocities with- where the action potential cannot be transmit-
out increased fiber diameter. ted to the next segment. No response occurs
• Membrane capacitance (charge storage with stimulation proximal to the block, and a
capacity) is inversely related to nerve fiber full response is seen with stimulation distal to
conduction velocity (e.g., high capacitance the block. Thus, conduction block in a whole
and slow conduction in unmyelinated nerve may be total, in which no axons trans-
fibers). Demyelination increases capaci- mit potentials across the site of damage, or
tance and thereby slows conduction. partial, in which only a proportion transmit
• Axonal internal resistance is inversely potentials across the block (Fig. 23–13A and
related to fiber diameter, resulting in 23–13B). In conduction block associated with
faster conduction in large fibers. Thus, a localized mononeuropathy, the CMAP area
nerve fiber narrowing or compression (or amplitude) obtained with stimulating just
leads to slow conduction. proximal to the site of the block is decreased
• High membrane resistance in myeli- compared with that just distal to the block.
nated fibers reduces current leakage and Conduction block generally means there are
increases conduction velocities. intact axons that are unable to transmit poten-
tials across a local area of damage. However, an
acute injury to a nerve that destroys all axons
will have the appearance of a conduction block
FINDINGS IN PERIPHERAL for a few days until Wallerian degeneration
NERVE DISORDERS occurs.47
Because many other factors may result in
The electrophysiologic findings in peripheral changes that have the appearance of conduc-
nerve disorders are conduction slowing, con- tion block, explicit criteria are required for
duction block, and reduced CMAPs or their identifying conduction block.48, 49 Slowing in
absence. Each may have a focal or a dif- some of the axons with dispersion of the CMAP
fuse distribution. Conduction slowing may be decreases the amplitude, but it increases the
seen as prolonged distal latencies, slow conduc- duration and area. Therefore, amplitude is
tion velocity, or prolonged F-wave latencies. less reliable in recognizing conduction block.
Segmental demyelination and the narrowing Because of normal dispersion over longer
of axons both slow conduction. Conduction segments of nerve, stimulation over short seg-
block can result from a metabolic alteration in ments is more reliable for identifying conduc-
the axonal membrane, such as local anesthetic tion block.50 With routine motor NCS, stimu-
block, or structural changes in the myelin, such lating a distal and a proximal site, the Mayo
as telescoping or segmental demyelination. Clinic EMG lab uses the following thresholds
Reduced or absent responses are the result of to define when the NCS should be extended to
total conduction block, Wallerian degeneration formally evaluate for possible conduction block
50% of distal. Similar consensus criteria for

conduction block have been published.48 Dis-
persion of a CMAP can also be associated with
decreased area because of phase cancellation,
that is, the summation of positive and negative
components of action potentials from different
axons.12 Thus, conduction block is more diffi-
cult to identify with low-amplitude potentials.
Conduction block increases with temperature
in a damaged nerve, providing another impor-
tant reason for monitoring temperature.28
Conduction slowing. In contrast to conduc-

tion block, conduction slowing is seen as pro-
longed latency and slowed conduction veloc-
ity. Although conduction block and slowing
may occur together, they often occur inde-
pendently. Conduction block is more common
in rapidly developing disorders, and conduc-
tion slowing is more common in chronic dis-
orders. Loss of strength is quantified most
accurately by reduction in CMAPs to stimula-
tion at a proximal site. CMAP amplitude and
area changes can help in categorizing nerve
damage into broad groups. For instance, in
traumatic injuries of a nerve, there is usually
conduction block (with an area or amplitude
change) or axonal disruption (with low ampli-
tude at all stimulation points) or some combi-
nation of the two. The clinical deficit caused
by either conduction block or axonal disruption
may have variable duration (Table 23–3). The
Table 23–3 Duration of Deficit

after Peripheral Nerve Injury
Figure 23–13. Example of partial conduction block in
the peroneal nerve. A, Routine peroneal motor NCS
demonstrating a partial conduction block between the fibu- Injury Duration of deficit
lar head and the knee [CMAP amplitudes of 2.1 mV at the
ankle, 1.9 mV at fibular head, and 0.5 mV at the knee]. Conduction block
B, Inching study with stimulation below the fibular head (amplitude change)
(top trace) and making incremental steps proximal along
the peroneal nerve (conduction block is identified between Metabolic Seconds to minutes
sites 1 and 2, at the fibular head). C, Traces shown in Myelin loss Days to weeks
superimposed view. Axonal distortion Weeks to months
Axonal disruption
(fibrillation potentials)
(i.e., these do not define conduction block,
but only cue one that it should be looked for Few axons No deficit
such as stimulating below the elbow or fibular Many axons Weeks to months
head): (1) for median nerve if proximal ampli- All axons Months to years
tude/area is less than 90% of distal; (2) for ulnar From Daube, J. R. 1992. Nerve conduction studies. In
and peroneal nerves if proximal amplitude/area Electrodiagnosis in clinical neurology, ed. M. J. Aminoff,
is less than 80% of distal; and (3) for tibial 3rd ed., 308. New York: Churchill Livingstone. By permis-
nerve if proximal amplitude/area is less than sion of WB Saunders Company.
Table 23–4 Patterns of Abnormality in NCS of Peripheral Neuromuscular

Disorders
Motor nerve studies Sensory nerve studies
Action potential Action potential
Disorder Amplitude Duration Conduction F-wave Amplitude Duration Conduction
velocity latency velocity
Axonal ↓ Normal >70% Mild ↑ ↓↓ Normal >70%

neuropathy
Demyelinating ↓ Proximal ↑ Proximal <50% ↑ ↓ ↑ Proximal <50%
neuropathy
Mononeuropathy ↓ ↑ ↓ ↑ ↓↓ ↑ ↓
Regenerated ↓ ↑ ↓ ↑ ↓ ↓ ↓
nerve
Motor neuron ↓↓ Normal >70% Mild ↑ Normal Normal Normal
disease
Neuromuscular (↓) Normal Normal Normal Normal Normal Normal
transmission
defect
Myopathy (↓) Normal Normal Normal Normal Normal Normal
Symbols: ↑, increase; ↓, decrease; ↓↓, greater decrease; (↓), occasional decrease.
From Daube, J. R. 1992. Nerve conduction studies. In Electrodiagnosis in clinical neurology, ed. M. J. Aminoff, 3rd ed.,
309. New York: Churchill Livingstone. By permission of WB Saunders Company.
results of NCS are a function of the underly- Findings in Focal Lesions

ing pathologic change, not the duration of the
disorder. No single change in NCS is typical A low-amplitude CMAP, slow conduction,
of the clinical phenomenon of neurapraxia, in or conduction block characterizes localized
which there is transient weakness without atro- peripheral nerve damage. The amplitude of
phy. If neurapraxia is caused by a metabolic the CMAP may be low at all sites of stimu-
alteration, it lasts only a few minutes; how- lation if Wallerian degeneration has occurred.
ever, if it is caused by axonal distortion with In conduction block, some axons are unable
telescoping of internodes, it may persist for to transmit action potentials through the
weeks or months.51 Some patterns of abnor- damaged segment but are functioning distal
mality are summarized in Table 23–4. CMAP to it.52 If all axons are blocked, no response
measurements cannot predict how long these is obtained proximal to the site of the lesion.
abnormalities will last. A block in conduction must be distinguished
from slowing in conduction, which may resem-
ble it. In conduction block, there is an abrupt
decrease in CMAP amplitude and area over
Key Points a short segment. Slowing of nerve conduc-
tion in some axons is associated with a gradual
• NCS assess large-diameter myelinated decrease in amplitude as stimulation is moved
fibers which are the most sensitive to dam- proximally because of dispersion of the CMAP.
age by localized pressure. The area of the evoked response remains con-
• The following distal to proximal ampli- stant with conduction slowing unless phase
tude differences require short segment cancellation also occurs.
stimulation for possible conduction block: The CMAP findings in focal nerve dam-
◦ Ulnar nerve >20% age evolve over time with the restoration of
◦ Median nerve >10% function. Local membrane changes, paranodal
◦ Peroneal nerve >20% demyelination, segmental demyelination, and
◦ Tibial nerve >50%. axonal telescoping are all followed by local
Table 23–5 Changes in Nerve Conduction Variables After Focal Nerve Injury∗
Changes
Variable Acute (<7 days) Subacute (weeks) Progressive Residual∗
Compound muscle action Normal Low if severe Low if severe Low if severe
potential
Motor conduction Normal <30% slow if <30% slow if (<30% slow if
velocity severe severe severe)
Motor distal latency Normal <30% long if <30% long if (<30% long if
severe severe severe)
F wave Prolonged or Absent or Absent or (Prolonged or
absent prolonged prolonged absent)
∗ Parentheses indicate changes that occur sometimes, but not always, at that stage.
repair within days to weeks. Wallerian degen- normal segment just distal to the area of abnor-
eration of some axons in a nerve is followed mality.56 The point of stimulation is noted,
by collateral sprouting of surrounding axons, and stimulation is reapplied at 2-cm inter-
resulting in reinnervation of the muscle and vals proximally along the nerve. The responses
the restoration of CMAPs within a few weeks, are superimposed and compared along the
depending on the amount of axonal loss. Axonal length of the nerve. A localized area of abnor-
sprouting within the nerve and the reinnerva- mality is indicated by a greater decrease in
tion of the muscle after the loss of all the axons amplitude or a greater increase in latency
are much slower and less complete than col- between two adjacent points of stimulation
lateral sprouting. The evolution of the nerve than between other sites. The anatomical loca-
conduction changes is outlined in Table 23–5. tion of this point is measured from a fixed
landmark (e.g., the medial epicondyle or fibu-
lar head) (Fig. 23–13). In this way, a con-
EVALUATION OF FOCAL duction block can be localized precisely along
NEUROPATHIES the nerve. Stimulation with near-nerve needle
electrodes can be used for inching in nerves
If a nerve is conducting slowly, it is impor- deep in the tissue, for example, the median
tant to identify whether the abnormality is nerve in the forearm.
localized or diffuse. Latencies and amplitudes In patients with hereditary neuropathy with
obtained with stimulation (or recording) over liability to pressure palsies (HNPP), even slight
short distances provide the best localization of traction or compression of a nerve may cause
focal nerve damage.55 If there is slowing of motor and sensory disturbances. Furthermore,
conduction velocity over any length of nerve the nerves of clinically unaffected relatives
(e.g., the median and ulnar nerves in the fore- also may have EMG and histologic abnormali-
arm or the tibial and peroneal nerves in the ties. These patients typically have evidence of
leg), the severity of slowing must be compared an underlying sensory and distal motor neu-
with that of other nerves in the patient. If ropathy.53 An increased incidence of pressure
the slowing is out of proportion to the slow- palsies has been observed among patients with
ing elsewhere or the decrease in amplitude is diabetes mellitus.54
more than the normal for that nerve, a localized Slowing of conduction in Guillain–Barré
abnormality must be sought. Stimulating proxi- syndrome (AIDP) is often most marked at sites
mally and distally to the suspected area of local commonly affected by pressure lesions (e.g.,
abnormality (e.g., knee or elbow) can iden- the median nerve at the wrist, the ulnar nerve
tify localized lesions. If conduction block or at the elbow, and the peroneal nerve at the
slowing is found between two points of stim- knee). Other conditions, including renal fail-
ulation, the method of short segmental stim- ure, alcoholism, and malnutrition, have been
ulation (inching) should also be used. Inching reported to increase susceptibility to focal
begins with supramaximal stimulation in the compression lesions.
Key Points nerves, often with features of a demyelinating

neuropathy.57 There may also be mononeu-
• Short segmental stimulation (inching) can
ropathies of the median nerve at the wrist,
precisely localize a conduction block along
the ulnar nerve at the elbow, or the per-
the nerve. oneal nerve at the knee, with localized slow-
ing or conduction block superimposed on a
generalized decrease in the CMAP ampli-
Findings in Diffuse Peripheral tude and mild generalized slowing of con-
Nerve Damage and Peripheral duction. F-wave latencies and sural sensory
Neuropathies nerve action potential amplitudes are the
most sensitive motor and sensory nerve con-
NCS can differentiate a disorder of primarily duction variables, respectively, for identifying
axonal loss from a disorder of primarily neuropathy.20, 58 Needle examination usually
demyelination. A disorder associated primar- shows only mild changes distally. However,
ily with axonal destruction, as in axonal some patients with diabetes mellitus have
dystrophies and dying-back neuropathies, is a lumbosacral polyradiculopathy, with diffuse
associated predominantly with low-amplitude fibrillation potentials in lumbar paraspinal
CMAPs at all sites of stimulation, with no more muscles. This pattern may be associated with
than about 30% slowing in conduction veloc- prolongation of F-wave latencies caused by
ity. Segmental demyelination is associated with proximal slowing of conduction.
pronounced slowing of conduction, usually to Although many patients with neuropathy
less than 50% of normal, and with a progressive have mixed findings on NCS, some patients
decrease in the CMAP amplitude at proximal may have a predominantly axonal or segmental
stimulation sites (Fig. 23–14). Slowing of con- demyelinating neuropathy (Table 23–6).
duction also occurs in severe, chronic axonal
disorders with axonal narrowing.
AXONAL NEUROPATHIES
The severity of peripheral nerve damage
can be well defined by NCS, but the patho- Axonal neuropathies primarily affect the axon
logic alteration cannot be predicted for mixed and produce either diffuse degeneration or
patterns or mild changes. Diabetes mellitus, dying-back of the distal portion of the axon.
the most common cause of peripheral neu- Axonal damage is particularly common with
ropathy, has a mixed pattern of abnormal- toxic and metabolic disorders. The major
ities. In this condition, there commonly is change on NCS is a decrease in the ampli-
a mild generalized distal neuropathy caused tude of the CMAP or the compound nerve
by multiple small additive lesions along the action potential (or both). This decrease is
STIMULATED:
In upper arm Just above elbow Just below elbow At wrist
5 mV
–
.001''
Conduction distance to hand, millimeters:
419 316 270 54
Conduction time to hand, milliseconds:
13.2 10.2 8.3 3.2
Conduction velocity upper arm to wrist, 36 meters per second
Figure 23–14. Hypothenar CMAP recorded with ulnar nerve stimulation in a patient with generalized peripheral
neuropathy. Velocity is slow throughout (36 m/second), with gradual dispersion of the CMAP to produce lower amplitudes
with proximal stimulation. Downward arrow indicates point of stimulation. (Courtesy of Dr. E. H. Lambert, Mayo Clinic).
Table 23–6 Patterns of Electromyographic Abnormality in

Conditions Associated with Peripheral Neuropathy
Predominant EMG/NCS Changes of Axonal Degeneration
Diabetes mellitus (some patients)
Guillain–Barré syndrome (some patients)
Toxic—vincristine, acrylamide, others
Alcohol
Uremia
Acute intermittent porphyria
Collagen-vascular diseases
Carcinoma
Amyloidosis
Predominant EMG/NCS Changes of Segmental Demyelination
Diabetes mellitus (some patients)
Guillain–Barré syndrome (some patients)
Dejerine–Sottas disease
Diphtheritic neuropathy
Chronic inflammatory neuropathy
Refsum disease
Leukodystrophies
Neuropathy with monoclonal protein
EMG, electromyographic; NCS, nerve conduction studies.
From Daube, J. R. 1992. Nerve conduction studies. In Electrodiagnosis in clinical neurol-
ogy, ed. M. J. Aminoff, 3rd ed., 312. New York: Churchill Livingstone. By permission of
WB Saunders Company.
proportional to the severity of the disease. axon has degenerated. Therefore, normal con-
Some axonal neuropathies, such as those asso- duction velocities should not be considered
ciated with vitamin B12 deficiency, carcinoma, evidence against the presence of neuropathy.
and Friedreich’s ataxia, chiefly affect sensory Often, the only finding in a case of axonal
fibers; others, such as lead neuropathies, have neuropathy is fibrillation potentials on nee-
a greater effect on motor fibers. Sensory dle examination of distal muscles, especially
axons commonly are involved earlier and more intrinsic foot muscles.
severely than motor axons. Occasionally, sen- If many large axons are lost because of
sory potentials can be low amplitude and asso- axonal neuropathy, conduction velocity may be
ciated with only mild sensory symptoms. In decreased but not to less than 70% of nor-
contrast to the change in amplitude, there usu- mal. Axonal neuropathies typically affect the
ally is little change in latency or conduction longer axons earlier and are first identified in
velocity in axonal neuropathies; conduction in the lower extremities. Tables 23–7A and 23–7B
individual axons is generally normal until the compare the electrophysiologic findings in
Table 23–7A Electrodiagnostic Features of Predominantly Axonal Neuropathies

CMAP amplitude Conduction velocity Distal latency Conduction block or
temporal dispersion
50–100% NL NL/minimally decreased NL No

<50% NL >70% NL <130%NL No
SNAPs—low amplitude with mild slowng in milder cases to absent in severe cases
EMG—clear fibrillations suggesting axonal loss and uncompensated denervation. In length-dependent pro-
cesses, this shows a distal gradation. In polyradiculopathy/neuropathy, fibrillations are seen proximally and
distally.
Table 23–7B Electrodiagnostic Features of Predominantly Demyelinating

Neuropathies
CMAP amplitude Conduction velocity Distal latency Conduction block or F wave or
temporal dispersion blink reflex
50–100% NL <70% NL >130% NL Yes∗ Prolonged

<50% NL <50% NL >130% NL Yes∗
SNAPs—lower amplitude, significant slowing; more likely absent than axonal

EMG—No fibrillations (or minimal if severe) in pure demyelinating process
Reduced recruitment out of proportion to fibrillations and neurogenic changes
In reality as these processes become chronic → mixed demyelinating/axonal
∗ Conduction block and temporal dispersion are typical of acquired demyelinating neuropathies; occur rarely in hereditary
demyelinating neuropathies.
demyelinating and axonal neuropathies based Similar patterns may be seen in the inher-
on the criteria used at the Mayo Clinic. Nerves ited chronic hypertrophic neuropathies, for
that are more susceptible to local trauma example, hereditary demyelinating motor and
because of their superficial location are also sensory neuropathy type I (Charcot–Marie–
more sensitive to axonal damage; sometimes, Tooth (CMT) type I) and Dejerine–Sottas
axonal neuropathies are manifested first as disease (CMT type III), as well as acquired dis-
peroneal neuropathies with low-amplitude or orders such as chronic inflammatory demyeli-
absent responses, while other motor nerves nating polyradiculoneuropathy (CIDP) and
remain intact. Axonal neuropathies may be variations of it (such as distal acquired demyeli-
associated with a change in the refractory nating sensory polyneuropathy (DADS),
period of the nerve and with a relative resis- multifocal motor neuropathy (MMN) with
tance to ischemia. Amyotrophic lateral scle- conduction block, and multifocal acquired
rosis, a disease of anterior horn cells, shows demyelinating sensory and motor neuropa-
changes on motor conduction studies typical thy (MADSAM)). Demyelinating neuropathies
of axonal neuropathy59 (Table 23–8) and is typically are associated with prolonged latencies
notable for large, repeater F waves resulting and a pronounced slowing of conduction, often
from collateral sprouting of axons.60 in the range of 10–20 ms.61 Commonly, the
amplitude is relatively preserved on distal stim-
SEGMENTAL DEMYELINATING ulation but is decreased proximally because of
NEUROPATHIES dispersion of the CMAP on proximal stimu-
lation. In some hereditary disorders, such as
Segmental demyelinating neuropathies are Dejerine–Sottas disease, the velocity may be
usually subacute inflammatory disorders, the only a few meters per second.
prototype being Guillain–Barré syndrome.
Table 23–8 Changes in Nerve Conduction Variables in Amyotrophic Lateral

Sclerosis
Changes
Variable Subacute/active Chronic active Inactive or residual
Compound muscle Normal unless severe Low if severe Low if severe

action potential
Motor conduction Normal <30% slow if severe <30% slow if severe
velocity/F wave
Motor distal latency Normal <30% long if severe <30% long if severe
Repetitive Decrement in some Decrement in some Normal
stimulation
The electrophysiologic findings also vary most common early electrophysiologic abnor-
among the inherited neuropathies, as shown malities are prolongation of the F wave, H
in Table 23–9.62 Acquired demyelinating wave, blink reflex, or distal motor latencies
neuropathies commonly affect sites of nerve and temporal dispersion.65–68 Distal record-
compression early and produce asymmetrical ing with stimulation at proximal sites, such as
neuropathies of the peroneal, ulnar, or median a spinal nerve or brachial plexus, also may
nerves at the knee, elbow, or wrist, respectively. show abnormalities correlating with proximal
The refractory period in demyelinating neu- segmental demyelination and slowing.69 More
ropathies is decreased, often to the extent that commonly, however, Guillain–Barré syndrome
repetitive stimulation at rates as low as 5 Hz is associated with prolonged distal latencies
causes a decrement. The decrement usually of a mild-to-moderate degree, dispersion or
does not appear, however, until rates of 10 or partial conduction block of CMAPs with prox-
20 Hz are used. Maximal voluntary contraction imal stimulation, and symmetrical or asym-
may increase the conduction block and weak- metrical slowing of conduction velocities. A
ness because of the axonal membrane changes waves are particularly prominent in Guillain–
in CIDP.63 Barré syndrome and are often the earliest
sign of the disease on conduction studies.64, 70
Sensory NCS demonstrate a unique pattern
with, often marked, abnormalities of median
Guillain–Barré Syndrome and ulnar antidromic sensory responses, but
Guillain–Barré syndrome (AIDP), predomi- sparing of the sural response. This is a
nantly a multifocal demyelinating disorder, has function of the larger nerve fiber diame-
a spectrum of electrophysiologic changes.64 ter and degree of myelination in the sural
NCS may show no abnormalities or the abnor- nerve (recorded behind the lateral malleolus)
malities may be limited early in the disease than the median or ulnar nerves (recorded
course and the study should be repeated antidromically over the digits) such that the
if there is a high clinical suspicion. The median and ulnar nerves are more prone to
Table 23–9 Electrophysiologic Findings of Inherited

Demyelinating Neuropathies
Inherited disorders with uniform conduction slowing
Charcot–Marie–Tooth 1A
Charcot–Marie–Tooth 1B
Dejerine–Sottas disease
Metachromatic leukodystrophy
Cockayne’s disease
Krabbe’s disease
Inherited disorders with multifocal conduction slowing
Hereditary neuropathy with liability to pressure palsies
Charcot–Marie–Tooth X
Adrenomyeloneuropathy
Pelizaeus–Merzbacher disease
Refsum’s disease
Inherited disorders with incompletely characterized electrophysiology
PM22 point mutations
P0 point mutations
Adult-onset leukodystrophies
Merosin deficiency
EGR 2 mutations
From Lewis, R. A., A. J. Sumner, and M. E. Shy. 2000. Electrophysiological features of
inherited demyelinating neuropathies: A reappraisal in the era of molecular diagnosis.
Muscle & Nerve 23:1472–87. By permission of John Wiley & Sons.
show electrodiagnostic abnormalities earlier in of the F-wave latency than would be expected
the disease process. Facial nerves or other cra- from the degree of conduction velocity slow-
nial nerves may be involved, with abnormalities ing. At the Mayo Clinic, we routinely compare
seen on blink reflex testing or facial nerve stim- F latencies to a calculated F estimate based on
ulation. Patients with mild nerve conduction the distal conduction velocity.
abnormalities and only mild changes on needle
examination have a good prognosis; others with
low-amplitude CMAPs and prominent fibrilla- Chronic Inflammatory
tion potentials have severe axonal destruction Demyelinating Polyradiculopathy
and a poor prognosis for rapid recovery.71 If
the initial diagnostic study is done early in There have been a number of proposed cri-
the disease process, before there has been teria (including electrophysiologic and clini-
adequate time for Wallerian degeneration and cal parameters) for the diagnosis of CIDP.76–78
active signs of denervation on the needle exam- Importantly, these include clinical, labora-
ination, a repeat study several weeks into the tory, pathologic, and electrophysiologic param-
illness may be useful for defining the relative eters. The most widely cited is the Ameri-
degree of axonal loss (CMAP amplitude reduc- can Academy of Neurology (AAN) consensus
tion and fibrillations) and hence prognosis. criteria which required three of four of the
Several electrodiagnostic criteria have been following to meet the electrophysiologic cri-
proposed for AIDP, although no definitive con- teria for CIDP: (1) conduction block (>20%
sensus has been reached. These criteria have between the proximal and the distal sites) or
been compared and it has been suggested that temporal dispersion in one motor nerve, (2)
those originally proposed by Albers, et al.65 abnormal conduction velocities in at least two
may be the most sensitive.72 However, sensi- motor nerves that are <80% of the lower limit
tivities and specificities of the various criteria of normal if the CMAP amplitude is >80% and
may vary depending on the AIDP subtype and <70% if the CMAP amplitude is <80%, (3)
patient population. The Albers criteria include prolonged distal latency in at least two motor
one of the following in two or more nerves: nerves (>125% if amplitude >80% of normal,
conduction velocity <95% if amplitude >50% >150% if amplitude <80% of normal), and
of normal or <85% if amplitude <50% of nor- (4) absent or prolonged F waves in at least
mal, distal latency >110% if the amplitude two motor nerves (>125% of normal if ampli-
is normal or >120% if the amplitude is less tude >80% of normal and >150% if ampli-
than normal, evidence of temporal dispersion tude <80%).76 These criteria were modified by
(proximal > distal duration by 30%), evidence Saperstein et al.77 such that only two of the
of conduction block (proximal to distal ratio four electrodiagnostic criteria were required
of <0.7), or F-wave latency >120% of nor- to make the diagnosis of CIDP and using
mal.65, 72 Studies have shown, however, that in the more stringent conduction block consen-
the majority of AIDP patients with conduc- sus criteria.48 These modified criteria as well as
tion velocity slowing, it is <75% of the lower those proposed by the European Federation of
limit of normal73 and similar findings have Neurological Societies/Peripheral Nerve Soci-
been reported for CIDP74 and paraprotein- ety78 have been shown to be more sensitive,
associated neuropathies.75 As increasing sen- without a significant loss of specificity to the
sitivity often sacrifices some specificity, at original AAN criteria.79 More recent diagnos-
the Mayo Clinic we have used more con- tic criteria for CIDP have tried to take into
servative criteria for demyelinating disorders account its varying clinical patterns (symmet-
(Table 23–7BB). In addition, F waves can rec- ric CIDP, DADS, MADSAM). These crite-
ognize focal slowing proximally in addition ria, particularly those of revised AAN and the
to peripheral slowing from demyelination by European Federation of Neurological Soci-
comparing F-wave latency with F estimate. eties/Peripheral Nerve Society, are improve-
A longer F estimate than F latency signifies ments in our strict research criteria for defining
greater slowing proximal than distal. It is not CIDP. However, it is important to recognize
possible therefore, without comparing the F that a number of patients with CIDP may not
latency to the F estimate, to determine if there meet the strict electrophysiologic criteria, yet
is a greater degree of slowing and prolongation based on clinical, laboratory, and pathologic
data may still meet criteria for possible or prob- ◦ Length-dependent, axonal sensorimo-
able CIDP and deserve a treatment trial in the tor, large fiber, peripheral
appropriate clinical context.77 Electrodiagnos- ◦ Mixed axonal and demyelinating, periph-
tic criteria proposed for the identification of eral
CIDP are not as sensitive as histologic crite- ◦ Polyradiculoneuropathy
ria in identifying patients who may respond to ◦ Lumbosacral radiculoplexus neuropa-
immune suppressive therapy.76, 80 thy
◦ Single or multiple mononeuropathies.
• AIDP (Guillain–Barré Syndrome) has
multiple
Multifocal Motor Neuropathy With ◦ Prolonged F wave and distal latencies
Conduction Block
early
One form of demyelinating neuropathy is ◦ Conduction block, temporal dispersion,
multifocal motor neuropathy with conduction marked velocity slowing
block, which may superficially resemble amy- ◦ Prolonged blink reflexes
otrophic lateral sclerosis. Although it may have ◦ Sural sensory sparing with median and
other features of a generalized demyelinating ulnar loss or slowing.
neuropathy, the classic finding is that of con- • Conduction block and/or temporal disper-
duction block, especially in the median nerve sion and side-to-side conduction asymme-
in the forearm. The conduction block can try favor acquired over inherited demyeli-
increase with activity81 and is hyperexcitable nating neuropathy.
with fasciculation potentials. The conduction • Toxic and metabolic disorders are typically
block may persist for years.82 Consensus crite- axonal neuropathies with greater sensory
ria for the diagnosis of MMN with conduction than motor involvement.
block have been published.83 To summarize, • CV >70% of the lower limit of normal
the diagnosis requires clinical weakness with- is seen in axonal neuropathies, even with
out objective sensory loss or upper motor neu- CMAP amplitude reduction.
ron signs in the distribution of two or more • Comparison of F latency with F estimate
named nerves that is due to conduction block defines relative proximal to distal slowing
outside of common entrapment sites and with even with marked distal slowing.
normal sensory NCS over these same seg- • Clinical, laboratory, and pathologic find-
ments (and throughout the remainder of the ings are important supplements to NCS in
study). the several diagnosis of CIDP.
At times, the pattern of abnormality in
demyelinating neuropathies helps differenti-
ate an acquired process from a hereditary Findings in Specific Focal
one.84 An acquired demyelinating neuropa- Mononeuropathies
thy has scattered areas of slowing, with some
areas being much more abnormal than oth- On NCS, changes in mononeuropathies vary
ers; hereditary disorders generally have a with the rapidity of development, the dura-
symmetrical pattern. Conduction block sug- tion of damage, and the severity of damage as
gests an acquired process. Acquired demyeli- well as with the underlying disorder.85 With a
nating disorders often show more disper- chronic compressive lesion, localized narrow-
sion with proximal stimulation than hereditary ing or paranodal or internodal demyelination
disorders do. This distinction is not always produces localized slowing of conduction. Nar-
reliable because some patients who have a rowing of axons distal to a chronic compression
hereditary demyelinating neuropathy with a slows conduction along the entire length of
low-amplitude CMAP may have pronounced the nerve. Telescoping of axons with intus-
dispersion at proximal sites of stimulation. susception of one internode into another dis-
torts and obliterates the nodes of Ranvier and
Key Points thus blocks conduction. Moderate segmental
demyelination and local metabolic alterations
• Diabetes can produce any of the variety of are often also associated with conduction block.
types of neuropathy: The segment of nerve with disruption of the
axons distal to an acute lesion may continue Key Points

to function normally for as long as 5 days; • Distal CMAP amplitude is decreased
then, as the axons undergo Wallerian degen-
by 7 days after acute axonal disruption
eration, conduction ceases and the amplitude
while it remains normal with conduction
of the evoked response diminishes and finally
block.
disappears. One week after an acute injury, the
• Needle/EMG findings after nerve injury
amplitude of the evoked response can be used
as an approximation of the number of intact, are dependent on the type of injury and
viable axons (Table 23–10). time since injury:
The evolution of electrophysiologic changes ◦ Reduced recruitment occurs immedi-
after peripheral nerve injury is also seen on ately and remains present indefinitely
needle examination and is an aid in char- after any injury.
acterizing mononeuropathies (Table 23–11). ◦ Fibrillation potentials and motor unit
Therefore, an adequate assessment of a periph- potential changes begin after two weeks
eral nerve injury should include both nee- with axonal disruption.
dle examination and NCS. The significance ◦ All findings except reduced recruitment
of changes with time after injury is out- and large motor unit potentials subside
lined in Table 23–12. The sequence of changes over two years after acute nerve injury.
shows that NCS can be important in assessing ◦ Both NCS and EMG are necessary
localized nerve injury within the first few days for full evaluation of an acute nerve
of injury. injury.
Table 23–10 Compound Action Potential Amplitude after

Peripheral Nerve Injury∗
Amplitude
Injury 0–5 Days After 5 Days Recovery
Conduction block
Proximal stimulation Low Low Increases
Distal stimulation Normal Normal Normal
Axonal disruption
Distal stimulation Normal Low Increases
∗ Supramaximal stimulation.
From Daube J. R. 1992. Nerve conduction studies. In Electrodiagnosis in clinical neurology, ed. M.
J. Aminoff, 3rd ed., 314. New York: Churchill Livingstone. By permission of WB Saunders Company.
Table 23–11 Findings on Needle Examination after Peripheral Nerve Injury

0–15 Days After 15 days Recovery
Conduction block
Fibrillation potentials None None None
Motor unit potentials ↓ Recruitment ↓ Recruitment ↓ Recruitment
Axonal disruption
Fibrillation potentials None Present Reduced
Motor unit potentials ↓ Recruitment ↓ Recruitment Nascent
↓, decrease.
From Daube J. R. 1992. Nerve conduction studies. In Electrodiagnosis in clinical neurology, ed. M. J. Aminoff, 3rd ed.,
314. New York: Churchill Livingstone. By permission of WB Saunders Company.
Table 23–12 Electromyographic Interpretations after Peripheral

Nerve Injury
Finding Interpretation
0–5 days
Motor unit potentials present Nerve intact, functioning axons
Fibrillations present Old lesion
Low compound action potential Old lesion
5–15 days
Compound action potential distal only Conduction block
Low compound action potential Amount of axonal disruption
Motor unit potentials present Nerve intact
After 15 days
Compound action potential distal only Conduction block
Fibrillation potentials Amount of axonal disruption
Distribution of damage
Recovery
Increasing compound action potential Block clearing
Increasing number of motor unit potentials Block clearing
Decreasing number of fibrillation potentials Reinnervation
“Nascent” motor unit potentials Reinnervation
From Daube J. R. 1992. Nerve conduction studies. In Electrodiagnosis in clinical neurology, ed.
M. J. Aminoff, 3rd ed., 315. New York: Churchill Livingstone. By permission of WB Saunders
Company.
MEDIAN NEUROPATHIES responses recorded at the thumb over the same

distance (10 cm), median and ulnar antidromic
The most common focal mononeuropathy is sensory responses to the ring finger over the
carpal tunnel syndrome, in which the median same distance (14 cm), and median and ulnar
nerve is compressed in the space formed by the palmar latencies (both over 8 cm).35, 89, 90 The
wrist bones and the carpal ligament. Many dif-
sensitivity of comparing median to ulnar pal-
ferent approaches have been suggested for the
mar (a mixed sensorimotor nerve response) is
electrodiagnosis of this condition.86–88 Early or
approximately 70%–75%, with a specificity of
mild compression of the median nerve in the 97%.35, 91 When there is a high clinical sus-
carpal tunnel may not show any electrophys-
picion, with symptoms preferentially affecting
iologic abnormalities, especially of CMAPs.
the thumb or fourth digit, and routine NCS and
However, more than 90% of symptomatic palmars do not show median slowing across
patients have localized slowing of conduction
the carpal tunnel, it can be useful to compare
in sensory fibers. The sensory latency through median to radial antidromic latencies (record-
the carpal tunnel is the most sensitive sin- ing over the thumb) or ulnar (recording over
gle measurement for identifying the earliest
the fourth digit) latencies over the same dis-
abnormality. This so-called palmar latency may tance.91 If these are nondiagnostic, performing
be compared with normal values but is more
the full CSI will increase the sensitivity to
reliable when compared with the latency in
83% with similar specificity for the diagnosis
ulnar sensory fibers over the same distance
of CTS.35
for a relative prolongation. Improved sensi-
Moderate nerve compression decreases the
tivity, specificity, and reliability in the diagno-
amplitude of the sensory nerve action potential
sis of CTS have been reported with a com-
and prolongs the latency to a greater extent.
bined sensory index (CSI). The CSI adds
Severe median neuropathy at the wrist also
the relative difference between the latencies
increases the distal motor latency to the thenar
of median and radial antidromic sensory
muscles and decreases the thenar CMAP. Very which are not slowed traversing the carpal tun-
severe median neuropathies at the wrist, which nel, but innervate the thenar eminence travel-
have no elicitable routine motor or sensory ing along the ulnar nerve. These crossing fibers
NCS, can be assessed with a comparison of reach the nearby ulnar hand muscles before
median to ulnar nerve distal motor latencies the uncrossed median fibers that traverse the
recording over the lumbricals and second dor- carpal tunnel (Fig. 23–15).
sal interosseous.92 It has only 56% sensitivity Many patients with CTS have bilateral
for CTS in all patients, but it has a high speci- abnormalities on NCS, even though the
ficity (98%) and is generally still elicitable when symptoms may be unilateral. Therefore, the
other routine responses are not.91 Table 23–13 conduction in the opposite extremity should
summarizes the electrophysiologic grading of be measured if a median neuropathy at the
the severity of median neuropathies at the wrist is identified. Several other considerations
wrist. It should be noted that some patients must be kept in mind when testing for CTS.
with electrophysiologically mild changes may A few patients have a normal sensory response
have severe, limiting symptoms and patients and a prolonged distal motor latency. Chronic
with electrophysiologically severe changes may neurogenic atrophy from a proximal lesion,
have little as far as symptoms are concerned. such as damage to a spinal nerve or anterior
The electrophysiologic severity correlates with horn cells, can result in distal motor slowing
the degree of sensory and motor axonal loss. and a normal sensory response. A radial sen-
This becomes an important factor in how soon sory response may be evoked inadvertently by
one should strongly consider surgical treat- high-voltage stimulation of the median nerve
ment options. Mild electrophysiologic abnor- and recorded as an apparent median sensory
malities mean treatment options from wrist potential. Occasionally, patients have sensory
splints, to corticosteroid injections, to surgical branches that innervate one or more fingers,
options can be considered based on the sever- which are anatomically separated from the
ity of clinical symptoms, whereas more severe motor fibers and relatively spared. Also, the
elctrophysiologic findings may push toward severity of compression may not be the same
earlier surgical interventions. Clinical features for all the fascicles of the median nerve, which
cannot predict NCS findings (or severity).93 would result in greater slowing in the axons to
A decrease in the CMAP is often associated some digital nerves than to others. This varia-
with mild slowing of motor conduction veloc- tion in involvement is the likely reason for the
ity in the forearm and fibrillation potentials in added value of the CSI. A median neuropathy
the thenar muscles. CTS of moderate severity may be an early finding in patients with more
is often associated with anomalous innervation diffuse neuropathies. To exclude this possibil-
of the thenar muscles, with the amplitude of ity, it is necessary to assess other nerves.
the response being higher on elbow stimula- Median neuropathies in the forearm are
tion than on wrist stimulation with an initial much less common and only rarely show
positivity before the M wave.94 This positivity abnormality on NCS, other than slightly low-
represents the median to ulnar crossover fibers amplitude sensory or motor responses (or
Table 23–13 Defining the Severity of CTS with NCS

Mild CTS
Prolonged sensory or palmar (mixed) distal latency with or without amplitude reduction
Moderate CTS
Prolonged sensory or mixed palmar distal latency with low amplitude and prolonged median motor
distal latency
Severe CTS
Absent sensory or mixed palmar and prolonged motor latency OR
Prolonged sensory or mixed palmar with low amplitude and prolonged motor with low amplitude
Very severe CTS
Absent sensory or mixed palmar and absent routine median motor response
Figure 23–15. An example of a median neuropathy at the wrist with a Martin-Gruber anastomosis. The response from
median nerve stimulation at the wrist with a prolonged distal latency 6.9 ms (normal <4.5 ms). The response from stim-
ulation at the elbow shows a large positive wave (arrow) preceding the M wave and a higher amplitude than at the wrist.
The preceding positive wave represents median to ulnar crossover fibers innervating the thenar eminence that do not tra-
verse the impinged carpal tunnel and hence arrive at the thenar eminence prior to the median fibers that must traverse the
carpal tunnel.
both).95 Anterior interosseous neuropathy and can identify slowing in the carpal tun-
pronator syndrome are usually manifested nel when standard studies show no
electrophysiologically by fibrillation poten- responses.
tials in the appropriate muscles. Infrequently,
patients have localized slowing of conduction
in the damaged segment of nerve.
ULNAR NEUROPATHIES
Findings in ulnar neuropathy vary with the
severity and location of the lesion.96 In most
Key Points
patients, the abnormality is at the elbow
• CTS is a clinical diagnosis that may have (Fig. 23–16). Various methods have been sug-
no electrophysiologic correlates. gested for electrodiagnostic evaluation of ulnar
• Difference between median and ulnar neuropathy.88 As in CTS, sensory fibers are
palmar distal latencies is the most sen- more likely to be damaged than motor fibers so
sitive single study for identifying median that the sensory nerve action potential is com-
slowing through the carpal tunnel. monly lost early. In some patients, focal slowing
• Other comparisons can supplement iden- can be demonstrated in ulnar sensory fibers
tification of focal slowing: across the elbow. Thus, direct measurement
◦ Median/thumb and radial/thumb anti- of the orthodromic compound nerve action
dromic latency differences. potential may be an efficient and accurate
◦ Median/digit-IV and ulnar/digit-IV anti- method for recognizing mild ulnar neuropa-
dromic latency differences. thy.96 Motor involvement often occurs later
• EMG and NCS abnormalities and may involve different fascicles selectively;
◦ May not correlate with the severity of thus, motor recordings from the first dorsal
clinical symptoms. interosseous as well as the hypothenar muscles
◦ Define the degree of axonal damage or may increase the sensitivity.97 Although there
loss. may be slowing of conduction to the flexor
◦ Help the clinician determine the urgency carpi ulnaris, this muscle usually shows little
or timing of surgery. or no change on NCS and needle examina-
◦ Comparison of median/lumbrical to tion. Measurements are more accurate with the
ulnar/dorsal interosseous II latencies arm extended laterally and the elbow flexed
A
Stimulate ulnar nerve
Record hypothenar muscle
10 mV
5 ms
Figure 23–16. Ulnar neuropathy at the elbow. A, Stimulation sites (from left to right: wrist, below elbow, above elbow, and
upper arm) are demonstrated on the arm schematic. Responses from the normal left upper extremity (L) follow on the next
line and from the affected right ulnar nerve (R) on the next line. A drop in amplitude is seen at the above elbow and upper
arm levels. B, Superimposed responses from inching across the elbow. The sixth response has a marked drop in amplitude
compared with the fifth response and despite the stimulator being moved the same distance between inching stimuli (i.e.,
approximately an inch between stimuli, hence inching) there is an evidence of focal slowing (larger than expected increase
in latency for distance stimulator was moved) from the fifth to sixth stimuli. By looking at the distance of the stimulator
between stimuli 5 and 6 and comparing it to the distance to the medial epicondyle, the focal partial conduction block and
focal slowing can be precisely localized relative to the medial epicondyle for the surgeon. (Courtesy of Dr. E. H. Lambert,
Mayo Clinic.)
360
to 45◦ , because of better access to and mea- Key Points

surement of the ulnar nerve at the elbow.98 • Ulnar neuropathy at the elbow may show
The most common localized finding in ulnar ◦ Localized conduction block
neuropathy of recent onset is conduction block ◦ Localized slowing
at the elbow that can be localized precisely ◦ Both localized conduction block and
by using the inching procedure99 (Fig. 23–16).
slowing.
This conduction block may be associated with • Ulnar/first dorsal interosseous conduc-
local slowing. Occasionally, focal slowing is
tion studies may be abnormal while
seen without conduction block. Chronic ulnar
ulnar/hypothenar are not, if there is fas-
neuropathy usually results in slowing of con-
cicular involvement of that segment of the
duction. If there has been significant axonal
ulnar nerve.
loss with low-amplitude motor responses, the • Acute compression often shows more
associated axonal atrophy results in slowing
focal motor changes than sensory.
in the forearm. Occasionally, a lesion proxi- • Chronic compression results in greater
mal to the elbow requires that stimulation be
sensory NCS changes.
applied in the upper arm as well.100 In ulnar • If standard NCS are normal and there
and median neuropathies, F-wave latency is
is a high clinical suspicion, side-to-side
prolonged proportional to the slowing in the
comparison should be performed.
peripheral segments. Ulnar neuropathies are • Chronic nerve focal compression results
commonly bilateral; if an ulnar neuropathy is
in slowing along the entire length of the
evident on one side, the opposite extremity
nerve distal to the damage.
should also be tested.
Figure 23–17. Right and left peroneal NCS with a compressive peroneal neuropathy with partial conduction block
localizing to the fibular head. Top image, Sites of stimulation and recording. Bottom images, Compound muscle action
potentials. Note the decrease in amplitude (conduction block) with stimulation at the fibula on the right with a normal
amplitude elicited below the level of the fibula. Ant. Tibial M., anterior tibialis muscle; Ext. Dig. B, extensor digitorum
brevis muscle. (Courtesy of Dr. E. H. Lambert, Mayo Clinic.)
PERONEAL NEUROPATHIES fascicular involvement and may require deep

needle electrode stimulation for sciatic NCS.
Neuropathy of the peroneal nerve at the
head of the fibula is another common focal
lesion. Peroneal neuropathy of recent onset
caused by compression is typically associated Key Points
with a conduction block that can be local- • Knee/ankle amplitude difference >20%
ized precisely using the inching procedure to requires stimulation at the fibular head in
identify the area where the evoked response search of focal conduction block.
decreases (Fig. 23–17). Conduction across this • Peroneal/anterior tibial testing is required
segment usually is not slowed, although in if no EDB response is recorded.
lesions of longer standing the slowing becomes • Focal peroneal neuropathy at the knee
prominent.101 without known nerve injury requires ultra-
NCS of the superficial peroneal nerve may sound of the region for ganglion cyst
be of value in differentiating a peroneal neu- invading the peroneal nerve, especially
ropathy from an L5 radiculopathy. If the super- deep branch.
ficial peroneal sensory nerve is normal, it is • Superficial peroneal sensory response
more likely that the damage is at the root level ◦ Most often reduced by postganglionic
rather than the peripheral level.102 However, damage
the superficial peroneal sensory response has ◦ May be spared in common peroneal
been shown to be abnormal in lumbar radicu- neuropathies
lopathies when the dorsal root ganglion (DRG) ◦ May be abnormal in lumbar radicu-
sits more proximally within the neural foramen lopathies.
(estimated to occur in 10%–40% of L5 DRG)
and there is a laterally placed disc compressing
the dorsal root ganglion as well as the root.102
It should also be noted that in almost 50% OTHER MONONEUROPATHIES
of common peroneal neuropathies (including
those with clinical sensory involvement), the Most other neuropathies are traumatic in ori-
superficial peroneal sensory response is spared gin and may be localized by motor conduction
(presumably secondary to fascicular sparing, studies.104 Neuropathies of the radial, tibial, or
anatomical variation where this branch comes phrenic nerves may similarly be localized by
off above the level of compression, or normal NCS but technically are more difficult.105, 106
sensory nerve function below a purely demyeli- Evaluation of most other nerves is not aided by
nating block without axonal loss).103 NCS because these neuropathies do not show
Patients with a moderately severe per- localized slowing. Reports of localized slow-
oneal neuropathy often do not have a CMAP ing of such nerves generally have not taken
response from the EDB. Recordings from the into account the distal slowing that occurs
anterior tibial muscle with stimulation at the with a long-standing proximal lesion.107, 108 In
head of the fibula and the knee may still facial neuropathies such as Bell’s palsy, stim-
demonstrate a block or slowing of conduc- ulation cannot be applied proximal to the site
tion in the nerve. Anomalous innervation of of the lesion. The usual findings in Bell’s palsy
the EDB muscle by a deep accessory branch with neurapraxia are normal amplitudes and
of the superficial peroneal nerve may make latencies; in axonal degeneration, the ampli-
it more difficult to recognize a peroneal neu- tude of the evoked response is decreased in
ropathy. In an apparent peroneal neuropathy proportion to the axonal destruction.109 Blink
without localized slowing of conduction, the reflexes can be used to measure conduction
short head of the biceps femoris muscle should across the involved segment, but they are
be tested on needle examination for fibrilla- commonly absent in Bell’s palsy. Conduction
tion potentials to exclude a sciatic nerve lesion, studies can help differentiate hemifacial spasm
as it is the most proximal peroneal innervated from other facial movements by demonstrating
muscle and the only one above the popliteal ephaptic activation of lower facial muscles dur-
fossa. Sciatic nerve lesions may present with ing periods of spasm, called the lateral spread
only peroneal deficit secondary to preferential response.
Brachial Plexus Lesions of the spinal nerve and nerve root are prox-
imal to the dorsal root ganglion, the sensory
Most brachial plexus lesions are traumatic, and potentials usually are normal, even in the dis-
motor NCS are of limited value. Generally, the tribution of a sensory deficit. This phenomenon
amplitude of the CMAP is reduced. Given pre- is valuable in identifying avulsion of a nerve
dictable sensory pathways, sensory NCS can root in which there is total anesthesia and loss
be extremely useful in brachial plexus lesions of motor function with normal sensory poten-
in localizing to the brachial plexus vs. the root tials. No evidence has been found of double
level (the latter has normal sensory NCS) and crush, a peripheral mononeuropathy related to
to localizing within the brachial plexus.110 In a radiculopathy.114
patients with lower trunk lesions, the ulnar
and medial antebrachial sensory responses are
reduced or absent, and in those with upper
trunk lesions, the median sensory response to
SUMMARY
the thumb and sometimes to the index fin-
CMAPs are among the most helpful recordings
ger, as well as the lateral antebrachial sensory
in the electrophysiologic assessment of periph-
response, is reduced or absent. In patients
eral neuromuscular disease. Compound mus-
with slowly evolving or compressive lesions of
cle action potentials are the recordings made
the plexus, such as tumors, a localized slow-
for all motor conduction studies, both of the
ing of conduction of motor fibers and occa-
directly recorded M wave used for periph-
sionally conduction block may be identified
eral conduction and the F-wave late response
on stimulation at the supraclavicular or nerve
used for testing proximal conduction. Reliable
root level. Neuralgic amyotrophy (Parsonage–
CMAP recordings require the use of stan-
Turner syndrome) has been reported to show
dard stimulating and recording electrode types
proximal conduction block with root stimula-
and locations and standard measurement crite-
tion.111, 112 Thoracic outlet syndrome, which has
ria.115 The sensitivity and specificity of motor
been reported to show abnormalities on NCS, conduction studies depend on comparing the
is usually a vascular syndrome with a change,
results obtained in a patient with the nor-
if any, only in sensory potential amplitudes that mal values obtained by using exactly the same
traverse the lower trunk (ulnar antidromic and methods. The normal values of motor con-
medial antebrachial sensory responses) and lit-
duction studies vary with physiologic factors
tle or no slowing of nerve conduction. such as age and temperature, which must be
controlled and adjusted.
Motor NCS with CMAPs localize focal
lesions in a nerve by identifying either localized
Radiculopathies conduction block or localized slowing of con-
duction. Conduction block is a change in size
Cervical and lumbosacral radiculopathies are of the CMAP when stimulating at two points
not usually associated with changes in motor near each other along the nerve. Both conduc-
NCS; however, if there is sufficient destruction block and slowing of conduction represent
tion of axons and Wallerian degeneration in pathophysiologic changes in the nerve, which
the distribution of the nerve being tested, the can sometimes be predicted by the changes
amplitude of the CMAP may be decreased.113 found on NCS. These changes can be help-
For example, in an L5 radiculopathy with ful in defining prognosis for improvement after
weakness, the response of the EDB muscle nerve damage.
on peroneal nerve stimulation is often of low
amplitude or absent. In the presence of atro-
phy and a low CMAP, there may be mild
slowing of conduction in the motor axons inner- REFERENCES
vating the atrophic muscle. In a few patients
with lumbosacral radiculopathy, measurements 1. Dyck, P. J., J. L. Karnes, P. C. O’Brien, W. J. Litchy,
P. A. Low, and L. J. Melton. 1992. The rochester
of F-wave or H-reflex latencies in mild lesions diabetic neuropathy study: Reassessment of tests and
have been valuable in identifying proximal criteria for diagnosis and staged severity. Neurology
slowing of conduction.113 Because most lesions 42:1164–70.
2. Blok, J. H., A. Ruitenberg, E. M. Maathuis, and G. and other test abnormalities. Muscle & Nerve 24:
H.Visser. 2007. The electrophysiological muscle scan. 307–10.
Muscle & Nerve 36(4):436–46. 19. Dyck, P. J., W. J. Litchy, J. R. Daube, et al. 2003.
3. Barkhaus, P. E., M. I. Periquet, and S. D. Nandedkar. Individual attributes versus composite scores of nerve
2006. Influence of the surface EMG electrode on the conduction abnormality sensitivity, reproducibility,
compound muscle action potential. Electromyogra- and concordance with impairment. Muscle & Nerve
phy and Clinical Neurophysiology 46: 235–9. 27:202–10.
4. Tjon, A. T. A. M., H. H. Lemkes, A. J. van der Kamp- 20. Andersen, H., E. Stalberg, and B. Falck.1997. F-
Huyts, and J. G. van Dijk. 1996. Large electrodes wave latency, the most sensitive nerve conduc-
improve nerve conduction repeatability in controls as tion parameter in patients with diabetes mellitus.
well as in patients with diabetic neuropathy. Muscle & Muscle & Nerve 20:1296–302.
Nerve 19:689–95. 21. Guiloff, R. J., and H. Modarres-Sadeghi. 1991. Pref-
5. van Dijk, J. G., I. van Benton, C. G. Kramer, and D. erential generation of recurrent responses by groups
F. Stegeman. 1999. CMAP amplitude cartography of of motor neurons in man. Conventional and single
muscles innervated by the median, ulnar, peroneal, unit F wave studies. Brain 114:1771–801.
and tibial nerves. Muscle & Nerve 22:378–89. 22. Young, M. S., and W. J. Triggs. 1998. Effect of stim-
6. Kincaid, J. C., A. Brashear, and O. N. Markand. 1993. ulator orientation on F-wave persistence. Muscle &
The influence of the reference electrode on CMAP Nerve 21:1324–6.
configuration. Muscle & Nerve 16: 392–6. 23. Nobrega, J. A., G. M. Manzano, N. F. Novo, and P.
7. Brashear, A., and J. C. Kincaid. 1996. The influence T. Monteagudo. 1999. Sample size and the study of F
of the reference electrode on CMAP configuration: waves. Muscle & Nerve 22:1275–8.
Leg nerve observations and an alternative reference 24. Weber, F. 1998. The diagnostic sensitivity of dif-
site. Muscle & Nerve 19: 63–7. ferent F wave parameters. Journal of Neurology,
8. Phongsamart, G., J. J. Wertsch, M. Ferdjallah, J. C. Neurosurgery and Psychiatry 65:535–40.
King, and D. T. Foster. 2002. Effect of reference 25. Rivner, M. H. 1998. The contemporary role of F-
electrode position on the compound muscle action wave studies. F-wave studies: Limitations. Muscle &
potential (CMAP) onset latency. Muscle & Nerve 25: Nerve 21:1101–4.
816–21. 26. Magistris, M. R., and G. Roth. 1992. Motor axon
9. Nandedkar, S. D., and P. E. Barkhau. 2007. Contribu- reflex and indirect double discharge: Ephaptic trans-
tion of reference electrode to the compound muscle mission? A reappraisal. Electroencephalography and
action potential. Muscle & Nerve 36: 87–92. Clinical Neurophysiology 85:124–30.
10. Jonas, D., C. Bischoff, and B. Conrad. 1999. Influ- 27. Dioszeghy, P., and E. Stalberg. 1992. Changes in
ence of different types of surface electrodes on motor and sensory nerve conduction parameters with
amplitude, area and duration of the compound temperature in normal and diseased nerve. Elec-
muscle action potential. Clinical Neurophysiology troencephalography and Clinical Neurophysiology
110:2171–5. 85:229–35.
11. Falck, B., and E. Stalberg. 1995. Motor nerve con- 28. Franssen, H., G. H. Wieneke, and J. H. J. Wokke.
duction studies: Measurement principles and inter- 1999. The influence of temperature on conduction
pretation of findings. Journal of Clinical Neurophysi- block. Muscle & Nerve 22:166–73.
ology 12:254–79. 29. Chaudhry, V., T. O. Crawford, and S. E. Derossett.
12. Kimura, J. 1998. Kugelberg lecture: Principles 1993. Thermal sensitivity in demyelinating neuropa-
and pitfalls of nerve conduction studies. Elec- thy. Muscle & Nerve 16:301–6.
troencephalography and Clinical Neurophysiology 30. Wang, A. K., E. M. Raynor, A. S. Blum, and
106:470–6. S. B. Rutkove. 1999. Heat sensitivity of sensory
13. Bril, V., R. Ellison, M. Ngo, B. Bergstrom, D. Ray- fibers in carpal tunnel syndrome. Muscle & Nerve 22:
nard, and H. Gin. 1998. Electrophysiological mon- 37–42.
itoring in clinical trials. Roche neuropathy study 31. Rutkove, S. B., M. A. Geffroy, and S. H. Lichten-
group. Muscle & Nerve 21:1368–73. stein. 2001. Heat-sensitive conduction block in ulnar
14. Buschbacher, R. M. 1999. Median nerve motor neuropathy at the elbow. Clinical Neurophysiology
conduction to the abductor pollicis brevis. Ameri- 112:280–5.
can Journal of Physics Medicine and Rehabilitation 32. Rutkove, S. B. 2001. Focal cooling improves neuronal
78(Suppl):S1–31. conduction in peroneal neuropathy at the fibular
15. Daube, J. R. 1996. Compound muscle action poten- neck. Muscle & Nerve 24:1622–6.
tials. In Clinical neurophysiology, ed. J. R. Daube, 33. Carcia, A., J. Calleja, F. M. Antolin, and J. Berciano.
199–234. Philadelphia: FA Davis Company. 2000. Peripheral motor and sensory nerve conduc-
16. Rivner, M. H. 1994. Statistical errors and their tion studies in normal infants and children. Clinical
effect on electrodiagnostic medicine. Muscle & Nerve Neurophysiology 111:513–20.
17:811–14. 34. Stetson, D. S., J. W. Albers, B. A. Silverstein,
17. Wang, S. H., and L. R. Robinson. 1998. Consid- and R. A. Wolfe. 1992. Effects of age, sex, and
erations in reference values for nerve conduction anthropometric factors on nerve conduction mea-
studies. Physical Medicine and Rehabilitation Clinics sures. Muscle & Nerve 15:1095–104.
of North America 9:907–23. 35. Robinson, L. R., P. J. Micklesen, and L. Wang. 1998.
18. Dyck, P. J., P. C. O’Brien, W. J. Litchy, C. M. Harper, Strategies for analyzing nerve conduction data: Supe-
J. R. Daube, and P. J. Dyck. 2001. Use of percentiles riority of a summary index over single tests. Muscle &
and normal deviates to express nerve conduction Nerve 21:1166–71.
36. Bromberg, M. B., and L. Jaros. 1998. Symmetry of neuropathy with liability to pressure palsies. Neurol-
normal motor and sensory nerve conduction mea- ogy 54:40–4.
surements. Muscle & Nerve 21:498–503. 54. Mulder, D. W., E. H. Lambert, J. A. Bastron, and R.
37. Krarup C. 1999. Pitfalls in electrodiagnosis. Journals G. Sprague. 1961. The neuropathies associated with
of Neurology 246:1115–26. diabetes mellitus: A clinical and electromyographic
38. Wells, M. D., and S. N. Gozani. 1999. A method to study of 103 unselected diabetic patients. Neurology
improve the estimation of conduction velocity distri- 11:275–84.
butions over a short segment of nerve. IEEE Trans- 55. van Dijk, J. G., J. Meulstee, M. J. Zwarts, and F.
actions on Biomedical Engineering 46: 1107–20. Spaans. 2001. What is the best way to assess focal
39. Tu, Y., S. Honda, and Y. Tomita. 1999. Estimation slowing of the ulnar nerve? Clinical Neurophysiology
of the conduction velocity distribution of periph- 112:286–93.
eral nerve trunks. Frontiers of Medical Biological 56. Campbell, W. W. 1998. The value of inching tech-
Engineering 9:189–97. niques in the diagnosis of focal nerve lesions. Inch-
40. Schulte-Mattler, W. J., M. Jakob, and S. Zierz. 1999. ing is a useful technique. Muscle & Nerve 21:
Assessment of temporal dispersion in motor nerves 1554–6.
with normal conduction velocity. Clinical Neurophys- 57. Wilson, J. R., J. D. Stittsworth, A. Kadir, and M.
iology 110:740–7. A. Fisher. 1998. Conduction velocity versus ampli-
41. Bostock, H., and T. A. Sears. 1976. Continuous con- tude analysis: Evidence for demyelination in diabetic
duction in demyelinated mammalian nerve fibers. neuropathy. Muscle & Nerve 21:1228–30.
Nature 263:786–7. 58. Shin, J. B., Y. J. Seong, H. J. Lee, S. H. Kim, H.
42. Nakamichi, K., and S. Tachibana. 2002. Ultra- Suk, and Y. J. Lee. 2000. The usefulness of mini-
sonographic measurement of median nerve cross- mal F-wave latency and sural/radial amplitude ratio
sectional area in idiopathic carpal tunnel syndrome: in diabetic polyneuropathy. Yonsei Medical Journal
Diagnostic accuracy. Muscle & Nerve 26: 798–803. 41:393–7.
43. Kele, H., R. Verheggen, H. J. Bittermann, and C. 59. Feinberg, D. M., D. C. Preston, J. M. Shefner, and
D. Reimers. 2003. The potential value of ultrasonog- E. L. Logigian. 1999. Amplitude-dependent slow-
raphy in the evaluation of carpal tunnel syndrome. ing of conduction in amyotrophic lateral sclerosis and
Neurology 61:389–91. polyneuropathy. Muscle & Nerve 22:937–40.
44. Beekman, R., and L. H. Visser. 2003. Sonography in 60. Ibrahim, I. K., and M. A. el-Abd. 1997. Giant
the diagnosis of carpal tunnel syndrome: A critical repeater F-wave in patients with anterior horn cell
review of the literature. Muscle & Nerve 27:26–33. disorders. Role of motor unit size. American Journal
45. Mesgarzadeh, M., C. D. Schneck, A. Bonakdarpour, of Physics Medicine Rehabilitation 76:281–7.
A. Mitra, and D. Conaway. 1989. Carpal tunnel: MR 61. Gutmann, L., A. Fakadej, and J. E. Riggs. 1983.
imaging. Part II. Carpal tunnel syndrome. Radiology Evolution of nerve conduction abnormalities in chil-
171:749–54. dren with dominant hypertrophic neuropathy of
46. Bayrak, I. K., A. O. Bayrak, H. E.Tilki, M. S. Nural, the Charcot–Marie–Tooth type. Muscle & Nerve
and T. Sunter. 2007. Ultrasonography in carpal tun- 6:515–9.
nel syndrome: Comparison with electrophysiological 62. Lewis, R. A., A. J. Sumner, and M. E. Shy. 2000. Elec-
stage and motor unit number estimate. Muscle & trophysiological features of inherited demyelinating
Nerve 35:344–8. neuropathies: A reappraisal in the era of molecular
47. McCluskey, L., D. Feinberg, C. Cantor, and S. Bird. diagnosis. Muscle & Nerve 23:1472–87.
1999. “Pseudo-conduction block” in vasculitic neu- 63. Cappelen-Smith, C., S. Kuwabara, C. S. Lin, I.
ropathy. Muscle & Nerve 22:1361–6. Mogyoros, and D. Burke. 2000. Activity-dependent
48. Olney, R. K. 1999. Consensus criteria for the diag- hyperpolarization and conduction block in chronic
nosis of partial conduction block. Muscle & Nerve inflammatory demyelinating polyneuropathy. Annals
8(Suppl):S225–9. of Neurology 48:826–32.
49. Pfeiffer, G., E. M. Wicklein, and K. Wittig. 64. Roth, G., and M. R. Magistris. 1999. Indirect dis-
2000. Sensitivity and specificity of different con- charges as an early nerve conduction abnormality in
duction block criteria. Clinical Neurophysiology the Guillain–Barré syndrome. European Neurology
111:1388–94. 42:83–9.
50. Schulte-Mattler, W. J., T. Muller, D. Georgiadis, M. 65. Albers, J. W., P. D. Donofrio, and T. K. McGona-
E. Kornhuber, and S. Zierz. 2001. Length depen- gle. 1985. Sequentional electrodiagnostic abnormali-
dence of variables associated with temporal dis- ties in acute inflammatory demyelinating polyradicu-
persion in human motor nerves. Muscle & Nerve loneuropathies. Muscle & Nerve 8: 528–39.
24:527–33. 66. Ropper, A. H., E. F. M. Wijdicks, and B. T. Shahani.
51. Ochoa, J., T. J. Fowler, and R. W. Gilliatt. 1972. 1990. Electrodiagnostic changes in early Guillain-
Anatomical changes in peripheral nerves compressed Barré: A prospective study in 113 patients. Archives
by a pneumatic tourniquet. Journal of Anatomy of Neurology 47: 881–7.
113:433–55. 67. Vucic, S., K. D. Cairns, K. R. Black, P. S. Chong,
52. Brown, W. F., and S. K. Yates. 1982. Percutaneous and D. Cros. 2004. Neurophysiologic findings in early
localization of conduction abnormalities in human acute inflammatory demyelinating polyradiculoneu-
entrapment neuropathies. Canadian Journal of Neu- ropathy. Clinical Neurophysiology 115:2329–35.
rological Science 9:391–400. 68. Cleland, J. C., K. Malik, P. Thaisetthawatkul, D. N.
53. Andersson, P. B., E. Yuen, K. Parko, and Y. T. Herrmann, and E. L. Logigian. 2006. Acute inflam-
So. 2000. Electrodiagnostic features of hereditary matory demyelinating polyneuropathy: Contribution
of a dispersed distal compound muscle action poten- diagnosis of multifocal motor neuropathy. Muscle &
tial to electrodiagnosis. Muscle & Nerve 33:771–7. Nerve 27:117–21.
69. Menkes, D. L., D. C. Hood, R. A. Ballesteros, and 84. Lewis, R. A., and A. J. Sumner. 1982. The elec-
D. A. Williams. 1998. Root stimulation improves trodiagnostic distinctions between chronic familial
the detection of acquired demyelinating polyneu- and acquired demyelinative neuropathies. Neurology
ropathies. Muscle & Nerve 21:298–308. 32:592–6.
70. Rowin, J., and M. N. Meriggioli. 2000. Electrodi- 85. Fowler, T. J., G. Danta, and R. W. Gilliatt. 1972.
agnostic significance of supramaximally stimulated Recovery of nerve conduction after a pneumatic
A-waves. Muscle & Nerve 23:1117–20. tourniquet: Observations on the hind-limb of the
71. Chiò, A., D. Cocito, M. Leone, et al. 2003. Register baboon. Journal of Neurology, Neurosurgery and
for Guillain–Barré Syndrome. Guillain–Barré syn- Psychiatry 35:638–47.
drome: A prospective, population-based incidence 86. Girlanda, P., A. Quartarone, S. Sinicropi, et al. 1998.
and outcome survey. Neurology 60:1146–50. Electrophysiological studies in mild idiopathic carpal
72. Alam, T. A., V. Chaudhry, and D. R. Cornblath. 1998. tunnel syndrome. Electroencephalography and Clin-
Electrophysiological studies in the Guillain–Barré ical Neurophysiology 109:44–9.
syndrome: Distinguishing subtypes by published cri- 87. Stevens, J. C. 1997. AAEM minimonograph #26: The
teria. Muscle & Nerve 21(10):1275–9. electrodiagnosis of carpal tunnel syndrome. Ameri-
73. Albers, J. W., and J. J. Kelly. 1989. Acquired can Association of Electrodiagnostic Medicine. Mus-
inflammatory demyelinating polyneuropathies: Clin- cle & Nerve 20:1477–86.
ical and electrodiagnostic features. Muscle & Nerve 88. American Association of Electrodiagnostic Medicine:
12: 435–51. Practice parameter for electrodiagnostic studies in
74. Barohn, R. J., J. T. Kissel, J. R. Warmolts, and J. R. carpal tunnel syndrome: Summary statement. 1999.
Mendell. 1989. Chronic inflammatory demyelinating Muscle & Nerve 8(Suppl):S141–3.
polyneuropathy. Clinical characteristics, course, and 89. Lew, H. L., L. Wang, and L. R. Robinson. 2000.
recommendation for diagnostic criteria. Archives of Test-retest reliability of combined sensory index:
Neurology 46: 878–84. Implications for diagnosing carpal tunnel syndrome.
75. Kelly, J. J. 1983. The Electrodiagnostic findings Muscle & Nerve 23:1261–4.
in polyneuropathy associated with monoclonal gam- 90. Kaul, M. P., K. J. Pagel, and J. D. Dryden. 2001.
mopathy. Muscle & Nerve 6: 504–9. When to use the combined sensory index. Muscle &
76. Report from an Ad Hoc Subcommittee of the Nerve 24:1078–82.
American Academy of Neurology AIDS Task Force: 91. AAEM. 2002. Practice parameter for electrodiag-
Research criteria for diagnosis of chronic inflam- nostic studies in carpal tunnel syndrome: Summary
matory demyelinating polyneuropathy (CIDP). 1991. statement. Muscle & Nerve 25:918–22.
Neurology 41:617–18. 92. Preston, D. C., and E. L. Logigian. 1992. Lum-
77. Saperstein, D. S., J. S. Katz, A. A. Amato, and R. J. brical and interossei recording with carpal tunnel
Barohn. 2001.Clinical spectrum of chronic acquired syndrome. Muscle & Nerve 15: 1253–7.
demyelinating polyneuropathies. Muscle & Nerve 93. Witt, J. C., J. G. Hentz, and J. C. Stevens. 2004.
24:311–24. Carpal tunnel syndrome with normal nerve conduc-
78. Hughes, R. A., P. Bouche, D. R. Cornblath, et al. tion studies. Muscle & Nerve 29:515–22.
2006. European Federation of Neurological Soci- 94. van Dijk, J. G., and P. A. Bouma. 1997. Recognition
eties/Peripheral Nerve Society guideline on man- of the Martin-Gruber anastomosis. Muscle & Nerve
agement of chronic inflammatory demyelinating 20:887–9.
polyradiculoneuropathy: Report of a joint task force 95. Buchthal, F., A. Rosenfalck, and W. Trojaborg. 1974.
of the European Federation of Neurological Societies Electrophysiological findings in entrapment of the
and the Peripheral Nerve Society. European Journal median nerve at wrist and elbow. Journal of Neurol-
of Neurology 13(4):326–32. ogy, Neurosurgery and Psychiatry 37:340–60.
79. Tackenberg, B., J. D. Lünemann, A. Steinbrecher, et 96. Merlevede, K., P. Theys, and J. van Hees. 2000. Diag-
al. 2007. Classifications and treatment responses in nosis of ulnar neuropathy: A new approach. Muscle &
chronic immune-mediated demyelinating polyneu- Nerve 23:478–81.
ropathy. Neurology 68:1622–9. 97. Kothari, M. J., M. Heistand, and S. B. Rutkove. 1998.
80. Haq, R. U., T. J. Fries, W. W. Pendlebury, M. J. Three ulnar nerve conduction studies in patients with
Kenny, G. J. Badger, and R. Tandan. 2000. Chronic ulnar neuropathy at the elbow. Archives of Physics
inflammatory demyelinating polyradiculoneuropa- Medicine Rehabilitation 79:87–9.
thy: A study of proposed electrodiagnostic and his- 98. Kincaid, J. C., L. H. Phillips, and J. R. Daube. 1986.
tologic criteria. Archives of Neurology 57: 1745–50. The evaluation of suspected ulnar neuropathy at the
81. Kaji, R., H. Bostock, N. Kohara, N. Murase, J. elbow. Normal conduction study values. Archives of
Kimura, and H. Shibasaki. 2000. Activity-dependent Neurology 43:44–7.
conduction block in multifocal motor neuropathy. 99. Herrmann, D. N., D. C. Preston, K. A. McIntosh,
Brain 123:1602–11. and E. L. Logigian. 2001. Localization of ulnar
82. Taylor, B. V., R. A. Wright, C. M. Harper, and P. neuropathy with conduction block across the elbow.
J. Dyck. 2000. Natural history of 46 patients with Muscle & Nerve 24:698–700.
multifocal motor neuropathy with conduction block. 100. Ochiai, N., J. Honmo, A. Tsujino, and Y. Nisiura.
Muscle & Nerve 23:900–8. 2000. Electrodiagnosis in entrapment neuropathy
83. Olney, R. K., R. A. Lewis, T. D. Putnam, and J. V. by the arcade of Struthers. Clinical Orthopaedics
Campellone Jr. 2003. Consensus criteria for the 12:129–35.
101. Singh, N., F. Behse, and F. Buchthal. 1974. Electro- 109. Olsen, P. Z. 1975. Prediction of recovery in
physical study of peroneal palsy. Journal of Neurol- Bell’s palsy. Acta Neurologica Scandinavica
ogy, Neurosurgery and Psychiatry 37:1202–13. 61(Suppl):1–121.
102. Levin, K. H. 1998. L5 radiculopathy with reduced 110. Ferrante, M. A., and A. Wilbourn. 1995. The util-
superficial peroneal sensory responses: Intraspinal ity of various sensory nerve conduction responses in
and extraspinal causes. Muscle & Nerve 21:3–7. assessing brachial plexopathies. Muscle & Nerve 18:
103. Kang, P. B., D. C. Preston, and E. M. Raynor. 2005. 879–89.
Involvement of superficial peroneal sensory nerve in 111. Lo, Y. L., and K. R. Mills. 1999. Motor root con-
common peroneal neuropathy. Muscle & Nerve 31: duction in neuralgic amyotrophy: Evidence of prox-
725–9. imal conduction block. Journal of Neurology, Neuro-
104. Robinson, L. R. 2000. Traumatic injury to peripheral surgery and Psychiatry 66:586–90.
nerves. Muscle & Nerve 23:863–73. 112. Watson, B. V., M. W. Nicolle, and J. D. Brown. 2001.
105. Trojaborg, W. 1970. Rate of recovery in motor and Conduction block in neuralgic amyotrophy. Muscle &
sensory fibres of the radial nerve: Clinical and elec- Nerve 24:559–63.
trophysiological aspects. Journal of Neurology, Neu- 113. Wilbourn, A. J., and M. J. Aminoff. 1998. AAEM
rosurgery and Psychiatry 33:625–38. minimonograph #32: The electrodiagnostic exami-
106. Cruz-Martinez, A., A. Armijo, A. Fermoso, S. nation in patients with radiculopathies. American
Moraleda,I. Mate, and M. Marin. 2000. Phrenic Association of Electrodiagnostic Medicine. Muscle &
nerve conduction study in demyelinating neu- Nerve 21:1612–31.
ropathies and open-heart surgery. Clinical Neuro- 114. Richardson, J. K., and G. M. Forman, B. Riley. 1999.
physiology 111:821–5. An electrophysiological exploration of the double
107. Post, M. 1999. Diagnosis and treatment of supras- crush hypothesis. Muscle & Nerve 22:71–7.
capular nerve entrapment. Clinical Orthopaedics 115. Stalberg, E., B. Falck, A. Gilai, J. Jabre, M. Sonoo,
368:92–100. and K. Todnem. 1999. Standards for quantifica-
108. Uludag, B., C. Ertekin, A. B. Turman, D. Demir, and tion of EMG and neurography. The International
N. Kiylioglu. 2000. Proximal and distal motor nerve Federation of Clinical Neurophysiology. Electroen-
conduction in obturator and femoral nerves. Archives cephalography Clinical Neurophysiology 52 (Suppl):
of Physical Medicine Rehabilitation 81:1166–70. 213–20.
Chapter 24
Assessing The Neuromuscular

Junction With Repetitive
Stimulation Studies
Andrea J. Boon
INTRODUCTION RAPID RATES OF

ANATOMY AND PHYSIOLOGY STIMULATION
OF THE NEUROMUSCULAR SELECTION OF NERVE–MUSCLE
JUNCTION COMBINATIONS
TECHNIQUE CLINICAL CORRELATIONS
CRITERIA OF ABNORMALITY SUMMARY
INTRODUCTION This chapter includes a brief review of the

anatomy and physiology of the neuromuscu-
Repetitive stimulation is a clinical neurophysi- lar junction as it applies to repetitive stimu-
ologic technique designed and used to evaluate lation, a detailed discussion of the technique
the function of the neuromuscular junction. involved, the pitfalls that can occur if not car-
The neuromuscular junction is the anatomical ried out correctly, criteria used to classify the
site where the motor nerve axon connects func- results as normal or abnormal, the patterns of
tionally with a striated (voluntary) muscle fiber. abnormalities that can be seen, and the clini-
The function of the neuromuscular junction is cal correlation of those abnormalities with the
disturbed and the test results are abnormal in various different disorders of neuromuscular
a group of rare diseases that includes myas- transmission.
thenia gravis, Lambert–Eaton myasthenic syn- Techniques that will be reviewed include
drome, botulism, and several rare congenital
myasthenic syndromes. Repetitive stimulation
is important not only in the detection, clarifica- • Slow-frequency repetitive stimulation of
tion, and follow-up of these unusual diseases, proximal and distal muscle/nerve combi-
but also in excluding these disorders in patients nations to evaluate for decrement (reduc-
presenting with symptoms of fatigue, vague tion in amplitude or area with repetitive
weakness, diplopia, ptosis, and malaise, or with stimulation) seen in myasthenia gravis,
objective weakness of uncertain origin. Lambert–Eaton myasthenic syndrome,
369
botulism, and congenital myasthenic syn- understanding the indications for, techniques
dromes. of, and results of repetitive stimulation.1
• The use of brief exercise to evaluate for Each muscle fiber is innervated by the ter-
postexercise facilitation seen in Lambert– minal branch of a motor neuron. The myeli-
Eaton myasthenic syndrome. nated axon of the motor neuron divides into
• The use of more prolonged exercise numerous branches (collaterals), each of which
to evaluate for repair and postexercise loses its myelin sheath near the muscle fiber
exhaustion seen in myasthenia gravis. and joins the muscle fiber midway along its
• The use of rapid rates of stimulation length. As the axonal branch nears the mus-
to bring out postexercise facilitation in cle fiber, it expands into a presynaptic terminal
patients who are too weak to exercise, bouton that lies within a depression in the
or to bring out decrement in some rare muscle cell membrane. The muscle cell mem-
congenital myasthenic syndromes. brane (postsynaptic membrane) beneath the
nerve terminal has a highly specialized struc-
Purpose and Role of Repetitive Stimulation ture, with numerous junctional folds. These
Studies specialized presynaptic neural and postsynap-
• To evaluate the function of the neuromus- tic muscle cell membrane structures consti-
tute the neuromuscular junction, that is, the
cular junction.
• To assess for disorders such as myasthe- synapse between nerve and muscle (Fig. 24–1).
The presynaptic nerve terminal has spe-
nia gravis and Lambert–Eaton myasthenic
cialized anatomical and metabolic features for
syndrome.
• To evaluate for neuromuscular disorders the formation, storage, release, and reuptake
of acetylcholine. Acetylcholine is required for
in patients with symptoms of fatigue or
chemical synaptic transmission and is stored in
weakness.
synaptic vesicles that release their contents into
the synaptic cleft under appropriate conditions
(Fig. 24–1). The amount of acetylcholine con-
ANATOMY AND PHYSIOLOGY tained in a single vesicle is called a quantum.
OF THE NEUROMUSCULAR The postsynaptic membrane contains acetyl-
JUNCTION choline receptor protein molecules concen-
trated on the crest of the junctional folds
Knowledge of the anatomy and function of (Fig. 24–1). When acetylcholine binds to
the neuromuscular junction is important in the postsynaptic acetylcholine receptor protein
Figure 24–1. Functional anatomy of the neuromuscular junction.

Repetitive Stimulation Studies 371
molecules, it causes a change in configuration Thus, if a nerve is stimulated repetitively,

of the receptor, opening a pore or channel in the amount of acetylcholine released with each
the membrane, resulting in sodium influx and stimulus varies depending on the rate of stim-
depolarization of the muscle cell membrane. ulation. At fast rates of repetition, that is, more
Randomly, presynaptic vesicles containing than 10 per second (short interval between suc-
acetylcholine join the presynaptic membrane cessive stimuli), the amount of acetylcholine
and release their quantal contents into the neu- released increases or is potentiated. After a
romuscular junction. The acetylcholine joins series of rapid stimuli (called tetanic stimula-
with the acetylcholine receptor and produces tion), the potentiation of acetylcholine release
a small depolarization of the muscle membrane may persist for 30–60 seconds. With slow rates
in the area around the neuromuscular junction. of repetitive stimulation, that is, less than 5
This local change is called a miniature end plate per second (long interval between stimuli), the
potential (MEPP). amount of acetylcholine released is less with
When an action potential reaches the nerve each of the first four stimuli. This decrease
terminal, voltage-gated channels in the nerve is more pronounced for 2–5 minutes after a
terminal open, allowing influx of calcium. This period of exercise or after repetitive, or tetanic,
triggers the release of a large number of stimulation.
vesicles (quanta) of acetylcholine in a short In normal subjects, the amplitude of the
time. The acetylcholine binds with the post- EPP is so much greater than that required to
synaptic receptors to produce a depolariza- reach threshold (known as the safety factor),
tion that is much larger than an MEPP. This such that small decreases or increases in ampli-
larger depolarization is termed an end plate tude of the EPP with repetitive firing have
potential (EPP). An EPP is large enough to no effect and each nerve action potential will
depolarize the muscle membrane around the produce contraction of the muscle fiber.
end plate sufficiently to generate a muscle However, in disorders in which the size
fiber action potential, which then spreads in of the EPP is decreased to the point where
an all-or-none fashion over the membrane the amplitude falls just above or just below
of the entire muscle fiber, releasing large threshold, minor physiologic fluctuations in
quantities of stored calcium into the sar- the amplitude of the EPP may assume major
coplasm. The calcium causes a change in con- importance. If the EPP is marginally above
figuration of the muscle fiber filaments and threshold, repetitive stimulation at slow rates
leads to excitation–contraction coupling. Thus, results in a lower amplitude EPP that may not
through excitation–contraction coupling, the reach threshold and neuromuscular transmis-
action potential results in contraction of the sion may fail, causing a decrease in the number
muscle fiber. In normal individuals, the EPP is of muscle fibers that contract. If the EPP is
much greater in amplitude than necessary for just below threshold for generating an action
the muscle cell membrane to reach threshold, potential, repetitive stimulation at rapid rates
and each EPP results in contraction of the mus- may result in an increased release of acetyl-
cle fiber. The size of the EPP is determined choline, producing an increased EPP ampli-
by the amount of acetylcholine in each vesicle, tude that may exceed threshold and result
the number of vesicles released, and the num- in facilitation of neuromuscular transmission,
ber and function of the acetylcholine receptors with an increment in the number of muscle
stimulated. fibers responding.
The amount of acetylcholine released at the
neuromuscular junction varies under different Key Points
conditions. The mechanisms involved in the
release of acetylcholine by an action potential • The neuromuscular junction is composed
are such that if another action potential occurs of a synapse where an axonal branch of a
within 200 ms after the first one, the amount of motor neuron joins with a muscle fiber.
acetylcholine released is greater with the sec- • The presynaptic membrane contains synap-
ond action potential (Fig. 24–2), but if a second tic vesicles, each of which contains one
action potential arrives later than 200 ms after quantum of acetylcholine.
the first, less acetylcholine is released with the • The postsynaptic membrane has numer-
second action potential. ous junctional folds with acetylcholine
Slow rates 2–3 Hz Rapid rates 50 Hz
gravis
End plate
potential
Muscle fiber
action potential
Compound muscle
action potential
Figure 24–2. The effects of repetitive stimulation at slow and fast rates on the release of acetylcholine (ACh), end plate
potential, individual muscle fiber action potentials, and compound muscle action potential in normal subjects, patients with
myasthenia gravis, and patients with myasthenic syndrome.
receptors concentrated on the ends of reduction in the number of fibers that

those folds. contract and a decrement in the com-
• When one quantum of acetylcholine binds pound muscle action potential.
to a receptor, a channel opens in the mus-
cle membrane, resulting in a small depo-
larization of the muscle membrane (an
MEPP). TECHNIQUE
• An action potential arriving at the nerve
terminal causes release of many quanta Repetitive stimulation testing demands an
of acetylcholine, resulting in an EPP that unusual degree of experience, attention to
depolarizes the muscle membrane suffi- detail, and technical expertise to avoid mis-
ciently to generate a muscle fiber action leading results. Technique is very important
potential. because, on the one hand, poor technique can
• The muscle fiber action potential spreads result in “abnormal” findings in patients with
in an all-or-none fashion and leads to con- normal neuromuscular transmission, leading to
traction of the muscle fiber. an erroneous diagnosis of a disorder of neuro-
• The amount of acetylcholine released with muscular transmission.2–4 On the other hand,
each action potential depends on the fre- poor technique may result in “normal” find-
quency of nerve stimulation. ings in patients with abnormal neuromuscu-
• In normal subjects, the EPP is much lar transmission, producing a falsely negative
greater than needed to ensure reli- result and a missed diagnosis.
able neuromuscular transmission (this is The basic techniques required are those
known as a high safety factor). used for routine motor nerve conduction stud-
• In diseases where the EPP is just above ies (see Chapter 23). These basic techniques
threshold, there will be less release of must be mastered before repetitive stimulation
acetylcholine with successive stimuli at is attempted. Attention to these basic techni-
low rates of stimulation, resulting in a cal details results in reliable and rapid testing.
Ignoring these technical details can produce

unreliable results, requiring repetition of the
procedure and causing great expenditure of
time and unnecessary patient discomfort.
Recording electrode placement is the same
as for routine motor nerve conduction stud-
ies, with the active, or G1, electrode over the
end plate or motor point and the G2 elec-
trode over the tendon. Repetitive stimulation
requires that the G1 electrode be positioned so
that the initial upward, or negative, deflection
is very sharp and there is no initial downward
(positivity at G1) deflection. The recording
electrodes and wires must be attached securely
so that no movement or loosening occurs dur- Figure 24–3. Example of a simple immobilization device
ing the study. The ground electrode should be for ulnar motor repetitive stimulation studies. Two boards
positioned to minimize stimulus artifact. are connected by adjustable clamps and padded for patient
comfort. The fingers are immobilized with a Velcro strap.
Patient cooperation is critical for a tech- The hand is immobilized in this fashion after the recording
nically satisfactory study. The patient has to electrodes have been attached securely.
understand the purpose and importance of the
study. Patient cooperation results in a shorter,
less painful study and produces more reliable greater the chance of stimulating unwanted
results. Instruct patients about the purpose of nearby nerves. For these reasons, the nerve
the study and the importance of remaining as must be carefully localized. In repetitive stim-
relaxed as possible. The patient should let the ulation, the stimulus site must be stable. This
movement produced by the stimulation occur is a challenge because movements produced
but avoid the contraction of other muscles, by the stimulation and the reaction of patients
especially between stimuli. tend to cause movement of the stimulating
Immobilization is very important in repeti- electrode in relation to the nerve. Stimulation
tive stimulation studies. Movement can result of the nerve distally limits the contraction of
in broken connectors and wires, loose elec- unwanted muscles and results in less artifact.
trodes, alterations in the relationship between Stimulation with a near-nerve needle electrode
the recording electrode and the muscles, intro- results in shorter duration, lower intensity, and
duction of unwanted activity from neighboring more localized stimulation. This may produce
muscles, and shifts in the location of the stim- less stimulus artifact, less patient discomfort,
ulator. Some form of physical restraint should and more reliable results. The needle can be
be used when possible to minimize movement. taped firmly to the skin, to avoid the sliding
Immobilization devices may include clamps, of the stimulator during the train of stimuli or
boards, straps, sheets, and towels (Fig. 24–3). between series of stimuli. This is particularly
Basically, any device that limits shifts between helpful if prolonged testing is necessary, such
the electrodes and recording or stimulating as when evaluating for periodic paralysis, or
sites without harming the patient or recording some congenital myasthenic syndromes.
setup can be used. The stimulus must be strong enough to
The stimulation technique is based on the excite all the motor axons, and the intensity
usual techniques of supramaximal stimulation must be increased 10%–25% above the level
with the smallest stimulus possible in terms of that produced a maximal response. This supra-
duration and intensity. Therefore, the stimu- maximal stimulus cancels the effect of small
lating cathode and anode must be as close to degrees of movement of the stimulating elec-
the nerve as possible. The longer the duration trode away from the nerve. Watch the response
and the greater the intensity of the stimulus, carefully for any change in amplitude or con-
the more uncomfortable the patient will be, figuration. If any such change is noted, the
the more movement that will occur, the more adequacy of the stimulus intensity should be
stimulus artifact that will be created, and the checked immediately.
The stimulus rate and the number of stim- Depending on the clinical problem and the
uli vary depending on the clinical problem. In results of the baseline 2-Hz repetitive stim-
most situations, slow rates of stimulation of ulation, a decision must be made about the
2 Hz, with an interstimulus interval of 500 ms, usefulness of further testing of neuromuscular
will maximize any potential decrement. The transmission with repetitive stimulation after
greatest decrease in acetylcholine release at exercise or tetanic stimulation (Fig. 24–5).
slow rates of stimulation occurs during the first In general, exercise is done for a brief
four stimuli. For these reasons, the best stan- period (10 seconds) or an intermediate period
dard approach is four or five stimuli at 2 Hz. (1 minute).
The slower the rate and the fewer the number Brief (10 seconds) periods of exercise in
of stimuli given, the better the patient is able a cooperative patient have almost the same
to tolerate the procedure. effect as rapid stimulation at 20–50 Hz for
The train of four stimuli should be repeated, 10 seconds, but are not nearly as uncomfort-
with at least 15–30 seconds of rest between able. After 10 seconds of exercise, the release
trains (Fig. 24–4). The trains are repeated of acetylcholine with each action potential is
to check for reproducible amplitudes, areas, potentiated for 30–60 seconds. During this
and configurations as well as the stability of period of postactivation, or posttetanic, poten-
the baseline, the presence of stimulus artifact, tiation, the amplitude of the EPP is increased,
patient relaxation or movement, and the stabil- and the amplitude of the evoked poten-
ity of the recording and stimulating electrodes. tial may be markedly increased in Lambert–
If any abnormalities are found, it is important Eaton myasthenic syndrome or in botulism;
to exclude any potential source of artifact that this phenomenon is known as postactiva-
could result in such an abnormality before pro- tion facilitation. In myasthenia gravis, the
ceeding with further testing. After excluding decrement at baseline may be decreased
any such artifact, three reproducible and tech- or repaired during this period (Figs. 24–4
nically satisfactory sets of four stimuli at 2 Hz and 24–5).
with 15–30 seconds between sets should be If there is no decrement or only a very ques-
obtained as a baseline. tionable decrement on baseline testing and the
B
minutes AE
Figure 24–4. Example of a repetitive stimulation study in a patient with myasthenia gravis, with stimulation of the ulnar
nerve and recording of the CMAP from the ADM muscle of the hand. A train of four stimuli were given at 2/second (s)
with the muscle rested on three occasions, separated by 30 seconds of rest for the three baseline studies (B1, B2, B3).
Next, the muscle was exercised voluntarily for 1 minute (m). The train of four stimuli at 2 Hz was repeated 3 seconds, 30
seconds, 1 minute, 2 minutes, 3 minutes, and 4 minutes after exercise (AE). Top, Histogram of the amplitudes of the four
responses of each train. This histogram is a good example of the pattern of abnormality that can be expected in disorders
of neuromuscular transmission. In each train of four, the greatest decrement is between the first and second response,
with less decrement between the second and third and the third and fourth responses. Immediately after exercise, the
decrement has repaired and there is some postactivation facilitation of the amplitude of the CMAP compared to baseline.
At 4 minutes after exercise (AE 4 m), the decrement is greater than it was at baseline (postexercise exhaustion). Middle,
The four responses to 2-Hz stimulation 4 minutes after exercise are displayed in the x-shifted fashion. Bottom, Numerical
display of the amplitudes and areas of each of the responses (Potential [pot] 1–4) and the percentage decrements (decr) in
amplitude (amp) and area 4 minutes after exercise.
Exercise 3'' later 2' later 10' later

Normal
30
seconds
5 mV
10 ms
Myasthenia gravis
10
seconds 5 mV
10 ms
Myasthenic syndrome
10
seconds 5 mV
10 ms
Figure 24–5. Examples of supramaximal repetitive stimulation at 3 Hz of the ulnar nerve at the wrist while recording over
the hypothenar muscle at rest and at 3 seconds, 2 minutes, and 10 minutes after exercise. Each waveform consists of three
superimposed CMAPs in a normal subject, in a patient with myasthenia gravis, and in another patient with myasthenic
syndrome. In the patient who has myasthenia gravis, the decrease in amplitude from the first to third response repairs
after 10 seconds of exercise and then becomes more pronounced 2 minutes after exercise. In a patient with Lambert–
Eaton myasthenic syndrome, the CMAP at rest is very low in amplitude; there is a decrement that is not appreciable at
this sensitivity. After brief exercise, there is a transient facilitation in amplitude of the CMAP. (From Lambert, E. H.,
E. D. Rooke, L. M. Eaton, and C. H. Hodgson. 1961. Myasthenic syndrome occasionally associated with bronchial
neoplasm: Neurophysiologic studies. In Myasthenia gravis [The Second International Symposium Proceedings], ed.
H. R. Viets, 362–410. Springfield, IL: Charles C. Thomas. By permission of the publisher.)
purpose of the electrophysiologic examination technical factors, including strength of stimu-

is to unmask a borderline or mild defect of neu- lation, should be checked.
romuscular transmission, the patient should be The display of the results varies with
exercised for 1 minute. The 1-minute period the machine, hardware, software, and display
of exercise should be performed in three sets devices. In general, the sensitivity should be
of 20 seconds of exercise interspersed with adjusted to display the potentials as large as
2–5 seconds of rest between each set, to sim- possible without overflowing or blocking. The
ulate prolonged stimulation. For 2–5 minutes sweep speed should be slow enough to spread
after exercise, the amount of acetylcholine the potential out so that it can be analyzed
released with each stimulus should be mini- visually and fast enough such that the entire
mal, providing the greatest chance for detect- potential is displayed, including any late com-
ing any defect of neuromuscular transmission ponents. The repetitive potentials should be
(Fig. 24–5). Usually, after the patient com- displayed in an x-shifted fashion. This means
pletes 1 minute of exercise, four stimuli are that the onset of the sweep for each successive
given at 2 Hz immediately after exercise, and stimulus is shifted to the right on the horizon-
at 30, 60, 120, 180, and 240 seconds after tal, or x-axis, or delayed so that every potential
exercise. As emphasized above, any change in can be analyzed individually. Thus, if there
amplitude, configuration, or area should ini- are changes, it is possible to determine which
tially be considered a technical problem, and potential in the sequence of four changed and
what the order of change was. Superimpo- Increments and decrements in responses are
sition of successive stimuli may allow closer measured by dividing the change in size by the
inspection and detection of changes in ampli- baseline size. For example, a baseline ampli-
tude, area, or configuration but does not allow tude of 10 mV with:
the determination of the sequence of changes
(Fig. 24–6). Results should be either printed with a postexercise amplitude of 8 or 12 mV
immediately or stored for later review and equates to 20% decrement and 20%
printing. increment, respectively,
The measurements of interest for repeti- with postexercise amplitudes of 1 and 19 mV
tive nerve stimulation studies are primarily the equates to 90% decrement and 90%
amplitude, duration, and area of the wave- increment, respectively,
form of the compound muscle action poten- with postexercise amplitudes of 20, 25, and
tial (CMAP) (Fig. 24–4), assuming that the 30 mV equates to 100%, 150%, and 250%
standard nerve conduction studies for that increments, respectively.
nerve/muscle combination have already been
completed. Measurements of amplitude can be
made either with manual or automated mark- Small apparent decrements (<10%) may occur
ers. Measurements of area are quickly and reli- for technical reasons in normal individuals.
ably made on digital electromyography (EMG) Decrements greater than 10% make it more
machines provided that the markers are placed likely that it is not a technical error. Small
at the initial negative deviation from baseline, increments in amplitude (usually <20%) may
and the return of the negative M-wave to base- be seen in normal individuals. Such an incre-
line. Digital EMG machines can quickly mea- ment is usually accompanied by a decrease in
sure changes in amplitude and area between duration with little or no change in area and
the first and subsequent responses. An increase is known as pseudofacilitation. This is due to
in size occurs with facilitation and is mea- an increased synchronization of the firing of
sured as the percent increment (increase) in motor units rather than an increase in the num-
the response; a decrease in size occurs with ber of units or the amplitude of the response of
post activation exhaustion and is measured as individual fibers.
percent decrement (decrease). There are several possible causes for abnor-
mal results in the absence of disease (false
positives) which should always be considered
before making the diagnosis of a disorder of
neuromuscular transmission. If there is move-
ment of the stimulating electrode in relation to
the nerve, this can produce a random variation
in amplitude or, less frequently, a sequential
decrement. This is more common immediately
after exercise and is more likely to occur when
the stimulus is not supramaximal. Submaximal
stimulation is suggested by a loss of amplitude
of the initial response in the train in com-
parison to the amplitude obtained during the
baseline testing (Fig. 24–7).
Movement of the recording electrode rela-
tive to the underlying muscle can produce a
change in configuration and amplitude that is
usually random but occasionally may show a
Figure 24–6. A, Superimposition and, B, staggering of decrement or increment. A shift in the baseline
four CMAPs. The potentials evoked by repetitive stimu- or visible muscle activity between the stimuli
lation can be displayed in a staggered, or x-shifted, fash- is suggestive of such movement. If technical
ion to allow inspection of each potential and determina-
tion of the sequence of any changes. Superimposition of
factors cannot be excluded and are suspected
the potentials allows easier visual identification of small to be the cause of the abnormality, the study
decrements. should be considered technically inadequate
Key Points
• Educate patients as to what to expect, and
how best to relax.
• If there is no risk, ensure that patients
have discontinued anticholinesterases at
least 6 hours before the test, and prefer-
ably overnight.
• Immobilize the limb and attach recording
wires and electrodes securely to prevent
movement.
• Keep the stimulator as still as possible
(consider using near-nerve needle stimu-
Figure 24–7. Technical problems such as poor relaxation lation, with needle taped in place).
or movement during repetitive stimulation can produce • Ensure that the stimulus is supramaximal
apparent decrements in normal subjects, as seen in these
examples.
(20% above that which gives a maximal
response).
• Ensure temperature is adequate; a cold
and nondiagnostic. This is preferable to making limb can lead to false-negative results.
a serious diagnosis on the basis of question- • Begin with a train of four stimuli at a
able data. Repeating the study at a later date slow rate of stimulus (2 Hz), and record
or under other circumstances may be helpful. three separate baseline trains with 15–30
False-negative results or normal results in seconds of rest in between.
a patient with a disorder of neuromuscu- • Consider repeating low-frequency train of
lar transmission can be caused by low tem- four stimuli after brief (10 seconds) exer-
perature, which will mask a mild defect of cise to evaluate for facilitation, especially
neuromuscular transmission, or continuation if CMAP is low at baseline.
of pyridostigmine or neostigmine during the • Consider repeating low-frequency train of
testing. Therefore, it is imperative that the four stimuli after 1 minute of exercise, to
patient discontinue such medications before bring out a mild decrement not seen at
the study and is warm before and during rest or postexercise exhaustion.
the study. Temperatures should be monitored • Technical factors should be suspected if
with the hand skin temperature greater than (1) the results are not reproducible; (2)
32◦ C and the foot skin temperature greater the pattern of decrement, increment, pos-
than 30.5◦ C. If the temperatures are cool, texercise potentiation, or exhaustion is
the patient should be warmed before further unusual; (3) there are baseline shifts or
studies are performed. When possible, treat- changes in configuration; or (4) there is
ment with acetylcholinesterase inhibitors such evidence of muscle activity or movement
as pyridostigmine or neostigmine should be between stimuli.
discontinued for at least 4–6 hours before
the test and preferably overnight. The risk to
the patient from discontinuing treatment must
be weighed against the importance of the test.
False-negative results may also occur if CRITERIA OF ABNORMALITY
treatment with immunosuppressants, intra-
venous immunoglobulin, or plasma exchange In normal subjects, no decrement should
is successful. Ideally, the test should be cond- occur with 2-Hz stimulation; however, techni-
ucted when the patient is most symptomatic— cal problems will often result in small decre-
usually late in the day when fatigued. Patients ments. Therefore, a conservative criterion of
for whom the diagnosis is in question should abnormality is to require a decrement in both
preferably be tested when the effects of area and amplitude of at least 10% in two
treatments such as plasma exchange, intra- different muscle/nerve combinations.
venous gamma globulin, and corticosteroids An abnormal test should meet the following
are minimal. criteria:
• Reproducible results should be obtained Eaton myasthenic syndrome and botulism in

on repeated testing (three baseline trains which there may be a marked increment with
of four stimuli are recommended). rapid rates of stimulation (Fig. 24–8). Rapid
• The pattern of decrement should be con- repetitive stimulation is painful. Brief exercise
sistent with that seen in neuromuscular with voluntary, strong contraction produces
disorders, with the greatest decrement the same effect without pain; however, some
in both amplitude and area occurring patients are unable to produce a strong con-
between the first and second stimuli in the traction because of extreme weakness, lack of
train of four (Fig. 24–4). understanding, or inability to cooperate.
• The changes induced by exercise with If a disorder is suspected such as Lambert–
potentiation and exhaustion should be Eaton myasthenic syndrome, rapid rates of
compatible with what is seen in the dis- stimulation may be necessary. Explain the
ease under question (Fig. 24–5). details of the test to the patient before pro-
• Edrophonium (Tensilon) or neostigmine ceeding. Stress the importance of remaining as
should decrease or correct the abnormali- still and relaxed as possible so that the results
ties seen in myasthenia gravis. of the test will be reliable. Before beginning,
check the machine and electrode setup to pre-
vent any technical errors. If a large facilitation
RAPID RATES OF STIMULATION is expected, the gain or sensitivity should be set
so that a much larger response can be recorded
Rapid rates of stimulation of 10–50 Hz are without blocking. Stimulate at 20–50 Hz for
helpful in some disorders such as Lambert– 2–10 seconds, depending on the situation. An
Figure 24–8. 1–10, Examples of facilitation of CMAPs of the hypothenar muscle during 40–50 Hz repetitive nerve stim-
ulation in 10 patients with Lambert–Eaton myasthenic syndrome. The potentials are shown at slow sweep speeds. The
facilitation ranges from 160%–1400%. Note the variety of patterns of facilitation that occur. In some recordings, a constant
amplitude shock artifact is seen with each CMAP (traces 1, 2, 3, 5, 7, and 9). (Courtesy of Dr. E. H. Lambert, Mayo Clinic.)
increment greater than 40% in adults or 20% transmission (Table 24–1), depending on the
in infants after 3 seconds of stimulation is individual patient’s clinical presentation. Clin-
considered abnormal. ically weak muscles are more likely to show
a decrement on repetitive stimulation, and
therefore those nerve–muscle combinations
should always be checked; however, it is often
SELECTION OF NERVE–MUSCLE more reliable to start with a distal muscle since
COMBINATIONS there is less movement and less pain. It also
allows the patient to get used to the tech-
Most of the diseases that are studied tend to nique before moving on to proximal muscles
affect certain muscles more than others. Prox- of more interest. Furthermore, in Lambert–
imal muscles are usually more involved than Eaton myasthenic syndrome, distal muscles
distal muscles. There are a number of differ- have actually been shown to be more sensi-
ent nerve–muscle combinations that can be tive than the biceps or trapezius in detecting
studied to search for defects of neuromuscular significant postexercise facilitation.5
Table 24–1 Specific Nerve–Muscle Combinations

Nerve Muscle Stimulation Advantages Disadvantages Immobilization
Site
Ulnar Abductor Wrist Reliably Distal muscle may Velcro strap,

digiti minimi immobilized, be spared clamp
well tolerated
Median Abductor Wrist Well tolerated Distal muscle may Thumb
pollicis brevis be spared, difficult restrained with
to immobilize towel
Radial Extensor Spinal groove Reliable, well Distal muscle may Restrain fingers
indicis tolerated, more be spared, may be manually
proprius sensitive than difficult to reliably
ADM stimulate the nerve
Musculocutaneous Biceps Lower edge axilla Proximal muscle Unstable stimulus, Board
difficult to
immobilize, painful
Axillary Deltoid Supraclavicular Proximal muscle Unstable stimulus, Large Velcro
difficult to strap or sheet
immobilize, painful
Spinal accessory Trapezius Posterior border Proximal muscle, Difficult to Strap or hands
upper sternoclei- well tolerated immobilize under chair
domastoid
Facial Nasalis Between mastoid Proximal muscle Painful, unstable None
and tragus stimulation, shock
artifact, cannot
immobilize.
Masseter contraction
Trigeminal Masseter Mandibular Reliable, well Not quite as None (gauze
notch tolerated sensitive as the facial pads between
proximal muscle nerve in myasthenia teeth to prevent
tooth trauma)
Peroneal Tibialis Knee Leg muscle Distal muscle may Board
anterior be spared
Femoral Rectus Femoral triangle Proximal leg Painful, difficult to Restrained
femoris muscle immobilize manually
The most common nerve muscle combi- response, the decrement levels off. After 10
nations studied are the ulnar nerve/abductor seconds of exercise, the decrement is partially
digiti minimi (ADM), spinal accessory nerve/ or completely repaired and there may be a
trapezius, and facial nerve/nasalis. Other com- small increment in the amplitude of the CMAP
binations should be considered based on the compared to baseline. In severe cases in which
reliability and results of the initial testing. the amplitude of the CMAP at baseline is low,
The peroneal/anterior tibial, musculocuta- there may be a marked increment in this ampli-
neous/biceps, and femoral/quadriceps may tude after brief exercise, as is typically seen
show abnormalities not detected with other in Lambert–Eaton myasthenic syndrome.4 At
nerves. A recent study demonstrated the util- 2–4 minutes after 1 minute of exercise, the
ity of the trigeminal nerve/masseter technique, decrement may be larger than at rest (termed
which is reliable, well tolerated, and almost postexercise exhaustion). Similar findings are
as sensitive as the facial nerve/nasalis com- seen in myasthenia gravis associated with the
bination.6 For distal muscles, techniques uti- use of the drug d-penicillamine.
lizing stimulation of the radial nerve with Lambert–Eaton myasthenic syndrome is a
recording over either the extensor indicis rare entity, which is associated with systemic
proprius or the anconeus muscle have been malignancies such as small cell carcinoma of
found to be reliable and more sensitive than the lung in about 40% of cases.11 This syn-
the ulnar nerve/ADM combination.7, 8 Even drome is an autoimmune disorder resulting
unusual combinations such as stimulation of from an antibody that alters the function of
the phrenic nerve while recording from the the voltage-gated calcium channels in the axon
diaphragm are feasible.9 terminal of the neuromuscular junction. The
result is decreased release of acetylcholine with
Key Points each action potential. The amount of acetyl-
choline released increases rapidly with rapid
• Begin with a distal muscle in the most rates of activation or stimulation. The weakness
affected limb (hypothenar and thenar). is more generalized than in myasthenia gravis
• Move to more proximal muscles that are but involves proximal muscles more than distal
clinically involved (trapezius, masseter, muscles and leg muscles more than arm mus-
and nasalis). cles. The bulbar muscles are not involved as
• Consider unusual nerve/muscle combina- prominently as they are in myasthenia gravis.
tions in situations where weakness is very Some patients have autonomic symptoms such
focal. as dry mouth, impotence, and constipation.
Baseline nerve conduction studies usually
demonstrate low-amplitude CMAPs at rest.
CLINICAL CORRELATIONS Repetitive stimulation at 2 Hz produces decre-
ments similar to but often more prominent
Myasthenia gravis is the classic disease of than those seen in myasthenia gravis. However,
the neuromuscular junction.10 It usually is unlike myasthenia where the decrement levels
the result of an autoimmune-mediated attack off after four or five stimuli, in Lambert–Eaton
on the acetylcholine receptor on the mus- myasthenic syndrome, the decrement usually
cle (postsynaptic) cell membrane. This results continues to increase with each stimulus, for up
in fewer functional receptors and fluctuating, to nine stimuli.12 Brief (10 seconds) exercise or
fatiguable weakness involving proximal mus- rapid stimulation at 50 Hz produces a marked
cles more than distal ones, particularly the bul- increment or facilitation of the amplitude of
bar muscles and often the extraocular muscles. the CMAP to more than 100% increase from
Experimentally, the amplitudes of MEPPs and the baseline amplitude (Figs. 24–5 and 24–8).
EPPs are low (Fig. 24–2). The resting CMAP This effect is transient and must be looked
is normal or minimally abnormal except in for in a well-rested, warm muscle immediately
more severe cases. With repetitive stimulation after brief exercise. After 60–120 seconds, the
at 2 Hz, there is a decrement, with the greatest amplitude returns to baseline and the decre-
decrease in amplitude occurring between the ment resumes and may be more prominent 3–4
first and second response and lesser decreases minutes after exercise (postexercise exhaus-
after that (Figs. 24–4 and 24–5). By the fifth tion). If the patient is too weak to exercise,
2–3 seconds of 50 Hz stimulation will induce acetylcholine.20–22 Results of repetitive stimu-

the same effect. LoMonaco et al. have reported lation are abnormal, but the pattern of abnor-
a similar facilitation in very weak patients by mality varies among the different disorders.
combining voluntary exercise with 3-Hz stimu- Special techniques are required in patients
lation applied while exercising.13 Although an with a defect of acetylcholine resynthesis or
increment of 100% or more has been used packaging, formerly named familial infantile
as the gold standard in diagnosing Lambert– myasthenia, as routine repetitive stimulation
Eaton myasthenic syndrome, Oh et al. recently may be normal, but prolonged exercise or
showed that an increment of 60% after exer- prolonged repetitive stimulation at 10 Hz for
cise is highly specific for this disorder, and 5–10 minutes may be required to bring out
increases the sensitivity of the test, particularly the decrement. On routine nerve conduc-
in seronegative cases.14, 15 tion studies in these cases, it is also impor-
Botulism, another rare disorder, is caused by tant to look for repetitive CMAPs, which fol-
exposure to one of the seven types of toxin pro- low the main CMAP by about 3–6 μs, and
duced by Clostridium botulinum. In adults, the may be seen in slow-channel congenital myas-
toxin is ingested in inadequately preserved or thenic syndrome or acetylcholinesterase defi-
prepared food. In infants, botulism is caused ciency.21 Such repetitive CMAPs are easily
by ingestion of spores that germinate into missed on routine nerve conduction studies as
bacteria that produce botulinum toxin in the they decrease in amplitude rapidly and may
gut. From 12 to 36 hours after the toxin is disappear with stimulus rates as low as 0.5 Hz;
ingested, blurred vision, dysarthria, dysphagia, they are best detected after delivery of a single
dry mouth, dyspnea, and generalized weak- supramaximal stimulus in a nerve/muscle com-
ness develop. The toxin markedly decreases the bination with a short duration CMAP such as
number of quanta of acetylcholine released by the ulnar nerve/ADM (Fig. 24–9).
an action potential, thus reducing the ampli- In myotonic disorders, repetitive stimula-
tude of the EPP. Routine nerve conduction tion may produce a small decrement at rest
studies are normal except for low-amplitude that may become more prominent as the
CMAPs. Rapid repetitive stimulation at 50 Hz rate of stimulation is increased. The decre-
or brief exercise produces an increment in ment in myotonic disorders is quite variable,
most cases (in 62% of adults and 92% of in regards to pattern, severity, and the stim-
infants). This may not be found in very severe ulus rate at which it occurs, depending on
cases but, when present, is in the range of the specific genetic mutation involved, tem-
30%–200%. In general, the degree of facili- perature, relationship to exertion, and several
tation seen in botulism is in the 130%–150% other variables.23 The decrement can even vary
range as opposed to Lambert–Eaton myas- from one stimulus to the next within a train
thenic syndrome where facilitation is typically of stimuli. Usually the decrement increases
greater than 200%. A small decrement at slow with increasing rates of stimulus or with pro-
rates of repetitive stimulation is rarely seen in longed exercise. In most myotonic disorders,
adults, but is present in more than 50% of cases there is a drop in amplitude of the CMAP
in infants; more severe cases tend to have less after exercise (and the decrement between
decrement and less facilitation. Postexercise the first and fourth stimuli may repair to
exhaustion is not seen.16–19 some degree for several minutes after exercise)
Congenital myasthenia is the general term (Fig. 24–10); however, in proximal myotonic
applied to a group of rare inherited disor- myopathy (myotonic dystrophy type 2), there
ders of neuromuscular transmission caused by is no loss in amplitude of the CMAP after
various structural or functional alterations of brief exercise, and this finding may help dis-
the neuromuscular junction, such as decreased tinguish the two disorders.24 In paramyotonia,
filling of the synaptic vesicles with acetyl- exercise or repetitive stimulation may fail to
choline, decreased number of synaptic vesi- bring out a decrement of the CMAP amplitude
cles available for release, absence of one form when the muscle is warm; however, a signifi-
of acetylcholinesterase, decreased number of cant decrement may be brought out by cooling
acetylcholine receptors on the junctional folds, the muscle, which may be more pronounced
or kinetic defects in the acetylcholine recep- after exercise or repetitive stimulation during
tor that increase or decrease the response to cooling, and may even progress to complete
Figure 24–9. Repetitive stimulation of the ulnar nerve at 2 Hz, recording from surface electrodes over the ADM muscle
in a patient with slow-channel congenital myasthenic syndrome. Note the repetitive CMAPs that follow the main CMAP
by about 6 ms, but are smaller and decrement to a greater degree than the main CMAP.
After 10 second voluntary contraction

Rest 3'' 30'' 3'
2 ms
Figure 24–10. Example of repetitive stimulation study of the ulnar nerve while recording from hypothenar muscles in
a patient with myotonia. There is a very small reproducible decrement at baseline. The decrement and the amplitude of
the CMAP are decreased after 10 seconds of exercise, but 3 minutes after exercise the amplitude and the decrement have
increased again.
electrical silence.23 Recessive myotonia con-

genita may show decrements of 60%–80% after
a short exercise test, and even more so with
a train of 10-Hz stimulation.23 Changes of this
magnitude are strongly suggestive of recessive
myotonia congenita.
Periodic paralysis may reveal a gradual
reduction in amplitude of the CMAP for 20–30
minutes after exercising the muscle for 3–5
minutes (Fig. 24–11), therefore the usual pro-
tocol is to leave the electrodes in place, and
record a single CMAP every 5 minutes for
30 minutes to demonstrate this loss of ampli-
tude.25 There usually is no decrement at slow
rates of repetitive stimulation.
In progressing neurogenic disorders such
as amyotrophic lateral sclerosis, a decrement
at slow rates of stimulation may be found
that is less prominent after brief exercise
Figure 24–11. Results of the prolonged exercise test in a
and increased after prolonged exercise. Such normal subject (top) and a patient with hypokalemic peri-
abnormalities are seen more frequently when odic paralysis (bottom). The subject is studied by recording
the disease is rapidly progressive, and there the CMAP amplitude with single stimuli at 1-minute inter-
can be a decrement as high as 30%–40% vals while the patient is at rest, during 3–5 minutes of
exercise, and for 20–40 minutes after exercise. In this
in such cases. Disorders of neuromuscular patient with hypokalemic periodic paralysis, the ampli-
transmission are rarely found in peripheral tude increases 39% during exercise and then decreases
neuropathies and inflammatory myopathies.26 dramatically by 62% after exercise.
Key Points test must be individualized for each case, with

the proper selection of nerve and muscle com-
• The usual pattern seen in myasthenia
binations, rates of stimulation, and types of
gravis is a decrement at rest, in proximal exercise. The examiner must be aware of the
muscles, that repairs after brief exercise, varied abnormalities that occur with different
and then recurs and sometimes increases disease entities.
after one minute of exercise (postexercise
exhaustion).
• The usual pattern seen in Lambert–Eaton REFERENCES
myasthenic syndrome is a low-amplitude
CMAP, with a decrement present at rest, 1. Engel, A. G. 1999. Anatomy and molecular architec-
and significant facilitation (>60%) after ture of the neuromuscular junction. In Myasthenia
brief exercise. Both proximal and distal gravis and myasthenic disorders, ed. A. G. Engel,
3–39. New York: Oxford University Press.
muscles are usually affected. 2. Keesey, J. C. 1989. AAEE minimonograph #33: Elec-
• The usual pattern seen in botulism is a trodiagnostic approach to defects of neuromuscular
low-amplitude CMAP on routine nerve transmission. Muscle & Nerve 12:613–26.
conduction studies, often with minimal or 3. Sanders, D. B. 1992. Electrophysiologic study of dis-
no decrement present on repetitive stim- orders of neuromuscular transmission. In Electrodiag-
nosis in clinical neurology, ed. M. J. Aminoff, 3rd ed.,
ulation, but marked facilitation after brief 327–54. New York: Churchill Livingstone.
exercise or 50-Hz stimulation. 4. Harper, C., Jr. 1999. Electrodiagnosis of endplate dis-
• Congenital myasthenic syndromes have ease. In Myasthenia gravis and myasthenic disorders,
variable presentation; in some cases rou- ed. A. G. Engel, 65–84. New York: Oxford University
Press.
tine repetitive stimulation may be normal 5. Maddison, P., J. Newsom-Davis, and K. R. Mills.
and a decrement may only be brought out 1998. Distribution of electrophysiologic abnormality
after prolonged repetitive stimulation at in Lambert–Eaton myasthenic syndrome. Journal of
10 Hz for 5–10 minutes. Neurology, Neurosurgery, and Psychiatry 65:213–17.
• No decrement is present in periodic paral- 6. Rubin, D. I., C. M. Harper, and R. G. Auger. 2005.
Trigeminal nerve repetitive stimulation in myasthenia
ysis, but the CMAP amplitude may grad- gravis. Muscle & Nerve 29(4):591–6.
ually decrease over 20–30 minutes follow- 7. Petertska, A., R. Jarrar, and D. I. Rubin. 2006. Radial
ing 5 minutes of exercise. nerve repetitive stimulation in myasthenia gravis. Mus-
• Recessive myotonia congenita may show cle & Nerve 33(6):817–19.
8. Kennett, R. P., and R. W. Fawcett. 1993. Repet-
decrements of over 40% with trains of itive nerve stimulation of anconeus in the assess-
10-Hz stimulation. ment of neuromuscular transmission disorders. Elec-
• A small decrement may be seen in disor- troencephalography and Clinical Neurophysiology
ders such as amyotrophic lateral sclero- 89:170–6.
sis, other myotonic disorders, and rarely 9. Zifko, U. A., M. W. Nicolle, W. Grisold, and
C. F. Bolton. 1999. Repetitive phrenic nerve stimula-
in inflammatory myopathies. For this rea- tion in myasthenia gravis. Neurology 53:1083–7.
son, it is important to document a decre- 10. Sanders, D. B, ed. 1994. Myasthenia gravis and myas-
ment of at least 10% in two or more thenic syndromes. Neurologic Clinics 12:231–442.
nerve/muscle combinations. 11. Newsom-Davis, J., and B. Lang. 1999. The Lambert–
Eaton myasthenic syndrome. In Myasthenia gravis
and myasthenic disorders, ed. A. G. Engel, 205–28.
New York: Oxford University Press.
12. Baslo, M. B., F. Deymeer, P. Serdaroglu, Y. Parman,
SUMMARY C. Ozdemir, and M. Cuttini. 2006. Decrement pattern
in Lambert–Eaton myasthenic syndrome is different
The electrophysiologic technique of repetitive from myasthenia gravis. Neuromuscular Disorders 16:
stimulation is an important component of the 454–8.
evaluation of patients with fatigue, weakness, 13. Monaco, M., M. Milone, L. Padua, and P. Tonaco.
1997. Combined low rate nerve stimulation and
ptosis, diplopia, dysphagia, or dysarthria. The maximal voluntary contraction in the detection of
technique requires a knowledge of the phys- compound muscle action potential facilitation in
iology and pathophysiology of neuromuscular Lambert–Eaton myasthenic syndrome. Muscle &
transmission and the basic techniques of nerve Nerve 20(9):1207–8.
14. Oh, S. J., K. Kurokawa, G. C. Claussen, and
conduction studies for proper application and H. F. Ryan. 2005. Electrophysiological diagnostic cri-
interpretation. Errors in technique can pro- teria of Lambert–Eaton myasthenic syndrome. Muscle
duce falsely positive or negative results. The & Nerve 32(4):515–20.
15. Oh, S. J., Y. Hatanaka, G. C. Claussen, and E. Sher. 21. Jones, H. R., C. F. Bo1ton, and C. M. Harper Jr. 1996.
2007. Electrophysiological differences in seroposi- Pediatric clinical electromyography, 353–85. Philadel-
tive and seronegative Lambert–Eaton myasthenic syn- phia: Lippincott-Raven.
drome. Muscle & Nerve 35(2):178–83. 22. Vedanarayanan, V. V. 2000. Congenital myasthenic
16. Pickett, J. B. III. 1988. AAEM case report #16: syndromes. Neurologist 6:186–96.
Botulism. Muscle & Nerve 11:1201–5. 23. Michel, P., D. Sternberg, P. Jeannet, et al.
17. Sheth, R. D., B. P. Lotz, K. E. Hecox, and 2007. Comparative efficacy of repetitive nerve
A. J. Waclawik. 1999. Infantile botulism: Pitfalls sitmulation, exercise, and cold in differenti-
in electrodiagnosis. Journal of Child Neurology ating myotonic disorders. Muscle & Nerve
14:156–8. 36:643–50.
18. Chaudhry, V., and T. O. Crawford. 1999. Stimulation 24. Sander, H. W., S. N. Scelsa, M. F. Conigliari, and
single-fiber EMG in infant botulism. Muscle & Nerve S. Chokroverty. 2000. The short exercise test is normal
22:1698–703. in proximal myotonic myopathy. Clinical Neurophysi-
19. Hatanaka, T., K. Owa, M. Yasunaga, et al. 2000. ology 111:362–6.
Electrophysiological studies of a child with presumed 25. McManis, P. G., E. H. Lambert, and J. R. Daube.
botulism. Childs Nervous System 16: 84–6. 1984. The exercise test in periodic paralysis (abstract).
20. Engel, A. G., K. Ohno, and S. M. Sine. 1999. Congen- Muscle & Nerve 7:579.
ital myasthenic syndromes. In Myasthenia gravis and 26. Rubin, D. I., and R. C. Hermann. 1999. Electrophys-
myasthenic disorders, ed. A. G. Engel, 251–97. New iologic findings in amyloid myopathy. Muscle & Nerve
York: Oxford University Press. 22:355–9.
Chapter 25
Motor Evoked Potentials

Jeffrey A. Strommen
INTRODUCTION MEP PHARMACOLOGY

TECHNIQUE APPLICATIONS
MEP Stimulation Demyelinating Disease
Technical Aspects of Electrical Stroke
Stimulation Spinal Cord Injury
Technical Aspects of Magnetic Other Disorders
Stimulation
MEP Recording CONTRAINDICATIONS AND RISKS
MEP Measurements SUMMARY
INTRODUCTION methods of stimulation, in this chapter this

will refer to potentials recorded after stimu-
Motor evoked potential (MEP) recording are lation of motor structures in the central ner-
being increasingly utilized as a means to moni- vous system. Although technically, stimulation
tor the central motor pathways during surgical of a peripheral nerve with recording over the
procedures at which time these structures, par- muscle is an MEP, this will be referred to as
ticularly the spinal cord, are at risk. Several a compound muscle action potential (CMAP)
studies have also evaluated the diagnostic yield which was discussed in Chapter 23. MEPs are
in various conditions including demyelinating the response elicited distally along the spinal
disease, spondylotic myelopathy, amyotrophic cord, peripheral nerve, or muscle with stimula-
lateral sclerosis (ALS), and peripheral nerve tion of the central motor pathways (spinal cord
disorders. The ability of MEP to predict recov- or cerebral hemispheres) using either tran-
ery, for prognostication, following central ner- scranial magnetic stimulation (TMS) or tran-
vous system insults, such as stroke or spinal scranial electric stimulation (TES). In intra-
cord injury, is being developed. Some authors operative monitoring, the goal is to stimulate
have used MEP to predict disability in multi- rostral to the structure at risk and record the
ple sclerosis. Although the term motor evoked potentials at a distal site, in order to identify
potentials has been used to describe poten- potentially reversible damage. In the diagnos-
tials from muscle or nerve elicited by several tic realm, central conduction time (CCT) and
385
the size of an MEP are measured to identify ranging from 20 to 50 Hz and that optimal acti-
focal slowing in the central motor pathways or vation occurred with the anode, rather than
loss of amplitude due to axonal loss. Clinical the cathode, placed over the motor area. Thus,
and experimental studies of the efficacy and stimulation of the motor cortex differs from
safety of TES and TMS have been performed that of peripheral nerves or motor fiber tracts
extensively in most countries and clinical appli- in the spinal cord in both the polarity of stimu-
cations of transcranial electrical MEP are now lation and the importance of rate of repetitive
permitted in all countries. TMS is permitted stimulation.
in all countries except the United States where Direct stimulation of the cerebral cortex may
the method is still considered experimental. activate the dendrites, cell bodies, or the axon
hillocks of the motor neurons in the precen-
tral gyrus. MEPs depend on activation at the
Purpose and Role of MEPs
axon hillocks in the depths of the cerebral
• Primary use is to monitor central motor cortex. With TES the anode is placed over
pathways that are at risk during surgical the area of the cortex to be stimulated. In
procedures. contrast to peripheral stimulation, the anode
• May have predictive value in central ner- more effectively stimulates the pyramidal cells
vous system injuries such as stroke and of the cortex because of its ability to induce
spinal cord injury. hyperpolarization of the apical dendrite and
• May provide additional diagnostic infor- depolarization of the axon hillock. This orienta-
mation in processes that lead to slowing tion enhances hyperpolarization at the cortical
of the central motor pathways such as surface and depolarization in the deep lay-
multiple sclerosis. ers of the cortical gray matter. With magnetic
stimulation, an intense current in an external
coil induces local depolarizing electric currents
that flow through the neuron and axon hillock.
TECHNIQUE Both depolarizations initiate descending action
potentials in the corticospinal pathway.
MEP Stimulation Cortical stimulation produces a series of
descending action potentials in the corti-
Theoretically, MEP can be obtained with stim- cospinal tracts. The D (direct) wave results
ulation anywhere along the peripheral or cen- from depolarization of the axon hillocks of the
tral motor axis. For clinical purposes, stim- large motor neurons. These are followed by a
ulation is generally performed in the central series of I (indirect) waves that reflect activa-
nervous system, either at the cerebral cortex tion of the cortical interneurons. The D and
or at the cervical spinal cord.1 At each loca- I waves can be recorded directly from the
tion, motor pathways can be activated by either spinal cord but with single pulse stimulation
electric or magnetic stimulation. Although the they generally will not depolarize the anterior
technical aspects of electric and magnetic stim- horn cell. MEPs are augmented with repet-
ulation differ, the physiology of MEP activation itive stimulation at both the cortical and the
is similar for both forms of stimulation. spinal cord levels. Trains of action potential
Direct electrical activation of the motor summate and facilitate depolarization of the
pathways at the level of the cerebral cortex has anterior horn cell above threshold for firing.
been used in experimental animals for many Figure 25–1 shows temporal summation at the
years to study the motor pathways. Penfield anterior horn cell.3 Anterior horn cell activa-
conducted the first extensive study of stimulation propagates axonal action potentials and
tion of the motor cortex in humans more than muscle depolarization that can be recorded
50 years ago during surgical procedures for as nerve and muscle action potentials. For
epilepsy.2 He noted that the responses were clinical purposes, a series of 3–10 stimuli to
attenuated substantially by anesthesia and so the brain generally produces optimal activation
conducted many of his operations with patients of MEP. Newly designed transcranial, electric
under local or light anesthesia. He also rec- and magnetic, stimulators provide repetitive
ognized that the cerebral cortex was activated stimulation that enhances MEP. Comparable
more readily with rapid repetitive stimulation, enhancement of MEP also can be obtained
Motor Evoked Potentials 387
the examined muscles, prestimulation of the

D1 appropriate mixed nerve, and mental simu-
I Cervical epidural lation of movement have been suggested to
Thenar muscle facilitate MEP in disorders in which the patient
is unable to voluntarily contract the muscle.5–7
D2 40 μV
100 μV Key Points
• Cortical stimulation can be performed
with either electrical or magnetic stimula-
D3 tion.
• Electrical stimulation is utilized routinely
for intraoperative monitoring whereas
magnetic stimulation is generally used for
diagnostic studies as it is better tolerated.
D4 • Transcortical electrical stimulation uses
anodal activation which differs from stim-
ulation of the peripheral nervous system
where cathode activation occurs.
D5 • Motor neurons are activated by the sum-
mation of the initial direct wave (D) and
series of indirect waves (I) arising from the
pyramidal axons and cortical interneurons
respectively.
• Responses are facilitated with paired stim-
ulation (3–10) and interstimulus interval
0 10 20 30 40 50 of 1–4 ms. These are most commonly facil-
MS
itated with slight voluntary activation.
Figure 25–1. TES epidural and muscle responses during
C2 to C3 meningioma surgery under propofol and fen-
tanyl anesthesia omitting neuromuscular blockade. Epidu-
ral recording (bandwidth 500–10,000 Hz) used a bipolar Technical Aspects of Electrical
electrode inserted caudal to the lesion after opening. Mus- Stimulation
cle recording (bandwidth 20–3000 Hz) used intramuscular
needle pairs. Traces consist of two superimposed aver-
ages of five trials. D1 through D5 are sequential D-wave Transcranial electric MEPs of the motor cortex
corticospinal volleys following one to five pulses (Nico- through the intact bony skull or percutaneous
let Viking, 0.5-ms pulse duration, 250-Hz train frequency, stimulation of the spinal cord are used for
250 V) through 9-mm cup electrodes collodion fixed 1 cm most intraoperative monitoring. Direct stimu-
anterior to C1/C2. A small I wave follows each. Temporal
summation produced progressive motor unit recruitment lation of the exposed cortex is used occasion-
with three through five-pulse trains. The brain stimulation, ally for location of the motor cortex during
epidural electrode, and induced movement raise safety tumor surgery and epilepsy surgery, and even
concerns discussed in this chapter. Single-pulse epidu- less commonly for clinical diagnostic purposes.
ral and five-pulse train muscle MEP monitoring were
successful without adverse effects in this patient. (From
The high resistance of bone to electric current
MacDonald, D. B. 2002. Safety of intraoperative transcra- requires much higher applied voltages to drive
nial electrical stimulation motor evoked potential moni- the few milliamps of current to the cerebral
toring. Journal of Clinical Neurophysiology 19(5): 416–29. cortex needed for cortical stimulation. TES
By permission of Lippincott Williams & Wilkins.) in normal adults often requires stimuli up to
1000 V which can be quite painful as there is
by low-level voluntary activation, which par- considerable contraction of the cranial muscu-
tially depolarizes the cell body and axon hillock lature. By using brief, high voltage shocks from
before the stimulus is applied4 (Fig. 25–2). On a stimulator with a low output impedance, the
electromyographic (EMG) recordings from the resistance of the scalp and calvarium can be
muscles used to record MEP, 15% activation overcome. Although there has not been consis-
has been found to give optimal facilitation of tent standardization of stimulation parameters,
MEP. Other methods including vibration of a recent study by Szelenyi et al. evaluated
0% Contraction 5% Contraction 10% Contraction
1.0RMT
S4
1.2RMT
1 mV
1.5RMT
1.0RMT S5
1.2RMT
1 mV
1.5RMT
0 25 50 75 ms 0 25 50 75 ms 0 25 50 75 ms
Figure 25–2. Variability of motor potentials evoked by TMS depends on muscle activation. TMS induced MEPs recorded
form the extensor digitorum communis (EDC). Each panel shows five superimposed MEPs at a single stimulus intensity
at one activation level. (From Darling, W. G., S. L. Wold, and A. J. Butler. 2006. Variability of motor potentials evoked by
transcranial magnetic stimulation depends on muscle activation. Experimental Brain Research 174:376–85. By permission
of Springer-Verlag.)
various parameters during intraoperative mon- stimuli with an interstimulus interval of 1–4 ms
itoring and recommended the following when are given at intensities of 200–500 V (Mul-
recording over a limb muscle:8 tiPulse Cortical Stimulator D185, Digitimer
Ltd, Welwyn Garden City, Hertfordshire, UK).
1. The lowest stimulus intensity to elicit These parameters are varied until a repro-
MEP in the tibialis anterior (TA) and ducible response can be recorded in all the
abductor pollicis longus (APB) is achieved muscles examined. The polarity of the simula-
with a train of individual 0.5-ms duration is also switched to assure maximal anodal
tion electric pulses (many commercial stimulation. Electric activation of the motor
machines have shorter duration pulses to pathways in the spinal cord occurs by depolar-
reduce pain). ization of the axons in the corticospinal tract
2. An interpulse interval of 4 ms gave the by the cathode, with hyperpolarization at the
lowest motor thresholds but did not differ anode, similar to that of a peripheral nerve.
significantly from 3 ms. This can be obtained by surface electrodes with
3. Stimulating electrode locations of C3/C4 stimuli greater than 1000 V, similar to the elec-
or C4/C3 (international 10–20 EEG sys- tric stimuli used in transcranial stimulation.
tem) gave the lowest stimulation thresh- However, in the operating room, activation is
olds. possible with lower intensity stimuli by a flat
plate cathode directly on the dura mater in the
MEP stimulation protocols vary among insti- surgical field or by a needle electrode cathode
tutions, but anodal stimulation, with 0.05 ms placed either in the interspinous ligament or
rapid rise time pulses to subcutaneously placed over the vertebral lamina. Anode placement for
electroencephalographic (EEG) electrodes at activation of the spinal cord may be epidural,
C3 and C4 is commonly used. Two to five laminar, esophageal, subcutaneous, or on the
surface of the skin at some distance from the cranial muscles that requires only a gauze
cathode. Stimulation of the spinal cord is also pad between the teeth; it is not reliable for
enhanced by repetitive stimulation at intervals intraoperative monitoring because of smaller
of 3–5 ms. In surgical settings, activation of responses, marked suppression with anesthetic
motor pathways in the cervical spinal cord is agents, and difficulty with stabilization of the
easier than cortical stimulation because anes- magnetic coil. Activation of the axons in the
thetics have less effect on the spinal cord than cervical corticospinal tract and the neurons in
on the cerebral cortex. However, spinal cord the motor cortex depends on the same mecha-
stimulation produces retrograde dorsal column nism. In both cases, an intense pulse of current
activation as well as anterograde activation of in an external coil induces a magnetic field
the motor pathways. Thus, a descending poten- that traverses the skin and bone without effect.
tial recorded from the spinal cord or individual In TMS, the electrically induced pulse of the
peripheral nerve will be from both motor and magnetic field induces current flow in the
sensory fibers.9 Such sensory responses are not underlying tissue of the brain or spinal cord.11
seen when recording from muscle because of At threshold the local current flow causes neu-
their relatively smaller size. With the tech- ronal or axonal depolarization; this depolariza-
nological development of modern stimulating tion initiates descending action potentials in
equipment, TES has become the preferred the corticospinal pathway. Magnetic stimula-
method with an overall success rate of obtain- tion differs from electric stimulation in that no
ing TES MEP of 94.8% in the upper extrem- current flow occurs in the superficial levels of
ities and 66.6% in the lower extremities.10 the skin, muscle, or bone. All current flow is
However, the ability to record MEP is signif- induced intracranially by the rapid change in
icantly reduced in patients under 7 years of current in the stimulating coil. With optimal
age, those over 64 years, and in patients with recording and no interference, TMS MEP can
more pronounced, preexisting clinical neuro- initiate supramaximal or nearly supramaximal
logic deficits. CMAP.
TMS is performed with coils of different
sizes and shapes for different clinical appli-
Key Points cations. In TMS an increase or decrease of
current or a change in direction of current in
• TES is the preferred method for eliciting
the coil induces comparable changes in the
MEP in the intraoperative setting. intracranial or intraspinal local currents. Thus
• MEP is most readily elicited with stim-
it is necessary to vary the orientation of the
ulation at C3–C4 for both the APB and coil and direction of current flow to induce
the TA (better responses occur in a few the current needed for activation of the neu-
patients from CZ–FZ.) ral tissue. In general, for lower limb MEP
• Maximal MEP is obtained with 3–5
the coil is centered around Cz whereas for
pulses, an interstimulus interval of 3–4 ms, upper extremity TMS the coil is placed just
and pulse duration of 0.5 ms. laterally. Single or paired stimuli of increas-
• Most responses will be obtained with
ing intensity are applied until a stable MEP
intensity of 200–500 V. is recorded or the maximal output (100%) is
• MEP is more difficult to obtain in those
reached. The coil can then be adjusted by rota-
under 7 or over 64 years of age and those tion and lateral movement to obtain a maximal
with more severe preexisting neurologic response.
deficits.
Technical Aspects of Magnetic Key Points

Stimulation • TMS is less painful and preferred for clin-
ical diagnostic testing.
In awake subjects, TMS is the preferred tech- • A magnetic field applied through the skull
nique since the magnetic pulse is painless induces current flow in underlying brain
while the high voltage electric shock is quite tissue with little current in the skin or
painful. TMS produces some contraction of subcutaneous tissue.
• For lower extremity MEP, the coil is cen- (Fig. 25–3). Although this has the potential
tered in the region of Cz and for the upper advantage of allowing complete neuromuscular
limb it is just lateral. block, it is a small response that is more likely
• TMS is less useful of the intraoperative arising from small responses of surrounding,
applications due to the marked sensitivity paralyzed muscles. This response may not be a
to anesthetic agents. pure motor response since it likely includes ret-
rograde somatosensory potential, which could
lead to a false-positive response.14
MEPs recorded from muscle have become
MEP Recording the recording of choice in most laboratories.
With either cortical or spinal cord stimula-
While magnetic coil positioning can focus stim- tion, CMAP can be recorded from the muscle
ulation more on one than another area of with surface, subdermal, or intramuscular elec-
the brain, stimulation of the motor pathways trodes (Fig. 25–4). MEP elicited by stimulation
at either the cortical or the spinal cord level of the cerebral cortex or the cervical spinal
activates multiple descending pathways with cord may be recorded relatively easily from
contraction of many muscles. Recording elec- most limb muscles with selection of the mus-
trodes can thus be placed anywhere along cles for recording determined by the clinical
the descending pathway. In the spinal cord, problem (Fig 25–5). Typically, the potentials
the motor volley can be recorded with elec- are recorded with surface electrodes, but sub-
trodes placed directly on the cord or in the cutaneous electrodes, intramuscular wires, or
epidural space. With the technique one can other electrodes can be used. Intramuscular
relatively easily record the D and I waves recording are used with caution as they lead
during surgical procedures with single pulses to less quantifiable CMAP. These potentials are
at a relatively low intensity (Fig. 25–1). As reduced by the neuromuscular blocking agents
cord MEPs are not affected by neuromuscu- usually required during surgical procedures,
lar blocking agents, they can be recorded with but with controlled blockade they can usually
complete neuromuscular block. Disadvantages be recorded with paired stimulation. The stan-
of cord MEP include possible hematoma from dard gains, filter settings, with longer sweep
catheter placement as well as inability to rec- speeds than used for peripheral motor con-
ognize unilateral loss of responses masking a duction studies are usually satisfactory. Some
persistent contralateral response.12 The motor MEPs during surgery are small enough to
response can also be recorded over periph- require averaging. Typically, the ankle extensor
eral nerves with spinal cord stimulation and and flexor muscles and the quadriceps mus-
has been referred to as neurogenic MEP13 cles are tested in the lower extremity, while the
Figure 25–3. Neurogenic MEPs recorded from the right and left sciatic nerves during spinal deformity surgery. (From
Owen, L. H., J. Laschinger, K. Bridwell, et al. 1988. Sensitivity and specificity of somatosensory evoked and neurogenic
MEPs in animals and humans. Spine 13:1111–18. By permission of Lippincott Williams & Wilkins.)
MEP during scoliosis surgery
Spinal stimulation: Paired ISI 3 ms, 80 mA, dur 1 ms

Esophageal-left cervical
left rect. fem.
left ant. tib
left soleus
right rect. fem.
right ant. tib.
right soleus
1 mV 10 ms
neuromuscular block 10%
Figure 25–4. Spinal stimulation with recording of CMAPs by surface electrodes over selected lower extremity muscles.
Stimulation in this case was with an anode–cathode set up from an esophageal to percutaneous laminar needle.
TES C3–C4 Cz
TMS Cz
C3
Stim: TC
rectus femoris
50 μV
A Tib
10 ms
Gastroc
narcotic anesthesia
Figure 25–5. Transcortical stimulation with recording of CMAPs by surface electrodes over selected lower extremity
muscles. These responses were recorded with electrical stimulation at C3–C4 during spine surgery with partial neuromus-
cular blockade. Magnetic stimulation will also elicit lower extremity CMAPs when the magnet is centered over Cz but these
are generally not used during surgery.
391
intrinsic hand muscles, forearm extensors, and sensitive to identify mild involvement by dis-
arm flexors are tested in the upper extremities. ease. Because MEPs are designed primarily to
The thenar muscle in the hand and the ante- detect disease of the central motor pathway,
rior tibial muscle in the leg are the sites most several methods have been developed to sub-
frequently used and provide the most defini- tract the latency of the peripheral segment of
tive normal data. MEP can also be recorded the motor pathway from the total latency, and
from facial innervated muscles during cranial thereby distinguish conduction over the cen-
base surgeries.15 tral portion of the motor pathways. Central
conduction time is the most commonly used
measure to identify central disorders.16 CCT
Key Points is calculated by subtracting the time needed
• Motor evoked recordings directly over the for the signal to travel over the peripheral seg-
spinal cord or peripheral nerve are eas- ment (spinal cord to the muscle) from the
ily obtainable and are less affected by total latency (site of stimulation to the mus-
anesthesia but have other drawbacks. cle). The latency of the peripheral segment
• MEPs recorded over the muscle are eas- is measured in one of two ways. The latency
ily recorded but requires either no or of a lumbar spine MEP is the most direct.
controlled neuromuscular blockade. Peripheral latency can also be obtained indi-
• Recording locations vary based on the rectly from F-wave measurements made dur-
clinical problem but are most reliable ing standard nerve conduction studies. F-wave
when obtained in the TA and thenar mus- traverses the peripheral motor pathway from
cles. the distal limb to the anterior horn cell and
• Surface or subdermal needle electrodes back to the muscle; thus, F-wave latency minus
provide the most quantifiable CMAP the distal M-wave latency divided by two plus
response. the distal latency. The direct or F-wave calcu-
lated spinal MEP latency subtracted from the
latency obtained with cortical stimulation gives
the CCT. Thresholds for activation of an MEP
have been used widely in physiologic studies of
MEP Measurements cortical function, but they have not proved of
value in clinical assessment.
With peripheral recordings, three measure-
ments of MEP have been used: latency, ampli- Key Points
tude, and threshold. MEP amplitude is more
variable in size and configuration than CMAP • Amplitude of MEP is highly variable and
evoked with peripheral stimulation. They can is thus of limited clinical utility.
vary with the attention and level of relaxation • An MEP amplitude reduction of >50%
in an awake patient. Thus, amplitude measures at critical periods during intraoperative
are of limited clinical diagnostic value. How- monitoring is generally considered signifi-
ever, major changes in amplitude can be used cant.
to detect motor pathway damage during surgi- • CCT is the most useful clinical diagnos-
cal monitoring. In general, a 50% reduction in tic measurement in various diseases but
amplitude from an average baseline value at a has limited utility in intraoperative mon-
critical time during surgery is considered sig- itoring.
nificant. Given the significant variability, some • CCT is most commonly measured using
centers have used only a loss of the response as calculated peripheral segment latencies
significant (see Chapter 44). based on F-waves subtracted by the total
The latency of responses from the time of conduction time from cortical stimulation.
stimulation to recording is the most reliable
measurement. The most direct method of mea-
suring MEP latency is by direct measurements MEP PHARMACOLOGY
that are compared with those of normal sub-
jects matched for age, height, and sex; how- The effect of drugs on the MEP of the awake
ever, absolute latencies may not be sufficiently patient is minimal. Minimal changes have
been reported with sedating agents such as approved in the United States, and is therefore
lorazepam, diazepam, and midazolam. These much more heavily used but only in monitor-
are not sufficient to preclude MEP testing in ing for surgery. In developed countries outside
patients on these medications. the United States, TES is as commonly used in
The primary factor affecting the MEPs in surgical monitoring.
the operating room is anesthesia. Anesthetic Normal data have been obtained in only
agents can suppress the response at multi- a few laboratories (and for specific stimula-
ple sites, particularly at central synapses in tion and recording methods that have not
the cortex, anterior horn cell, and less so been widely adopted).18 Furthermore, vari-
at the neuromuscular junction. Halogenated ability of the responses has raised concern
inhalation agents easily abolish MEPs by block- about the reliability of the responses for clin-
ade at the cortex and anterior horn cell.17 ical interpretation.6, 19 Although some reports
If used at all, the concentration of these agents include normative data, most do not. Conse-
need to remain very low (<0.5%). While not quently, generally accepted values for MEP
universally the case, nitrous oxide can also amplitude and latencies do not exist. For
suppress MEP dramatically. Since MEPs are clinical studies, amplitude measures gener-
recorded from muscle, neuromuscular block- ally are not used because of the marked
ing agents will also suppress or eliminate differences among normal subjects. Latency
them. MEP can still be reliably recorded with measures can show marked slowing or dis-
neuromuscular blockade up to 50%. Contin- persion of the response. Such a finding is
uous monitoring of the level of neuromus- strong evidence for a demyelinating process, as
cular junction block is valuable in assuring in multiple sclerosis. Finally, although MEPs
stable MEP. The remaining movement with have been shown to have clinical value in
MEP requires that the surgeon be warned patients with central motor process, clinical
before each stimulus. Unless core body tem- examination, magnetic resonance imaging, and
perature is quite low, temperature produces other laboratory testing are often more spe-
only a mild, gradual increase in stimulation cific. Despite this, a large study in Italy using
threshold.17 TMS revealed that this technique can be
highly accurate as a diagnostic test in many
Key Points conditions.20 In this study the overall agree-
ment between clinical and electrophysiologic
• Inhalation of anesthetic agents plays a abnormalities was 87% with higher sensitiv-
major role in suppressing MEP during ities in spinal cord pathology (0.85), heredi-
intraoperative monitoring. tary spastic paraplegia (0.8), and motor neuron
• In the awake patient there is little sup- disease (0.74). MEPs have been tested and
pression from oral anxiolytic, or sedative reported on as a diagnostic aid in many neu-
agents. rologic disorders, including cerebral infarcts,
• Neuromuscular blocking agents should Parkinson’s disease, other movement disor-
be minimized and closely monitored ders, motor neuron disease, cervical spondylo-
but are acceptable with blockade up sis, and less common disorders. More recent
to 50%. reports have attempted to use MEP to pre-
dict prognosis in injuries to the central ner-
vous system such as in stroke or spinal cord
APPLICATIONS injury.
MEP recordings are commonly used in
MEPs are used for two major purposes: clinical monitoring neural function during surgery in
diagnosis and operative monitoring of neural major medical centers, but have not other-
function. TMS are widely used for diagnos- wise been widely adopted. In the operative
tic purposes outside the United States, but setting the goal is to protect central motor
are not approved for this use in this country pathways which may be at risk at a time
because of the time and effort that would be when clinical examination is not possible. More
required by the manufacturers. TMS in the recently, they have become a valuable tool
United States is limited to studies conducted in brachial plexus exploration and reconstruc-
under research protocols. In contrast, TES is tion, in assessing whether the motor root is
avulsed. These applications will be discussed in Key Points

Chapters 43–45.
• MEP is frequently abnormal and may
serve as a valuable diagnostic tool in the
assessment of central demyelinating
disease.
Demyelinating Disease • CCT and motor dispersion may be evident
in central demyelinating diseases.
In clinical diagnostic neurology, the most com-
mon use of MEP is the search for evi-
dence of demyelinating disease.21 MEPs show
marked slowing and dispersion in the pres- Stroke
ence of demyelinating disease in the spinal
cord or brain stem, particularly in transverse MEPs have been studied as a predictor of out-
myelitis22, 23 (Fig. 25–6). Such changes may be come in stroke. Many variations in MEP meth-
present even in the absence of any major or ods have been used to evaluate different stroke
clear-cut neurologic deficit. Therefore, MEP populations, but there still does not appear to
can be used to identify subclinical lesions in be a clear consensus on the predictive value
patients with multiple sclerosis and help define of functional return after stroke.26, 27 MEP test-
the significance of fatigue.24 Many reports have ing has been applied in studies of recovery
shown the sensitivity of MEP in detecting after severe traumatic injury to the brain and
and characterizing slowing of conduction in in acute brain stem lesions and may be helpful
the motor pathways that may not produce a in predicting the outcome.28, 29
measurable motor deficit. Virtually all patients
with multiple sclerosis who have weakness
in a limb also have abnormal MEP in that Spinal Cord Injury
limb. Recent studies have attempted to use
MEP as an objective measure to document MEPs are frequently abnormal in patients with
functional consequences of multiple sclero- cervical or lumbar spondylosis.30 The changes
sis, thus providing an outcome measure to may be a prolongation of absolute latency at a
estimate prognosis and document treatment single site or a change in latency and configura-
effects and assist with interpretation of clinical tion between two sites of recording. In cervical
treatment trials. Kalkers et al. analyzed TMS as spondylosis, abnormality on MEP testing may
a valid measure of evaluating neurologic dys- be caused by bony compression of the spinal
function and found that the degree of abnor- cord or the spinal nerves in the interverte-
mality did correlate with standard multiple bral foramina. The abnormalities are increased
sclerosis.25 latency and decreased amplitude.31, 32 A recent
Multiple sclerosis
Normal subject with normal strength
Right ADM Left ADM Right ADM Left ADM
Wrist Wrist
5 mV
C7/T1 C7/T1
13.9 14.3 15.3 14.9
2 mV
Scalp Scalp
10 ms 19.5 19.2 31.8 29.0
10 ms
Figure 25–6. Hand MEPs in a normal subject (left) and in a patient with documented multiple sclerosis (right). Function
and clinical testing were normal in the hand and arm. ADM, abductor digiti minimi of the hand. Black triangles, MEP
latencies. (From Hess, C. W., K. R. Mills, N. M. Murray, and T. N. Schriefer. 1987. Magnetic brain stimulation: Central
motor conduction studies in multiple sclerosis. Annals of Neurology 22:744–52. By permission of John Wiley & Sons.)
study has also found significant correlation with from loss of amplitude and loss of facilitation
MRI abnormalities and suggests MEP as a and, thus, MEP testing is unlikely to add to
screening tool to detect cervical spondylotic the standard EMG evaluation despite some
myelopathy.33 Generally, a change in latency is reported abnormalities.36, 37 Cortical hyperex-
more prominent than a reduction in amplitude. citability has been shown with MEP testing as
Amplitude reduction is particularly difficult to an early feature of motor neuron disease estab-
assess in MEP as it may be difficult to obtain lishing the presence of both upper and lower
a supramaximal response in many normal sub- motor neuron but without clear evidence of
jects, particularly in the lower extremities. diagnostic value.38 MEP testing also has been
Assessing functional preservation of the applied extensively in analyzing in Parkinson’s
motor tracts and predicting recovery in spinal disease. One unconfirmed TMS study reported
cord injury has become a major topic of early diagnosis of central motor changes and
research with the increasing number of treat- difference in compensational capacity based on
ment trials addressing spinal cord repair with the type of disease.39 In completely paralyzed
growth factors, stem cells, and bridging grafts. patients with an apparent psychogenic paraly-
MEP recording has been used in this setting sis, MEP can be useful if the motor pathway
to determine level of injury, completeness of from cortex to muscle is intact. MEP should
injury, and to monitor recovery after injury. In be markedly abnormal or absent in a struc-
a study by Curt et al., it was shown that an tural disorder that produces complete paral-
absence of a recordable abductor digiti minimi ysis, in contrast to the normal MEP seen in
(ADM) MEP with TMS was very highly pre- psychogenic disorders.40
dictive of poor recovery of intrinsic hand func-
tion. In addition, ambulatory capacity could
be predicted in that those with normal MEP
recorded in the TA regained full ambulatory CONTRAINDICATIONS AND
ability whereas if this response was absent 78% RISKS
showed no or only minimal ambulatory capac-
ity.34 Lack of facilitation of MEP amplitudes Reports of complications of MEP testing
has been utilized as a measure of motor tract are few. The most extensive safety review
involvement in incomplete spinal cord injuries reported remarkably few adverse events.3 In
but requires more study to determine the clin- over 15,000 patients the most common adverse
ical and prognostic significance.35 event was tongue or lip lacerations which
could be prevented by placement of a bite
Key Points block. Seizures were very rare occurring in
only five reported cases, usually with underly-
• The value of MEP after strokes remains to ing epilepsy, cardiac arrhythmia in five patients
be determined. and intraoperative awareness in one. There
• Latency abnormalities correspond well to were no recognized adverse neuropsychologic
structural damage to roots or cord in cer- effects, headaches, or endocrine disturbances.
vical spondylosis. Based on these findings the relative contraindi-
• MEP may well become important for fol- cations include epilepsy, cortical lesions, con-
lowing the course of recovery with new vexity skull defects, raised intracranial pres-
treatments for spinal cord injury. sure, cardiac disease, being on medication for
epilepsy testing under anesthesia, intracranial
electrodes, vascular clips or shunts, and cardiac
Other Disorders pacemakers or other implantable biomedical
devices. Additional exclusion criteria provided
Reports in a small number of patients with by the manufacturer of the D185 MultiPulse
other disorders who have had MEP testing are Cortical Stimulator (Digitimer, Welwyn Gar-
available, but their clinical significance is dif- den City, Hertfordshire, UK) include history
ficult to determine. In ALS, MEP testing is of stroke or psychiatric disorder. Few studies
reported to show abnormalities of latency and have assessed MEP in children,41 but current
threshold, but these are difficult to distinguish age ranges as per Digitimer’s operator manual
from the changes that would be expected solely report no other risks in ages from 3 to 99 years.
SUMMARY 9. Machida, M., S. L. Weinstein, T. Yamada, J. Kimura,

T. Itagaki , and T. Usui. 1988. Monitoring of motor
MEP recordings are a safe and effective means action potentials after stimulation of the spinal cord.
Journal of Bone and Joint Surgery (American Volume)
of assessing conduction along the central and 70:911–18.
peripheral motor pathways in a variety of clin- 10. Chen, X., D. Sterio, X. Ming, et al. 2007. Success
ical settings. MEP is proven to be effective in rate of motor evoked potentials for intraoperative neu-
monitoring central motor pathways that are at rophysiologic monitoring: Effects of age, lesion loca-
tion and preoperative neurologic deficits. Journal of
risk during surgical procedures and is finding Clinical Neurophysiology 24(3):281–5.
usefulness in the diagnosis and prognosis of 11. Counter, S. A., and E. Borg. 1992. Analysis of the coil
several central nervous system disorders such generated impulse noise in extracranial magnetic stim-
as multiple sclerosis, stroke, and spinal cord ulation. Electroencephalography and Clinical Neuro-
injury. The responses can be elicited by a vari- physiology 85:280–8.
12. Burke, D., and R. G. Hicks. 1998. Surgical monitoring
ety of stimulation techniques using either mag- or motor pathways. Journal of Clinical Neurophysiol-
netic or electrical stimulation to the cortex, ogy 15(3):194–205.
spinal cord, or peripheral nerve. Recordings 13. Owen, L. H., J. Laschinger, K. Bridwell, et al. 1988.
can be obtained from several structures includ- Sensitivity and specificity of somatosensory evoked and
neurogenic-motor evoked potentials in animals and
ing the spinal cord, peripheral nerve, or mus- humans. Spine 13:1111–18.
cle dependent on the clinical application. The 14. Toleikis, J. R., J. P. Skelly, A. O. Carlvin, and
clinical neurophysiologist should have a basic J. K. Burkus. 2000. Spinally elicited peripheral nerve
understanding of the techniques as well as the responses are sensory rather than motor. Clinical Neu-
potential physiologic and technical factors that rophysiology 111:736–42.
15. Akagami, R., C. Dong, and B. D. Westerberg. 2005.
need to be accounted for in the interpretation Localized transcranial electrical motor evoked poten-
of these studies. tials for monitoring cranial nerves in cranial base
surgery. Neurosurgery 57:ONS-78–85.
16. Claus D. 1990. Central motor conduction: Method
and normal results. Muscle & Nerve 13:1125–232.
REFERENCES 17. Sloan, T. B., and E. J. Heyer. 2002. Anesthe-
sia for intraoperative neurophysiologic monitoring of
1. Rossini, P. M. 1990. Methodological and phys- the spinal cord. Journal of Clinical Neurophysiology
iological aspects of motor evoked potentials. 19(5):430–43.
Electroencephalography Clinical Neurophysiology 18. Eisen, A., S. Siejka, M. Schulzer, and D. Calne.
41(Suppl):124–33. 1991. Age-dependent decline in motor evoked poten-
2. Penfield, W., and H. Jasper. 1954. Epilepsy and the tial (MEP) amplitude: With a comment on changes
functional anatomy of the human brain. Boston: Little, in Parkinson’s disease. Electroencephalography and
Brown. Clinical Neurophysiology 81:209–15.
3. MacDonald, D. B. 2002. Safety of intraoperative tran- 19. Triggs, W. J., L. Kiers, D. Cros, J. Fang, and
scranial electrical stimulation motor evoked poten- K. H. Chiappa. 1993. Facilitation of magnetic motor
tial monitoring. Journal of Clinical Neurophysiology evoked potentials during the cortical stimulation silent
19(5): 416–29. period. Neurology 43:2615–20.
4. Darling, W. G., S. L. Wold, and A. J. Butler. 2006. 20. Di Lazzaro, V., A. Oliviero, P. Profice, et al. 1999. The
Variability of motor potentials evoked by transcranial diagnostic value of motor evoked potentials. Clinical
magnetic stimulation depends on muscle activation. Neurophysiology 110:1297–307.
Experimental Brain Research 174:376–85. 21. Jones, S. M., L. J. Streletz, V. E. Raab, R. L. Knobler,
5. Rossini, P. M., and S. Rossi. 1998. Clinical applications and F. D. Lublin. 1991. Lower extremity motor evoked
of motor evoked potentials. Electroencephalography potentials in multiple sclerosis. Archives of Neurology
and Clinical Neurophysiology 106:180–94. 48:944–8.
6. Magistris, M. R., K. M. Rosler, A. Truffert, and 22. Ho, K. H., M. Lee, K. Nithi, J. Palace, and K. Mills.
J. P. Myers. 1998. Transcranial stimulation excites vir- 1999. Changes in motor evoked potentials to short-
tually all motor neurons supplying the target muscle. interval paired transcranial magnetic stimuli in multi-
A demonstration and a method improving the study of ple sclerosis. Clinical Neurophysiology 110:712–19.
motor evoked potentials. Brain 121:437–50. 23. Kalita, J., and U. K. Misra. 2000. Neurophysiolog-
7. Han, T. R., J. H. Kim, and J. Y. Lim. 2001. Opti- ical studies in acute transverse myelitis. Journal of
mization of facilitation related to threshold in transcra- Neurology 247:943–8.
nial magnetic stimulation. Clinical Neurophysiology 24. Colombo, B., F. Martinelli Boneschi, P. Rossi, et al.
112:593–9. 2000. MRI and motor evoked potential findings
8. Szelenyi, A., K. F. Kothbauer, and V. Deletis. 2007. in nondisabled multiple sclerosis patients with and
Transcranial electrical stimulation for intraopera- without symptoms of fatigue. Journal of Neurology
tive motor evoked potential monitoring: Stimulation 247:506–9.
parameters and electrode montages. Clinical Neuro- 25. Kalkers, N. F., R. L. M. Strijers, M. M. S Jasperse, et al.
physiology 118: 1586–95. 2007. Motor evoked potential: A reliable and objective
measure to document the functional consequences in cervical spondylosis. Journal of Neurological Sci-
of multiple sclerosis? Relation to disability and MRI. ences 244:17–21.
Clinical Neurophysiology 118:1332–40. 34. Curt, A., M. E. Keck, and V. Dietz. 1998. Functional
26. Pennisi, G., G. Rapisarda, R. Bella, et al. 1999. outcome following spinal cord injury: Significance of
Absence of response to early transcranial magnetic motor-evoked potentials and ASIA scores. Archives of
stimulation in ischemic stroke patients: Prognostic Physics Medicine and Rehabilitation 79:81–6.
value for hand motor recovery. Stroke 30:2666–70. 35. Diehl, P., U. Kliesch, V. Dietz, and A. Curt. 2006.
27. Feys, H., J. van Hees, F. Bruyninckx, R. Mercelis, Impaired facilitation of motor evoked potentials in
and W. De Weerdt. 2000. Value of somatosensory and incomplete spinal cord injury. Journal of Neurology
motor evoked potentials in predicting arm recovery 253:51–7.
after a stroke. Journal of Neurology, Neurosurgery and 36. Kohara, N., R. Kaji, Y. Kojima., et al. 1996. Abnor-
Psychiatry 68:323–31. mal excitability of the corticospinal pathway in patients
28. Moosavi, S. H., P. H. Ellaway, M. Catley, M. J. Stokes, with amyotrophic lateral sclerosis: A single motor
and N. Haque. 1999. Corticospinal function in severe unit study using transcranial magnetic stimulation.
brain injury assessed using magnetic stimulation of the Electroencephalography and Clinical Neurophysiol-
motor cortex in man. Journal of Neurological Sciences ogy 101:32–41.
164:179–86. 37. Truffert, A., K. M. Rosler, and M. R. Magistris. 2000.
29. Schwarz, S., W. Hacke, and S. Schwab. 2000. Mag- Amyotrophic lateral sclerosis versus cervical spondy-
netic evoked potentials in neurocritical care patients lotic myelopathy: A study using transcranial magnetic
with acute brain stem lesions. Journal of Neurological stimulation with recordings from the trapezius and
Sciences 172:30–7. limb muscles. Clinical Neurophysiology 111:1031–8.
30. Travlos, A., B. Pant, and A. Eisen. 1992. Transcra- 38. Vucic, S., and M. C. Kiernan. 2006. Novel thresh-
nial magnetic stimulation for detection of preclinical old tracking techniques suggest that cortical hyperex-
cervical spondylotic myelopathy. Archives of Physics citability is an early feature of motor neuron disease.
Medicine and Rehabilitation 73:442–6. Brain 129:2436–46.
31. Tavy, D. L., H. Franssen, R. W. Keunen, A. R. Wat- 39. Guekht, A., M. Selikhova, M. Serkin, and E. Gusev.
tendorff, R. E. Hekster, and A. C. van Huffelen. 2005. Implementation of the TMS in the early stages
1999. Motor and somatosensory evoked potentials in of Parkinson’s disease. Electromyography and Clinical
asymptomatic spondylotic cord compression. Muscle Neurophysiology 45(5):291–7.
& Nerve 22:628–34. 40. Pillai, J. J., S. Markind, L. J. Streletz, H. L. Field,
32. Nakanishi, K., N. Tanaka, N. Kamei, et al. 2007. Signif- and G. Herbison. 1992. Motor evoked potentials in
icant correlation between corticospinal tract conduc- psychogenic paralysis. Neurology 42:935–6.
tion block and prolongation of central motor conduc- 41. Nezu, A., S. Kimura, S. Uehara, T. Kobayashi,
tion time in compressive cervical myelopathy. Journal M. Tanaka, and K. Saito. 1997. Magnetic stimulation
of Neurological Sciences 256(1–2):71–4. of motor cortex in children: Maturity of corticospinal
33. Lo, Y. L., L .L. Chan, W. Lim, et al. 2006. Transcranial pathway and problem of clinical application. Brain
magnetic stimulation screening for cord compression Development 19:176–80.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART D
Assessing the Motor Unit
Contraction of somatic muscles underlies all horn cells; radiculopathies, mononeuropathies,

human movement. Central and peripheral and peripheral neuropathies involve the periph-
motor pathways that control movement ulti- eral axons; and primary myopathies and dis-
mately act on muscle to produce contraction. orders such as myasthenia gravis affect the
The lower motor neuron forms the connection muscles and neuromuscular junctions.
between motor pathways in the central nervous Several electrophysiologic techniques may
system and the muscles, and may be consid- be used to test the integrity of the motor
ered the “final common pathway” of motor units. Motor nerve conduction studies mea-
function. The lower motor neuron is made sure alterations in function of the periph-
up of motor units, each of which consists of eral motor axons, neuromuscular junction,
a cell body, an axon, and all of the muscle or muscle, but only indirectly measure the
fibers innervated by that axon. The cell body extent of axonal destruction or anterior horn
of a motor neuron is located in a cranial nerve cell loss. The electrophysiologic assessments
motor nucleus within the brain stem or the described in the chapters in this section com-
anterior horn of the spinal cord. The peripheral plement nerve conduction studies in defin-
motor axon is a myelinated fiber that travels ing the character, severity, and distribution of
in a cranial nerve or peripheral nerve to the neuromuscular disease. Distinguishing among
muscle, where it branches into multiple nerve primary muscle diseases, disorders of the neu-
terminals. Each terminal branch innervates a romuscular junction, and neurogenic disor-
number of muscle fibers, and the number of ders depends on needle electromyography
terminal branches in the muscle determines (Chapter 26). In addition, needle electromyo-
the innervation ratio (the number of muscle graphy helps characterize the defects of neu-
fibers in a motor unit). Innervation ratios may romuscular transmission and the number of
be as small as 50 (in extraocular muscles and functioning axons or anterior horn cells. These
other small muscles requiring fine control) or two aspects of neuromuscular disease can be
as large as 2000 (in large, powerful muscles quantified more precisely with the special
such as the gastrocnemius). techniques of quantitative EMG (Chapter 27),
Most peripheral neuromuscular diseases single fiber electromyography (Chapter 28),
involve one or more components of a motor and motor unit number estimate (MUNE)
unit. Motor neuron diseases, such as amy- (Chapter 29)—techniques that are discussed in
otrophic lateral sclerosis, involve the anterior this section.
401
Chapter 26
Assessing the Motor Unit with

Needle Electromyography
Devon I. Rubin
INTRODUCTION Single Fiber Electrodes

KNOWLEDGE BASE OF NEEDLE EMG Macroelectrodes
TECHNIQUE OF NEEDLE
SKILLS OF EMG WAVEFORM
EXAMINATION
RECOGNITION
Clinical Evaluation
Pattern Recognition
CONDUCTING THE NEEDLE Semiquantitative EMG
EXAMINATION ORIGIN OF EMG POTENTIALS
Preparing the Patient NORMAL EMG ACTIVITY
Muscle Selection Normal Spontaneous Activity (End
Needle Insertion plate Activity)
Needle Movement Normal Voluntary Activity (Normal MUPs)
Recording Display During Needle Firing Rate and Recruitment of MUPs
Examination MUP Configuration
Data Collection of a Resting Muscle Rise Time
Data Collection from a Contracting Duration and Amplitude
Muscle Phases
Measurement of MUPs Stability
POTENTIAL COMPLICATIONS ABNORMAL SPONTANEOUS
DURING NEEDLE EXAMINATION ELECTRIC ACTIVITY
Anticoagulation or Bleeding Disorders Insertional Activity
Lymphedema and Skin Problems Fibrillation Potentials
Infection Precautions Myotonic Discharges
Examining Peri-Pleural Muscles Complex Repetitive Discharges
Examination of Patients with Fasciculation Potentials
Pacemaker Myokymic Discharges
Obese Patients Neuromyotonic Discharges
Low Pain Tolerance (Neuromyotonia)
Cramp Potentials (Cramp Discharge)
EMG SIGNAL ANALYSIS Synkinesis
NEEDLE ELECTRODE
CHARACTERISTICS ABNORMAL ELECTRICAL
Standard Concentric Electrodes ACTIVITY—VOLUNTARY MUPs
Monopolar Electrodes Abnormal Recruitment
403
Long-Duration MUPs ABNORMAL ELECTRICAL

Short-Duration MUPs ACTIVITY—DISORDERS OF
Polyphasic MUPs CENTRAL CONTROL
Mixed Patterns: Long-Duration Tremor
and Short-Duration Dystonia, Rigidity, Spasticity,
MUPs Stiff-man syndrome
Varying or Unstable
MUPs PATTERNS OF ABNORMALITIES
Doublets (Multiplets) SUMMARY
INTRODUCTION • Complements nerve conduction stud-

ies in the localization of neuromuscular
Needle electromyography (EMG) is one of the disorders.
major diagnostic tools for identifying and char- • Assists in the assessment of the tempo-
acterizing disorders of the motor unit, includ- ral profile and activity of neuromuscular
ing anterior horn cells, peripheral nerves, diseases.
neuromuscular junctions, and muscles. Most • Identification of specific spontaneous dis-
electrodiagnostic studies require needle EMG charges may help to identify a specific eti-
evaluation as a complement to the nerve con- ology or narrow the differential diagnosis.
duction studies to adequately assess for neuro-
muscular disorders. Similar to nerve conduc-
tion studies, needle EMG requires a unique
combination of knowledge and skills. Special- KNOWLEDGE BASE OF
ized knowledge of peripheral nervous system NEEDLE EMG
anatomy, physiology, pathophysiology, diseases,
techniques, electricity, and patient interaction An initial step in learning EMG is to under-
is necessary. The knowledge is familiar to prac- stand the range of information that EMG can
ticing clinicians but requires a detail beyond provide to extend the clinical evaluation of
that of the general neurologist. Successful, patients with suspected neuromuscular dis-
high-quality, reproducible EMG depends on eases.1 Information provided by electrodiag-
the skills of a clinician practiced in physi- nostic testing includes
cally inserting, moving, and recording with a
needle electrode, assessing clinical problems, • Confirming a clinically suspected diagnosis
and on the unique art and skills of analyz- • Excluding other potential mimicking
ing electric signals recorded from muscle using diseases
auditory pattern recognition and semiquantita- • Identifying unrecognized or subclinical
tion.1, 2 Many of these facets of needle EMG disease
depend on the unique analysis and process- • Localizing abnormality or lesion within a
ing capabilities of the human brain and have specific region of the peripheral nervous
precluded the replacement of skilled physician system
electromyographers by automated equipment. • Defining the severity of a disease
This chapter will review the steps and tech- • Defining the pathophysiologic mechanism
niques of needle EMG and waveform analysis, of a disease
and will describe the types of EMG waveforms • Defining the evolution, stage, and progno-
recorded during the needle examination. sis of a disease
Purpose and Role of Needle EMG

The expert clinical electromyographer under-
• Method used to assess the integrity of the stands how EMG can provide each of these
motor unit, including lower motor neuron, types of information and keeps each one in
neuromuscular junction, and muscle. mind when interpreting and reporting the
Assessing the Motor Unit with Needle Electromyography 405
results of EMG testing. Several steps are nec- hip abduction. There is decreased pinprick
essary throughout the EMG evaluation for the sensation over the dorsum of the foot and a
electromyographer to efficiently and accurately reduced left Achilles reflex. On the basis of
provide the appropriate information. These the clinic examination, the primary hypothesis
steps include is an L5 radiculopathy and possible S1 radicu-
lopathy. Alternate hypotheses include peroneal
• Performing a thorough clinical evaluation. mononeuropathy, sciatic neuropathy, or sacral
• Conducting the needle examination, plexopathy.
including: On the basis of these hypothesis, nerve con-
◦ Preparing the patient for the study duction studies selected were the left peroneal
◦ Selecting the appropriate muscles to motor, tibial motor, superficial peroneal sen-
test sory, and sural sensory. Abnormal conduction
◦ Inserting and moving the needle studies were compared with the conduction
electrode studies on the right leg. Decision regarding
◦ Collecting the data muscles to examine during needle EMG are
◦ Recognizing special situations related to discussed below.
the ability to examine muscles.
• Analyzing the recorded activity.
Special attention to aspects of each of these

CONDUCTING THE NEEDLE
steps is necessary during an EMG examination EXAMINATION
to ensure a comfortable, comprehensive, and
reliable study. The ability to efficiently and effectively record
the electric activity from muscle depends con-
siderably on an electromyographer’s skills of
patient interaction and using a needle record-
TECHNIQUE OF NEEDLE ing electrode. These skills are learned by expe-
EXAMINATION rience. Special attention to each aspect of
the needle examination will help to ensure
Clinical Evaluation a more comfortable and reliable study. In
addition, careful attention to several special
A clinical electromyographer’s approach to the problems—skin infection or other cutaneous
evaluation of a patient referred for an EMG conditions, bleeding disorder, cardiac valvular
study is an extension of the type of hypothesis disease, obesity—presented by a few patients
generation and testing that is commonly used must be considered before a needle examina-
in clinical neurology. The electromyographer tion is performed.
must review all of the clinical data, confirm
and obtain additional history as needed, and Preparing the Patient
perform a focused neuromuscular examination
to define the clinical deficits. On the basis of Prior to the study, most patients will have
this information, a set of hypotheses listing the received information about the needle exam-
possible localizations and causes of the clini- ination and may have a few questions. It is
cal problem can be generated and prioritized. helpful to explain briefly that a needle will be
These hypotheses determine which electro- inserted into several muscles and may cause
physiologic tests, including nerve conduction some discomfort. The patient will appreciate
studies and needle EMG, are required to knowing approximately how long the study will
test the hypotheses and distinguish among the take and approximately how many muscles will
potential disorders. be examined.
Case Example. A 52-year-old man presents
with a 6-month history of left leg pain, Muscle Selection
foot weakness, and foot numbness. The clin-
ical examination demonstrates weakness in The muscles to be tested are selected initially
foot dorsiflexion, inversion, eversion, and on the basis of the clinical hypotheses. In
evaluating certain diseases, the distribution of The muscle should be palpated during inter-
findings will often vary between muscles as mittent contraction to localize its borders. The
well as in different regions of the same mus- skin is pulled taut to decrease the pain that
cle. For example, in many myopathies nee- occurs during insertion of the needle through
dle EMG abnormalities are more commonly the skin, and pulled a short distance over the
seen in proximal muscles, but in some etiolo- muscle to reduce bleeding following removal
gies, such as with inflammatory myopathies, of the needle after the study. The needle elec-
the superficial layers of the muscle may show trode should be held firmly in the fingers
more prominent changes than deeper portions. and, after alerting the patient to an immi-
Motor neuron disease may show widely dis- nent stick, the needle is then inserted smoothly
tributed findings and therefore multiple distal and quickly through the skin into the sub-
and proximal limb muscles supplied by dif- cutaneous tissue or superficial layers of the
ferent roots and nerves may be necessary to muscle.
demonstrate a widespread disease of motor
neurons. Some types of motor neuron diseases,
such as Kennedy’s disease, may demonstrate Needle Movement
findings more prominently in cranial mus-
cles. In contrast to generalized disorders, in a During needle EMG, three types of activity
patient with a suspected radiculopathy or sin- are recorded: insertional activity, spontaneous
gle mononeuropathy, the needle examination activity, and voluntary activity. Since the nee-
will be more focused on a single limb. dle electrode primarily records activity from a
Regardless of the distribution of mus- small area in a muscle, the electrode must be
cles being examined, the individual muscles moved to record the activity in several differ-
selected should ideally be superficial, easily ent regions of the muscle in order to obtain
palpated, and readily identified. They should a more complete assessment of the underly-
be located away from major blood vessels, ing changes that may have occurred in the
nerve trunks, and viscera and should be those motor units. The movement of the needle
that cause the least discomfort for the patient. through the muscle is the predominant gen-
For example, testing the thenar or small foot erator of the discomfort experienced during
muscles often makes patients more uncomfort- the examination. To reduce this discomfort,
able than testing other muscles. Hence, these the muscle should be examined by moving the
muscles should be tested only when the infor- needle along a straight line into the muscle
mation is not available from other muscles. in short steps (0.5–1 mm). Large movements
Since the location and method of activation are more painful.3 The pace of needle move-
of muscles and the appearance of motor unit ment should not be rushed. A brief pause
potentials (MUPs) can vary greatly among dif- (2 seconds or longer) between each step is
ferent muscles, the examiner should become needed to listen and watch for slow firing
familiar with how to test each muscle and the abnormal activity, such as fibrillation poten-
range of normal findings within the muscle. tials or fasciculation potentials. The needle is
advanced in 5–30 such steps depending on
Case Example. In the case described earlier, the muscle diameter. After the diameter of the
muscles initially planned to be examined with muscle has been traversed, the needle is with-
needle EMG in the patient with leg pain, foot drawn from the muscle—but not from the
weakness, and foot numbness include ante- skin—and reinserted from a different angle at
rior tibialis, posterior tibialis, medial gastrocne- the same location. Two to four such passes
mius, vastus medialis, tensor fascia lata, gluteus through the muscle are made until an adequate
maximus, and lumbar paraspinals. number of sites in the muscle have been exam-
ined. Adequate control during needle manip-
ulation can only be obtained manually with
Needle Insertion small advances of the needle. The examiner’s
hand should be resting on the patient and
Once the appropriate muscle to be examined the needle should be held firmly and steadily
is identified, the skin is wiped over each punc- in the hand without release throughout the
ture site with alcohol prior to needle insertion. examination.
Recording Display During Needle muscle, distracting the patient with conversa-
Examination tion, and providing continuous verbal feedback
and reassurance.
Each electromyographer develops a prefer- Several types of electrical signals normally
ence for how to display the electric activity; occur in a resting muscle. Insertional activ-
however, certain variables should be familiar to ity is the electric response of the muscle
all examiners because of their common use and to the mechanical damage by a small move-
advantages in certain situations. Oscilloscope ment of the needle (Fig. 26–1). Evaluation of
sweep speeds of 5–10 ms/cm are best for char- insertional activity requires a pause of 0.5–1
acterizing the appearance of motor units, but second or more following cessation of nee-
slower speeds of 50 or 100 ms/cm are help- dle movement to see any repetitive poten-
ful to characterize firing patterns and assess tials that may be activated. Insertional activity
firing rates during recruitment analysis. Ampli- may be increased, decreased, or show specific
fication settings of 50 μV/cm and 200 μV/cm waveforms, such as myotonic discharges. End
are most useful for examining spontaneous plate activity is made up of many, individ-
and voluntary activity, respectively. Filter set- ual, miniature end plate potentials (MEPPs)
tings of approximately 30 and 10,000 Hz or (end plate noise) or action potentials of indi-
more should be used for routine studies. If vidual muscles fibers due to discharge from
formal quantitation of MUPs is to be per- terminal nerve irritation from the examining
formed with comparison of results to those needle tip (see Normal Spontaneous EMG
published by Buchthal, measurements of the Activity).
duration of MUPs should be made with a gain
of 100 μV/cm at a sweep speed of 5 ms/cm
(10 ms/cm if long duration) and with a low filter
frequency of 2–3 Hz. The convention of dis- Data Collection from a
playing negative potentials at the active elec- Contracting Muscle
trode as upward deflections is used in clini-
cal EMG. The majority of the time spent during the nee-
dle examination occurs during the assessment
of MUPs in a contracting muscle. The contract-
ing muscle is best examined with the muscle
Data Collection of a Resting at a level of contraction that activates only
Muscle a few motor units (low to moderate effort).
Selective activation of the muscle of interest
Examination of the muscle at rest is per- may be needed to determine needle position
formed to assess for abnormal spontaneous when examining deep muscles, muscles that
discharges that may be indicators of an are difficult to palpate, or small muscles. The
underlying disease (see Abnormal Sponta-
neous EMG Waveforms). The resting muscle
is tested for spontaneous activity at a gain
of 50 μV/div. When the needle is well within
the muscle, it should not be moved for sev-
eral seconds so the examiner can listen for
fasciculations or slowly firing fibrillation poten-
tials. In some cases, obtaining complete mus-
cle relaxation may be difficult or impossible,
such as in patients experiencing pain, patients
with spasticity or tremor, in children, or in
muscles such as the diaphragm or anal sphinc-
ter. In tense patients or during a painful
examination, relaxation can be enhanced by
Figure 26–1. Top, Normal short burst of insertional
certain techniques, such as positioning the activity with needle movement. Bottom, Increased inser-
muscle in neutral or relaxed position, passively tional activity with a train of repetitive firing potentials after
manipulating the limb, activating an antagonist the insertional burst.
steps in testing contracting muscle include the only at minimal-to-moderate levels of contrac-
following: tion while the needle is advanced through dif-
ferent areas of the muscle. Currently, digital
• Withdraw the needle to a subcutaneous EMG machines have automated the measure-
position prior to voluntary muscle contrac- ments. Two other quantitative EMG programs
tion to reduce bending of the needle and use computer algorithmic template matching,
discomfort. called multiple motor unit potential (multi-
• Position the joint across which the muscle MUP) and decomposition quantitative EMG
acts to limit the activity of synergistic and (DQEMG), allow for assessment of individ-
adjacent muscles. ual MUPs during a stronger contraction where
• Prior to moving the needle into the mus- 3–5 MUPs are firing at one time. Quantitative
cle, ask the patient to hold the limb in EMG is reliable and often needed in question-
a position that requires activation only of able cases to increase the certainty of a diagno-
the muscle being examined. sis. Objective measurements may be a neces-
• Palpate the contracting muscle as a guide sity in recognizing mild diseases, such as an
to needle movement. early neurogenic process or mild myopathies.
• Advance the needle until encountering Details of individual characteristics of MUPs
MUPs with a rapid rise time (identified by cannot be measured reliably during a strong
a sharp clicking sound). voluntary contraction, which normally produces
a dense pattern of multiple superimposed
The MUPs recorded from muscle can be ana- potentials called an interference pattern. Inter-
lyzed in several different ways.4–6 The usual ference pattern analysis summates the effect of
method in clinical studies is to display and recruitment with the duration and amplitude
measure isolated potentials (described below of the potentials and records the number of
under Measurement of Motor Unit Potentials); turns and total amplitude of the electric activity
however, other approaches that analyze the during a fixed time with an automatic count-
entire sequence of waveforms in an interfering device.7 With examination at these stronger
ence pattern when multiple motor units are levels of contraction, less dense patterns may
firing have also been used. Such analyses are occur if there is a loss of motor units, poor
applied almost exclusively to motor unit activ- effort, an upper motor lesion, or if the muscle
ity, but not to spontaneous activity. is powerful. The latter three conditions can be
distinguished from a loss of motor units only
by estimates of firing rates. This method varies
Measurement of MUPs with patient effort, which must be accounted
for in measurements.
The recruitment and appearance of MUPs are Recording EMG by isolation of single MUPs
examined during voluntary activity. Multiple, and by interference pattern provides reliable
different MUPs (a minimum of 20) in different estimates of the electric activity in a muscle.
areas of the muscle must be assessed to obtain Because of the number and variety of normal
a complete assessment of the integrity of the MUPs, both of these methods require multi-
motor units composing that muscle. ple measurements and a statistical description
Measurements may be made in two ways: of the results obtained from different areas of a
by isolation and measurement of a single muscle. The results of these two methods cor-
MUP (quantitative EMG) and by interfer- relate well with each other and with muscle
ence pattern analysis. Quantitative EMG is histology and neither method has been shown
the classic method of measuring an MUP to be superior to the other.
by isolating and recording at least 20 single
potentials and then manually measuring the Key Points
duration, number of phases, and amplitude.
These measurements must be compared with • A thorough clinical evaluation and exam-
the values recorded from the same muscle ination by the physician performing the
in normal subjects of the same age. This electromyogram prior to the study is nec-
method provides no quantitative assessment essary to define the hypotheses and deter-
of recruitment and makes the measurements mine which muscles should be examined.
• The technique used during needle inser- extremely low. There are only a few reports of
tion and movement through the muscle paraspinal hematoma, calf hematoma, and calf
should consist of small needle movements artery pseudoaneurysm development following
with pauses to minimize discomfort. needle examination in patients on anticoagu-
• Brief pauses of one second or longer lation.9–11 In a recent study using ultrasound
are necessary during spontaneous activ- to identify hematoma formation in the ante-
ity assessment to observe for slowly firing rior tibialis muscle following needle EMG in
potentials. patients on anticoagulants and antiplatelets,
• Sensitivities of 50 μV/cm are optimal for only 2 of 100 patients on Coumadin (with
assessment of spontaneous discharges and INRs up to 4.0 assessed) and one of 60
200 μV/cm are optimal for assessment of patients on antiplatelet agents developed small
voluntary MUPs. (2–3 mm wide by 2–3 cm long), subclinical
• Assessment of a resting and mildly con- hematomas.12
tracting muscle is necessary to assess In a survey of 47 EMG laboratories in the
for abnormal spontaneous activity and United States, 9% of laboratories reported
changes in MUPs. experiencing at least one episode of bleed-
• Assessment of at least 20 different MUPs ing complication requiring medical or surgical
is necessary to adequately assess the intervention due to needle EMG.9 However,
changes that may be occurring in the in this same survey, 66% of laboratories indi-
muscle. cated that they were willing to perform EMG
• Single MUP (quantitative EMG) and on all limb muscles in anticoagulated patients,
interference pattern analysis are both while the other laboratories indicated that they
complementary techniques used to assess would limit the needle examination. Further-
the motor units. more, half of the laboratories limited needle
examination of cranial muscles and 28% lim-
ited paraspinal muscle examination.
POTENTIAL COMPLICATIONS At this time, there is no standard of practice
in electrodiagnostic medicine regarding the
DURING NEEDLE EXAMINATION highest level of anticoagulation at which a nee-
dle examination can safely be performed with-
Needle EMG is a safe procedure; however,
out additional risk and no consensus statement
potential complications related to needle inser-
by the American Association of Neuromuscu-
tion and movement through a muscle may
lar and Electrodiagnostic Medicine (AANEM)
rarely occur.8 In special circumstances, limita-
related to the performance of needle EMG
tions in the needle examination or adjustments
in anticoagulated patients. Nonetheless, each
in the examination technique may need to be
case must be examined individually and the
considered to reduce potential risks.
necessity and benefits of the study must be
weighed with the potential risks. In the ideal
situation, anticoagulants should be discontin-
Anticoagulation or Bleeding ued prior to the study, although in most cases,
Disorders this increases the risk of potential thrombotic
complications. Most electromyographers pre-
Needle examination can generally be per- fer to know the level of anticoagulation (INR)
formed without complications in patients on before the study to determine the level of
anticoagulants, antiplatelet agents, or with risk. If the prothrombin time is in the thera-
bleeding complications, although adjustments peutic range (especially low range), the study
in technique and limitations may apply. The can be performed safely in most instances.
risk of performance of the needle examina- If it is above therapeutic range, all or part
tion in patients on anticoagulation is excessive of the needle examination may need to be
bleeding or hematoma formation. If this were deferred. If the needle examination is per-
to occur in a closed compartment, there is formed on patients on anticoagulation, special
the potential for development of compartment attention and adjustment in the technique of
syndrome and tissue necrosis. Despite this the- the study should be made and the physician
oretical risk, the magnitude of the risk is likely should
1. Examine the minimal number of muscles type of valvular heart disease or with pros-
necessary to answer the referring physi- thetic valves is similar to that of repeated
cian’s question. venipuncture and prophylactic antibiotics are
2. Avoid deep muscles, for example, para- not necessary.
spinals and diaphragm.
3. Avoid tight fascial spaces, for example,
tibialis anterior. Examining Peri-Pleural Muscles
4. Avoid muscles in close proximity to arter-
ies, for example, iliopsoas, flexor pollicis Examination of muscles adjacent to or near the
longus. lungs produces a risk of puncturing the pleura
5. Place pressure on the puncture site for and inducing pneumothorax. This may occur
1–2 minutes following the examination. with examination of the diaphragm, rhom-
boids, serratus anterior, trapezius, supraspina-
Similar precautions should be considered in tus, and cervical paraspinals. Experience in
patients with thrombocytopenia. If the platelet examining these muscles and precise knowl-
count is above 30,000/mm,2 the study can usu- edge of the location and anatomy of these
ally be performed safely. For hemophilia and muscles is crucial to prevent this complication.
uncommon bleeding disorders, the patient’s Techniques used to reduce the risk of pneu-
hematologist should be consulted prior to per- mothorax when examining these muscles have
formance of the needle examination. been reviewed.8 In all cases, when any of these
muscles are examined, the needle electrode
should be advanced very slowly and smoothly,
Lymphedema and Skin Problems listening for the sharp, clicky sound of the
MUPs (indicating close proximity). When the
Several dermatologic conditions should lead to sound of the potentials becomes more dulled
avoidance or limitation of the needle exam- with needle advancement, the needle is likely
ination. The needle electrode should not be nearing the distant portion of the muscle and
inserted into an infected area of skin (such as should be withdrawn. Listening for a respi-
one with cellulitis) or in an area of prominent ratory pattern of MUP firing, indicating the
vasculature (such as varicose veins). Addition- approach to the peri-pleural muscles, should
ally, patients with thin skin, such as those on prompt discontinuation of forward movement
corticosteroids, may be more prone to bleeding of the needle. When the needle is close to
or tearing of the skin and extra caution should MUPs with this pattern of firing in deeper mus-
be taken during the examination. cles, caution should be made against further
Examining a limb with lymphedema poses advancement of the needle.
the risk of persistent leaking of serous fluid,
potentially increasing the risk of the develop-
ment of cellulitis. Despite the absence of stud- Examination of Patients with
ies assessing this risk, a position statement by Pacemaker
the AANEM suggested that “reasonable cau-
tion should be exercised in performing needle There is no contraindication to performing the
examinations in lymphedematous regions.”13 needle examination in patients with a pace-
maker or other automated defibrillator. Recog-
nition of pacemaker artifact is important, in
Infection Precautions order to avoid misinterpretation of the artifact
as a fibrillation potential.
Universal precautions should be taken with
every study. In patients with dementia (with
possible Creutzfeldt–Jakob disease), HIV infec- Obese Patients
tion, viral hepatitis, or other transmissible dis-
ease, added precautions to avoid inadvertent Examination of certain muscles may be diffi-
needle stick should be made. The risk of nee- cult in obese patients. Positioning the needle
dle EMG in patients with rheumatic or other electrode at a steeper angle allows for deeper
Table 26–1 Guidelines for Needle Size According to BMI

lb kg Needle size for hip girdle muscles by BMI
353 160
342 155
331 150
320 145
308 140
297 135
286 130
275 125
264 120 120
253 115
242 110
231 105 90
220 100
209 95 75
198 90
187 85
176 80
165 75
154 70
143 65
132 60
cm 145 150 155 160 165 170 175 180 185 190 195 200 205
in 57 59 61 63 65 67 69 71 73 75 77 79 81
BMI<30: use 50mm

needle
BMI>30: use 75mm
needle
BMI>35: use 90mm
needle
BMI>40: use 120mm
needle
penetration through the tissue into the mus- Pain minimization requires attention to all
cle. Deep muscles, such as hip girdle muscles, interactions with the patient, in particular the
may require a 75-, 90-, or 120-mm needle. The techniques of the needle examination itself.
patient’s body mass index (BMI) can be used to Techniques such as distraction, continued reas-
guide the electromyographer in the selection surance, and an empathetic approach to the
of the appropriate needle length (Table 26–1). patient during the study may improve the
Some muscles, such as the deep paraspinal patient tolerance of the study. The technique
muscles, may be difficult to reach, even in of needle movement has a significant impact
average-sized patients, without a long needle. in pain reduction. Studies have demonstrated
Needles up to 120 mm long should be avail- that needle movements of less than 1 mm when
able. Caution should be taken when examining using concentric needle electrodes are signifi-
peri-pleural muscles or muscles neighboring cantly less painful than needle movements of
risky structures. Selective activation of mus- approximately 1 cm.13
cles may be necessary to ensure correct muscle
localization.
Key Points
Low Pain Tolerance • Needle examination can be carried out
safely in most patients on therapeutic anti-
Most patients are able to tolerate the discom- coagulation or antiplatelet agents.
fort of the needle examination without dif- • In anticoagulated patients, bleeding com-
ficulty, but a few need a special approach. plications can be reduced by minimizing
the number of muscles examined, avoid-

ing deep muscles or tight fascial com-
partments, and placing extra pressure on
puncture sites.
• In patients with thrombocytopenia with
platelet counts of less than 30,000/mm2 or
in patients with hemophilia, consultation
with a hematologist should be considered
prior to needle EMG.
• Caution should be used when examin-
ing patients with lymphedema or peri-
pleural muscles to avoid complications,
such as the development of cellulitis or
pneumothorax.
• There is no contraindication to needle
EMG in patients with pacemaker; pace- Figure 26–2. Electrode types used in recording EMG
signals. The recording areas of each electrode type are
maker artifact should not be misinter- shown by the hashed regions.
preted as a fibrillation potential.
electric signals are distorted by intervening
skin, subcutaneous tissue, and other muscles.
EMG SIGNAL ANALYSIS Needle electrodes inserted into the muscle
can depict the electric signals accurately, but
The ultimate goal of the needle examination is depending on needle type, these electrodes
to obtain data that reflects the underlying mor- record from different numbers of muscle fibers
phology of the muscle fibers and motor units, and from muscle fibers in different locations14
and identify the changes that occur with dif- (Fig. 26–2, for color image, see color plates).
ferent diseases. The electrical signals that are Standard concentric and monopolar electrodes
recorded are dependent on a number of fac- are the ones most commonly used for diagnos-
tors, including the electrode type that is used tic clinical EMG.
during the examination and characteristics of
the muscles.
Standard Concentric Electrodes
NEEDLE ELECTRODE A bare, 24- to 26-gauge hollow needle with a
CHARACTERISTICS fine wire down the center is beveled at the
tip to expose an active, oval recording surface
The selection of the type of needle electrode of 125 μm by 580 μm. The electrode is ref-
used in EMG depends on a number of patient- erenced to the shaft of the needle, thereby
related and examiner considerations. Needle canceling unwanted activity from distant, sur-
electrodes must be sterile, sharp, and straight rounding muscle. Although these electrodes
and the recording surface must be absolutely were expensive, inexpensive disposable models
clean. A thin, poorly conducting film on the are now available. The common sizes available
electrode surface can cause a low-voltage, are 25 mm (26 gauge), 50 mm (26 gauge), and
irregular, positive waveform, popping artifact 75 mm (20 gauge). The needle is a detach-
that can be mistaken for end plate noise or able electrode connected to the preamplifier
positive waves. Electric impedance should be by a cable. Because of the narrow gauge, elec-
checked if a break or short is suspected (correct trodes are particularly delicate and need to be
impedance at 60 Hz is 5–20 ). handled carefully. They are most fragile at the
Different types of electrodes have been used junction of the shaft and hub and may bend
to record the electric activity of normal and dis- or break at this location. This electrode type
eased muscles. Surface electrodes can depict has several advantages: (1) its ability to record
the extent of EMG activity and measure the EMG activity with a minimum of interference
conduction velocity in muscle fibers, but the from surrounding muscles, (2) its fixed-size
recording surface, (3) the absence of a separate in a cylinder along the shaft of the needle.
reference electrode, and (4) the extensively This recording summates the activity of many
defined quantitation of the sizes of normal MUPs, which cannot be differentiated from
MUPs for ages and muscles. one another.18 The potential from a single
motor unit is isolated with the help of simul-
taneous recording of potentials from single
Monopolar Electrodes muscle fibers with a 25-μm diameter electrode
halfway along the shaft of the macroelectrode
A Teflon-coated fine needle electrode, usually on a second channel. The second channel is
made of stainless steel, can have a very fine used to identify the firing pattern of a single
gauge and an extremely sharp point. Monopo- motor unit. The electric activity recorded from
lar electrodes consist of a solid 22-gauge to the macroelectrode at the time of the firing of
30-gauge needle with a bare tip approximately a single fiber potential on the small electrode is
500 μm in diameter. These electrodes record averaged over multiple discharges. This results
essentially the same activity as recorded with in an averaged potential from all muscle fibers
standard concentric electrodes, but MUPs are along the macroelectrode, which are inner-
slightly longer in duration and have a higher vated by the same motor unit as the single
amplitude.15 Monopolar electrodes are pre- muscle fiber. Thus, the averaged potential gives
ferred by some electromyographers because an estimate of the activity in a larger portion
they are less expensive and, depending on the of the muscle fibers of the motor unit. Occa-
technique of needle movement, may be less sionally, macroelectrode recordings are able to
uncomfortable for patients.16 identify changes in the whole motor unit that
are not apparent with smaller electrodes.19
Single Fiber Electrodes Key Points

• Several different types of needle elec-
Recordings made with electrodes with small
(25 μm) recording surfaces referenced to the trodes may be used for needle examina-
shaft of the needle with filtering of the low- tion.
• Concentric needle electrodes record less
frequency components focus on a small num-
ber of muscle fibers in the immediate vicinity noise and with a more stable baseline than
of the electrode (see Chapter 28). Single fiber monopolar electrodes.
• The MUPs recorded with a monopo-
EMG needles record from small areas of mus-
cle and cannot be used to characterize the lar electrode are of higher amplitude
size of MUPs. This method has been used pri- and slightly longer duration than those
marily in studying disorders of neuromuscular recorded with a concentric needle
transmission because it can detect variation in electrode.
• Single fibers needles are used predomi-
motor units (jitter between single fiber poten-
tials) not seen with other needle electrodes. nantly to assess for disorders of neuromus-
Single fiber EMG can also be used to quan- cular transmission and the density of fibers
tify the density of muscle fibers in a motor unit in a motor unit.
• Macroelectrodes record from a larger sur-
(fiber density), a measurement closely related
to the percentage of MUPs that are polyphasic face area of the electrode and summate
and the number of turns on the MUP. activity of several motor units.
Macroelectrodes SKILLS OF EMG WAVEFORM

RECOGNITION
A larger needle electrode is the macro needle,
or macroelectrode.17 A macroelectrode record- EMG waveform analysis requires measure-
ing is made from 15 mm of the shaft of a needle ment of multiple different types and parame-
electrode referenced to a surface electrode. ters of waveforms. It would be ideal to have
The macroelectrode records from a large num- formal, quantitative measures of each of the
ber of muscle fibers of multiple motor units variables of the potentials that are assessed
during needle EMG, such as are available Auditory pattern recognition, like visual pat-
for nerve conduction studies. The limitations tern recognition, is so intrinsic to cortical func-
of current EMG equipment and the time tion that once learned it occurs essentially
required to accomplish such measurements instantaneously. The skills of auditory pattern
preclude this for routine EMG. However, a recognition form the basis of learning the
skilled electromyographer can achieve accu- major distinct patterns of firing of EMG dis-
rate and reliable waveform recognition and charges. EMG waveforms fire with distinct pat-
analysis by applying the well-defined skills of terns, and these patterns help to identify and
pattern recognition and semiquantitation. Pat- define the waveform. The patterns of firing of
tern recognition is used to identify and name EMG waveforms are defined by the manner
a waveform while semiquantitation is the skill of change of the interpotential interval of suc-
used to analyze the changes that occur in cessively firing potentials. These patterns can
MUPs with diseases. Both of these skills rely also be considered as the predictability of when
heavily on auditory recognition and analysis. the next potential of a repetitively firing wave-
Similar to learning the technique of the needle form will occur (Fig. 26–3). The different firing
examination, the skills of pattern recognition patterns of EMG waveforms are
and semiquantitation are learned by experi-
ence but can be continually improved and
enhanced throughout the electromyographer’s • Semirhythmic—recurring in orderly, but
career. (The accompanying CD is a useful tool not precise, intervals. The variation in
to assist with learning pattern recognition and the change of interpotential interval is
semiquantitation of EMG waveforms.) approximately 10%. Potentials that fire
in a semirhythmic pattern are voluntary
MUPs.
• Regular (no change or linear change)—
Pattern Recognition recurring at precisely defined intervals
that may be identical, may be changing
A major component of EMG is auditory pat- slowly or rapidly, or may be changing in
tern recognition, a skill that most persons have a linear manner. Regular firing potentials
that allows them to recognize the voice of a include fibrillation potentials and complex
friend and to recognize and name the enor- repetitive discharges (CRDs).
mous range of sounds in the environment. • Regular (exponential change)—recurring
Only a limited number of automated systems at precisely defined intervals that change
have been able to make these distinctions.20 slowly or rapidly in an exponential
Figure 26–3. Firing patterns of EMG potentials.

manner. Potentials firing with this pattern Key Points

include myotonic discharges. • Auditory pattern recognition is the skill
• Irregular—recurring in random intervals
used to identify and name EMG potentials.
with no predictability. Potentials that fire • Assessment of the firing pattern of a dis-
in an irregular pattern include end plate
charge helps to identify the discharge.
spikes, fasciculation potentials, and cramp • Firing pattern is defined by the manner
discharges.
• Burst—groups of discharges firing at one in which the interpotential interval of suc-
cessively firing potentials change.
interval in the burst, with the burst recur- • Potentials with a regular firing pattern
ring at slower intervals. Potentials that
have a defined change in the interpoten-
fire in this pattern include myokymic dis-
tial interval, which may be a fixed interval
charges, hemifacial spasm, and tremor.
or one that changes linearly or exponen-
tially.
• Fibrillation potentials fire in a regular pat-
Semiquantitative EMG tern with a fixed or linear change in firing
rate.
Reliable estimates of numerical values for the • Myotonic discharges fire in a regular pat-
duration, phases, rise time, firing rate and tern with an exponential change in firing
number of active motor unit potentials can be rate.
obtained by estimates based auditory recogni- • CRDs fire in a regular firing pattern with
tion. This skill of semiquantitative EMG (also a fixed firing rate.
called rapid quantitation) requires taking the • Fasciculation potentials and end plate
time to learn the methods and then apply- spikes fire in an irregular firing pattern.
ing them consistently to each recording until • MUPs fire in a semirhythmic firing
the techniques are mastered. Semiquantitation pattern.
(rapid quantitation) takes less time than for- • Semiquantitation allows for rapid assess-
mal quantitative EMG. Semiquantitative EMG ment of the parameters of MUPs.
includes the following steps:
• Patient contracts the muscle to activate

only a few (1–4) MUPs from a single area ORIGIN OF EMG POTENTIALS
of muscle. The MUPs are recorded and
displayed. The electric activity of all of the EMG poten-
• The number of individual MUPs and their tials recorded with a needle electrode in a
individual rates of firing are determined muscle is derived from the action potentials
by auditory recognition. of the muscle fibers that are firing singly or
• The variables of MUPs (rise time, dura- in groups near the electrode.21, 22 The muscle
tion, amplitude, phases, turns, stability) fiber action potentials, and MUPs, normally
are determined by auditory recognition have a triphasic configuration, since the action
for each of the potentials. potentials typically propagate toward and then
• The stored potentials are visually reviewed away from the recording electrode. If a poten-
to assess accuracy in the determination of tial is recorded from a region of a muscle fiber
the parameters. that is unable to generate a negative compo-
• With no change in activation, the steps nent of the potential (e.g., if the membrane
above are repeated in additional areas of of the muscle fiber is damaged), the poten-
muscle (using 0.5-mm movements). tial will be recorded as a large positivity fol-
• Recordings in different areas are repeated lowed by a long low negativity.23 For example,
until a minimum of 30 potentials are fibrillation potentials recorded from damaged
recorded. The findings at each location are areas of muscle fibers are recorded as positive
averaged mentally. waves (Fig. 26–4). The amplitude of the exter-
nally recorded action potential and the rate
With mastery of semiquantitation, the elec- of rise of the positive–negative inflection (rise
tromyographer will be able to estimate each of time) is proportional to the distance between
the parameters with more than 90% accuracy. the muscle fiber and the recording electrode
potentials occur involuntarily at the end plate

zone or in diseased states. For example, muscle
fibers that are not innervated by an axon have
an unstable muscle fiber membrane potential
and fire individually without external stimu-
lation, usually with a regular rhythm. These
Figure 26–4. Simultaneous recording of a single muscle are fibrillation potentials. Normally, muscle
fiber potential from one needle electrode that mechani- fibers are under neural control and fire only
cally initiates the potentials and from another electrode, in response to an end plate potential that
a few millimeters distant, recording from the same mus-
cle fiber. (From Daube, J. R. 1991. AAEM minimonog- reaches threshold. This usually occurs after vol-
raphy #11: Needle examination in clinical electromyo- untary activation and is mediated by central
graphy. Muscle & Nerve 14:685–700. By permission of neural control. These potentials are motor unit
John Wiley & Sons.) potentials.
A variety of normal and abnormal EMG
and both fall off exponentially as the distance
waveforms may be recorded from the mus-
increases.24 Therefore, in order to accurately
cle (Fig. 26–5). EMG waveforms generated
assess the potentials, only those with a rapid
from single muscle fiber action potentials fir-
rise time should be analyzed. The size of single
ing individually include end plate spikes, fib-
fiber action potentials is also related directly to
rillation potentials, and myotonic discharges.
the diameter of the muscle fiber and can par-
Waveforms generated by groups of muscle
tially be used clinically to judge the duration of
fiber action potentials include CRDs, fascicula-
denervation.25 Normally, the size and shape of
tion potentials, myokymic discharges, and neu-
the potentials are constant each time they fire.
romyotonic discharges. Voluntary MUPs are
Action potentials of individual muscle fibers
groups of muscle fibers firing together, linked
may occur spontaneously or may be initi-
as part of the same motor unit firing under
ated by external excitation. Muscle fiber action
voluntary control (Table 26–2).
Figure 26–5. Examples of EMG waveforms recorded from muscle fibers. (From Daube JR. AAEM minimonography
#11: needle examination in clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of the
publisher.)
Table 26–2 Origin of EMG Waveforms

Single fibers MUPs
Firing alone Firing in groups Firing Firing Firing
(adjacent fibers) spontaneously spontaneously voluntarily
(individual) (in bursts) (individual)
End plate spikes Insertional activity Fasciculation Myokymic MUPs

Fibrillation CRDs potentials discharges
potentials Neurotomyotonic Synkinesis
Myotonic discharges Tremor
discharges Myoclonus Hemifacial spasm
Dystonia
Stiff-man
syndrome
CRDs, complex repetitive discharges.
Key Points recorded with needle electrodes as monopha-

sic negative waves that have amplitudes less
• All of the EMG potentials recorded dur-
than 10 μV and durations of 1–3 ms or less.
ing needle EMG are generated from the Individual potentials occur irregularly but usu-
action potentials of single muscle fibers. ally cannot be distinguished. This activity is
• Single muscle fibers action potentials may
usually seen as an irregular baseline called end
fire spontaneously and individually, spon- plate noise and has a typical seashell sound
taneously in groups, or under voluntary (Fig. 26–6).
control. The action potentials of some individual
• Individual muscle fiber action potentials
muscle fibers may be recorded in the end
are typically triphasic. plate region as brief spike discharges called
• When recorded from a region of mus-
end plate spikes.26 End plate spikes are caused
cle that cannot generate a negative
by mechanical activation of a nerve terminal
component, such as a damaged muscle
with secondary discharge of a muscle fiber.
fiber, they have a predominantly positive
They have a rapid irregular firing pattern, often
configuration.
• The amplitude and rise time of a muscle with interspike intervals of less than 50 ms.
Although usually initially negative, end plate
fiber action potential is proportional to the
distance between the muscle fiber and the
electrode.
• The size of a single fiber action potential
is directly related to the diameter of the
muscle fiber.
• MUPs occur under central neural control.
NORMAL EMG ACTIVITY
Normal Spontaneous Activity

(End plate Activity)
Normal muscle fibers show no spontaneous Figure 26–6. Top, Spike form of normal spontaneous
electric activity outside of the end plate region. activity in the end plate region. Middle, End plate noise is
seen as an irregularity in the baseline at a low amplification
In the end plate region, MEPPs occur ran- in addition to the spike form. Bottom, End plate spikes and
domly due to spontaneous release of individual noise are seen as small triphasic potentials at a fast sweep
quanta of acetylcholine. These MEPPs may be speed and high amplification.
spikes may be triphasic or, if the needle elec- be characterized by their firing pattern, fir-
trode has damaged the muscle fibers, may also ing rates (recruitment), and configuration or
be recorded as rapid, irregularly firing positive appearance.
waves. End plate spikes sound like sputtering
fat in a frying pan or slowly ripping Velcro. Key Points
End plate activity is normal, occurs in every
• Voluntary MUPs are the sum of the
individual, and has no clinical significance.
However, since recording from the end plate potentials of the individual muscle fibers
region is usually uncomfortable, identification innervated by a single anterior horn cell
of end plate should prompt repositioning of the that occur in the region of the needle
needle electrode. Additionally, end plate spikes electrode.
• MUPs are characterized by their firing
should not be confused with fibrillation poten-
tials or short-duration MUPs, both of which pattern, firing rates (recruitment), and
fire in a different pattern. configuration or appearance.
• MUPs fire in a semirhythmic pattern.
Key Points
Firing Rate and Recruitment of
• End plate spikes and noise are normal MUPs
spontaneous discharges.
• End plate noise is the recording of the Clinical EMG judges the number of motor
MEPPs. units present in a muscle. The number of
• End plate spikes are the action potentials motor units in a muscle may be considered
of muscle fibers that occur with release in two ways. The first is the total number of
of several quanta of acetylcholine due to motor units that could be fired if the ante-
irritation of a nerve terminal branch. rior horn cell pool received adequate cen-
• End plate spikes and noise fire in an tral nervous system input. This refers to the
irregular pattern and have no clinical actual number of motor units within an indi-
significance. vidual muscle. The second is the actual num-
ber of motor units that are activated when a
patient attempts a voluntary contraction. Both
of these are used to assess the presence or
Normal Voluntary Activity absence of disease involving the lower motor
(Normal MUPs) neuron although the second is quite variable
and changes with the patient’s cooperation, the
All voluntary muscle activity is mediated by strength of the muscle, pain, and the pres-
lower motor neurons and the muscle fibers ence or absence of disease of the upper motor
they innervate (motor units) and is recorded neuron.
electrically as motor unit potentials. All MUPs Judgment of the number of motor units
under voluntary control fire in a semirhyth- within a muscle can be performed by assess-
mic pattern and at a relatively constant fre- ing MUP recruitment—which is defined as
quency, although this frequency continuously the initiation of the firing of additional motor
changes as the voluntary activation increases or units as the rate of discharge of the active
decreases. The MUP is the sum of the poten- MUP increases. Recruitment can be assessed
tials of the individual muscle fibers innervated by comparing the rate of firing of single units
by a single anterior horn cell. Since the mus- with the total number of motor units that are
cle fibers in the region of the needle electrode firing. In most normal muscles, motor units
discharge in near synchrony and as a result of initiate firing rates at approximately 5–8 Hz
factors such as proximity of the fibers to the and gradually increase up to 20–40 Hz as the
electrode, length of the innervating terminal effort exerted by the patient increases.27 The
axon, and conduction time along the muscle rate of firing is used as a gauge of the inten-
fiber, the MUP has a more complex configu- sity of excitation of the anterior horn cell by
ration of higher amplitude and longer duration the central nervous system. As the firing rates
than a single fiber action potential. MUPs may increase, additional motor units begin to fire
(are recruited). The determination of the rate motor units than full interference pattern anal-
of firing is one of the more difficult steps to ysis. However, since there may be selective loss
master in standard EMG because of difficulty of higher threshold motor units, recruitment
in obtaining sufficient control by the patient of analysis should include levels of effort associ-
motor units to isolate one or two units. Slow ated with firing rates in the range of 15 Hz.
firing is a term referring to individual MUPs Recruitment may be defined as normal,
that fire at rates slower than 10 Hz and rapid reduced (sometimes referred to reduced num-
firing refers to individual MUPs firing faster bers or discrete firing), rapid (sometimes
than 12 Hz. When possible, the rate of firing referred to as full recruitment), or poor
of the motor unit initially activated is mea- activation.
sured at the time the second unit begins to fire.
In most muscles, this occurs at 6–8 Hz. Nor- • Normal recruitment—the pattern of
mally, recruitment of additional MUPs occurs recruitment is normal for that muscle,
at low levels of effort and at slow rates of firing with an adequate number of MUPs being
(Fig. 26–7). recruited for the frequency of firing
Recruitment can be characterized by recruit- present. If maximal effort can be obtained,
ment frequency, which is the frequency of fir- a full interference pattern is seen, but
ing of a unit when the next unit is recruited individual motor unit firing rates of 15 Hz
(begins to discharge). This is a function of the are sufficient for recruitment analysis.
number of units capable of firing and is usually • Reduced recruitment—a higher recruit-
between 7 and 10 Hz for motor units in normal ment frequency or a smaller number of
limb muscles and up to 16 Hz for motor units MUPs recruited for firing rates in the
in cranial muscles during mild contraction.28 range of 15 Hz than expected for that mus-
Recruitment frequencies vary in different muscle. Reduced recruitment is characteristic
cles and for different types of motor units. of neurogenic disorders in which axonal
Recruitment may also be characterized by the loss on conduction block is the patho-
ratio of the rate of firing of the individual motor physiologic mechanism. In patients with
units to the number that are active. For most severe or end-stage myopathic disorders,
normal limb muscles, this ratio averages less reduced recruitment may also occur due
than 5 and therefore there will be two MUPs to the loss of all muscle fibers within a
firing if one of them is firing at 10 Hz, three at motor unit. This term should not be used
15 Hz, and four at 20 Hz. Therefore, the ratio to describe the condition of patients in
of the number of units firing to the rate of fir- whom relatively few MUPs fire because of
ing can provide a rough gauge of the number pain, strong muscles, upper motor neuron
of motor units. lesions, or poor cooperation. In these sit-
In the presence of lower motor neuron dis- uations, few potentials are fired although
eases, where the number of motor units in a they fire slowly with a normal pattern of
muscle is decreased from axonal loss, or in recruitment (i.e., poor activation).
disorders characterized by conduction block, • Poor activation—a normal recruitment
recruitment frequency increases and, there- pattern and normal recruitment fre-
fore, MUPs fire more rapidly before additional quency, but with relatively few motor
motor units are recruited. Conversely, the rate potentials firing. These potentials fire
of firing of those MUPs already firing will be slowly, but recruitment of additional
unduly fast for the number of MUPs that have potentials is normal. This occurs with
been activated. Or, less commonly, the first upper motor neuron disorders, poor coop-
unit begins firing at a higher rate than nor- eration by the patient, pain, excessively
mal (more than 10 Hz). If the ratio is greater strong muscle, or two-joint muscles, such
than 5 (e.g., 2 units firing at 16 Hz), there is as the gastrocnemius. It is not evidence of
virtually always some decrease in the number lower motor neuron disease.
of motor units. Thus, firing rate of MUPs is • Rapid recruitment—increased number of
an important measure of the loss of axons.29 motor units relative to the force of con-
This semiquantitative method of determining traction. With this type of recruitment,
reduced recruitment provides a more accurate the occurrence of large numbers of MUPs
and reproducible estimate of the number of with normal recruitment frequencies and
Figure 26–7. MUP firing under voluntary control showing minimal reduction in recruitment in an extensor carpi radialis
muscle with normal strength. Top, Two motor units (A and B) initially fire at 5 and 6 Hz. Middle, With increased voluntary
effort, firing rate of A and B increases to 8 and 9 Hz, with recruitment of a third unit (C). Bottom, With greater effort,
the rates increase to 10 and 11 Hz, with no additional nearby units recruited. Only a small, distant unit begins firing at
7 Hz (D). (From Daube, J. R. 2000. Electrodiagnostic studies in amyotrophic lateral sclerosis and other motor neuron
disorders. Muscle & Nerve 23:1488–502. By permission of John Wiley & Sons.) (From Daube JR. AAEM minimonography
publisher.)
420
normal patterns of recruitment occur with the total number of motor units that are
minimal patient effort. This must be firing; for most normal limb muscles, this
graded in proportion to the force exerted, ratio averages less than 5.
because the patterns of firing are entirely • Loss of motor units leads to reduced
normal. It is the only estimate described recruitment (few MUPs firing at fast rates)
that requires consideration of the force and is most commonly seen in neurogenic
exerted by the muscle. It is evidence of disorders.
disease involving the muscle directly. • Identification of rapid recruitment requi-
res consideration of the force exerted by
Although recruitment analysis is reasonably the patient.
reproducible and clinically reliable, it is usually • Poor activation is seen as a normal recruit-
a subjective judgment made by electromyog- ment pattern and frequency, but with rel-
raphers on the basis of experience. It requires atively few MUPs firing. This occurs with
taking into account differences in recruitment upper motor neuron disorders, poor coop-
in different areas of individual muscles and eration by the patient, pain, or an exces-
the even greater differences among differ- sively strong or two-joint muscle, such as
ent muscles. Automated methods for formally the gastrocnemius.
quantitating the recruitment pattern have been
developed.30 In automated studies, individual
MUPs were isolated in human muscles under
voluntary control in an experimental setting. MUP Configuration
The interpotential interval (the inverse of fre-
quency of firing) was determined for a pop- An MUP is also characterized by its appear-
ulation of normal subjects and for patients ance, including duration, amplitude, number
with ALS. The normal onset frequency in the of turns, area, and rate of rise of the fast
biceps muscle ranged from 6 to 8 Hz, with component (rise time) (Fig. 26–8). Each of
the recruitment frequency of the second motor these characteristics has multiple determi-
unit at 7–12 Hz. In patients with ALS, the onset nants, including technical, physiologic, and
frequency was from 8 to 20 Hz, with recruit- pathologic factors. Technical factors that have
ment frequencies of 12–50 Hz. These studies a major influence on the appearance of MUPs
provided quantitative measures of motor unit include the type of needle electrode used to
number estimate. Formal quantitative mea- record the potentials, the area of exposed sur-
sures can provide evidence of the reliability of face of the active leads of the electrode, the
the clinical methods; however, they are so time characteristics of the metal recording surfaces,
consuming and complex that they have not and the electric characteristics of the cables,
been applied clinically (see Chapter 27). Fur-
ther studies and technical developments may
eventually allow recruitment analysis to pro-
vide more accurate estimates of the number of
motor units in a muscle.
Key Points
• MUP recruitment judges the number of
motor units in the muscle, and is defined
by the ratio of the rate of firing of the
potentials to the number of potentials
firing.
• In most normal muscles, motor units ini-
tiate firing rates at approximately 5–8 Hz
and gradually increase up to 20–40 Hz as
the effort exerted by the patient increases.
• Recruitment can be assessed by compar- Figure 26–8. Schema of MUP showing characteristics
ing the rate of firing of single units with that can be measured.
If these technical and physiologic factors

are controlled, the normal anatomical and his-
tologic features of the motor unit and any
pathologic changes that may affect these fea-
tures will determine the characteristics of the
MUPs. The anatomical and histologic features
include innervation ratio (number of mus-
cle fibers in the motor unit), fiber density
(number of muscle fibers per given cross-
sectional area), the distance of the needle
tip from the muscle fibers and from the end
Figure 26–9. Schema of variation of MUP morphology
with electrode position. plate region, and the direction of the axis
of the muscle fiber. The characteristics of
the action potentials generated by individual
preamplifier, and amplifier. The appearance of muscle fibers depend on muscle fiber mem-
an MUP from one motor unit also varies with brane resistance and capacitance (which may
electrode position, since only a small propor- be affected by the amount of connective tissue,
tion of the fibers in a motor unit are near the blood vessels, and fat between the electrode
electrode and those at a distance contribute and muscle fibers), intracellular and extracel-
little to the recorded MUP (Fig. 26–9, for lular ionic concentrations, muscle fiber diam-
color image, see color plates). Thus, no single eter, and conduction velocity. The synchrony
MUP characterizes a motor unit but rather the of firing of the muscle fibers in a motor unit
characteristics of multiple MUP recorded from depends on the length, diameter, and con-
different sites allow for the optimal assess- duction velocity of the nerve terminals, the
ment of the morphology of the motor unit diameter of the muscle fibers, and the relative
(Fig. 26–10). The appearance of MUPs also location of the end plates on the muscle fibers.
changes with several normal physiologic vari- The firing characteristics of the motor unit
ables, including the subject’s age, the muscle depend on the amount of overlap with other
being studied, the location of the needle in the motor units, the number of motor units in
muscle, the manner of activation of the poten- the muscle (or per given area), the differential
tials (minimal voluntary contraction, maximal response to sources of activation (monosynap-
voluntary contraction, reflex activation, or elec- tic, local spinal cord, higher centers), and the
tric stimulation), and the temperature of the rates and patterns of discharge of the anterior
muscle.31, 32 horn cell.
Figure 26–10. Normal MUPs in the biceps muscle showing the variety of configurations that can be seen. (From Daube,
J. R. 1991. AAEM minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700.
By permission of John Wiley & Sons.)
Rise Time plus one (Fig. 26–8). The configuration of an

MUP may be monophasic, biphasic, or tripha-
The rise time is the duration of the rapid sic or it may have multiple phases. The config-
positive–negative inflection and is a function uration depends on the synchrony of firing of
of the distance of the muscle fibers from the the muscle fibers in the region of the electrode.
electrode (Fig. 26–8). It is less than 500 μs if Usually, only a small proportion of MUPs have
the electrode is near muscle fibers in the active more than four phases; those that do are called
motor unit. When the needle electrode is more polyphasic potentials. The percentage varies
distant to the muscle fibers in the motor unit, with the muscle being tested and the age of
the rise time will increase and the amplitude the patient, but is usually no more than 15%
and duration of the motor unit may decrease. of MUPs in most muscles. A late spike, distinct
As a result, only MUPs that are near the elec- from the main potential, that is time locked to
trode, with a rise time of 0.5 ms or less, should the main potential is called a satellite potential
be analyzed. (Fig. 26–8). The satellite potential is generated
by a muscle fiber in a motor unit that has a
long nerve terminal, narrow diameter, or dis-
Duration and Amplitude tant end plate region. If an MUP is recorded
from damaged muscle fibers or from the end of
The duration of the MUP is the time from the the muscle fibers it may have the configuration
initial deflection away from baseline to the final of a positive wave with low, long, late negativity
return to baseline (Fig. 26–8). It varies with the phases.
muscle, muscle temperature, and the patient’s
age. The duration is the parameter which most
accurately reflects the area of the motor unit.
The amplitude of the potential is the maximal Stability
peak-to-peak amplitude of the main spike of
the potential and varies with the size and den- Variability of an MUP is any change in its con-
sity of the muscle fibers in the region of the figuration, amplitude, or both in the absence
recording electrode and with their synchrony of movement of the recording electrode as the
of firing. The amplitude typically consists of the motor unit fires repetitively. Normally, MUPs
action potentials of a few muscle fibers within are stable and appear identical each time
the motor unit that are closest to the record- they fire. In disorders affecting neuromuscu-
ing tip of the electrode. It also differs with the lar transmission, variation of the potential may
muscle, muscle temperature, and the patient’s occur.
age. Decreased muscle temperature produces
higher amplitude and longer duration MUPs.
Key Points
The polarity of all potentials recorded on
needle EMG depends on recording the poten- • Motor unit configurational changes occur
tial with the active (G1 amplifier input) elec- when there has been loss or remodel-
trode. In most cases, where the potential is ing of the motor unit, in myopathic or
recorded by the active central recording wire neurogenic disorders.
at the tip of the electrode, MUPs will appear • Duration and amplitude of the MUPs are
predominantly negative. However, if an MUP the parameters which best reflect the area
is recorded with the shaft of a standard con- of the MUP.
centric electrode or with the reference of • Motor unit variation indicates unsta-
a monopolar electrode, it will be displayed ble neuromuscular transmission and can
as an inverted triphasic potential (apparently occur in neuromuscular junction disorders
negative–positive–negative). as well as in progressing neurogenic and
myopathic diseases.
• Technical factors that have an influence
Phases on the appearance of MUPs include the
type of needle electrode used, the area of
A phase of an MUP can be defined as the num- exposed surface of the active leads of the
ber of times the potential crosses the baseline electrode, the characteristics of the metal
recording surfaces, and the electric char- be altered or there may be abnormal pat-
acteristics of the cables, preamplifier, and terns of activation, as in tremor and synki-
amplifier. nesis. All of these abnormal EMG discharges
• Several normal physiologic variables affect are recognized most accurately and reliably by
the configuration of the MUPs, includ- auditory pattern recognition and semiquantita-
ing the subject’s age, the muscle being tion. The trained ear of an electromyographer
studied, the location of the needle in can define the discharge frequency, rise time,
the muscle, the degree of activation of duration, and number of turns/phases of EMG
the potentials, and the temperature of the potentials. The techniques of pattern recogni-
muscle. tion and semiquantitation have been described
• A short rise time (typically less than previously (see Skills of EMG Waveform
500 μs) indicates recording close to the Recognition).
muscle fibers in the motor unit.
• MUP duration is measured from initial
baseline deflection to final return to base- Insertional Activity
line, and varies with muscle, temperature,
and age. Insertional activity is the electrical activity that
• MUP amplitude consists of the action occurs with mechanical depolarization of the
potentials of a few muscle fibers within the muscle fibers due to needle insertion and
MUP. movement through the muscle. Insertional
• Only up to 15% of normal MUPs in a activity is generated by single muscle fiber
muscle should be polyphasic (>4 phases). action potentials and is composed of combina-
• Normal MUPs have a stable configuration tions of positive and negative spikes depending
each time they fire. on the site of origin of the generated action
potential.33 In a normal muscle, the burst of
insertional activity reflects the number of mus-
cle fibers that depolarize due to mechanical
ABNORMAL SPONTANEOUS irritation; with larger needle movements, the
ELECTRIC ACTIVITY length of the bursts of insertional activity is
longer, and with smaller needle movements
Neuromuscular diseases are best described by the length is shorter. Regardless of the length
a combination of clinical findings, histologic of the insertional bursts, the activity ceases
changes, and the pattern of abnormal find- almost immediately following cessation of nee-
ings on needle EMG. Needle EMG findings dle movement.
are combinations of different specific types of Insertional activity may be increased
abnormal electric waveforms described in the (Fig. 26–11) or reduced from the brief burst
following sections. The clinical electromyogra- that occurs in normal subjects. Increased inser-
pher must recognize specific discharges and tional activity may occur as two types of
know what diseases are associated with them. normal variants, as a result of denervated mus-
In most cases, a specific discharge may be cle, or associated with myotonic discharges.
associated with several different diseases. The The normal variants are recognized by their
following discussion describes the types of widespread distribution, most often occurring
abnormal electrical activity recorded with a in younger, muscular persons, especially in
needle electrode and the diseases associated their calf muscles. One normal variant is com-
with them. Neuromuscular diseases may show posed of short trains of regularly firing positive
abnormal spontaneous discharges, abnormal waves. Some patients with this type of dif-
voluntary MUPs, or both. Abnormal sponta- fuse increased insertional activity have been
neous activity includes fibrillation potentials, found to have mutations in the CLCN1 gene
fasciculation potentials, myotonic discharges, associated with myotonia congenita.34 The sec-
CRDs, myokymic discharges, cramps, and neu- ond type is characterized by short recurrent
romyotonic discharges. MUPs may have an bursts of irregularly firing potentials, some-
abnormal duration and amplitude, abnormal times termed snap, crackle, pop.
number of phases, or vary in morphology. The Increased insertional activity may be the
recruitment pattern of the potentials may also initial early sign, within the first 2–3 weeks,
Figure 26–11. Increased insertional activity. (From Daube JR. AAEM minimonography #11: needle examination in
clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
of denervation following an acute neurogenic acute neurogenic disorder, such as an

disorder, such as an early radiculopathy or early radiculopathy or mononeuropathy.
mononeuropathy. This often occurs prior to • Rarely, benign forms of insertional activ-
the development of more sustained fibrillation ity, such as snap, crackle, pop, may be seen
potentials. In addition, since needle movement in young, muscular patients.
often leads to the generation of fibrillation • Diffuse increased insertional activity of
potentials in denervated muscle, most muscles the positive waveform type may be seen in
that demonstrate more sustained fibrillation patients with the CLCN1 gene associated
potentials have increased insertional activity. with myotonia congenita.
In rare cases, decreased insertional activity • Decreased insertional activity may occur
may occur when the muscle fibers are unable to when muscle fibers are unable to produce
produce action potentials in response to mem- action potentials, such as with end-stage
brane irritation. This most commonly occurs in neuromuscular diseases where the muscle
severe or end-stage neurogenic or myopathic is replaced by connective tissue or fat or
disorders where the muscle is completely with disorders such as periodic paralysis or
atrophic or has been replaced by connective metabolic myopathies.
tissue or fat. Additionally, disorders of mus-
cle membrane dysfunction, such as periodic
paralysis (during paralysis) or myophosphory- Fibrillation Potentials
lase deficiency myopathy (McArdle’s) (during
a contracture), may demonstrate decreased Fibrillation potentials are the action potentials
insertional activity or electrical silence during of single muscle fibers that are twitching spon-
needle movement through the muscle. taneously in the absence of innervation. These
potentials typically fire in a regular pattern at
Key Points rates of 0.5–15 Hz (Fig. 26–12). Infrequently,
they may be intermittent or irregular, par-
• Insertional activity is the electrical activity ticularly early after a denervating process; in
that occurs with mechanical depolariza- these cases, the interspike interval is longer
tion of the muscle fibers due to needle than 70 ms, distinguishing them from end
movement through the muscle. plate spikes. Fibrillation potentials have one of
• Insertional activity reflects the number two forms, either a brief spike or a positive
of muscle fibers that depolarize due to wave.35, 36 (Fig. 26–13) Fibrillation potentials
mechanical irritation; with larger needle that occur as brief spikes (spike form) may be
movements, the length of the bursts of triphasic or biphasic, 1–5 ms in duration, and
insertional activity is longer, and with 20–200 μV in amplitude, with an initial posi-
smaller needle movements the length is tivity or negativity (when recorded at the site
shorter. of origin). Fibrillation potentials that occur as
• Increased insertional activity may be the positive waves (positive waveform) are often of
initial early sign of denervation from an longer duration (10–30 ms) and biphasic, with
50 μV
C –
+
0.001″
Figure 26–12. Fibrillation potentials. A, Spike form. B, Positive waveform. C, Development of a positive waveform from a
spike form (serial photographs taken after insertion of needle electrode). (From Daube, J. R. 1991. AAEM minimonography
#11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of John Wiley & Sons.)
the action potential cannot propagate along

the muscle fiber past the recording electrode.
Rarely, fibrillation potentials are observed to
transform from a spike to a positive waveform
or vice versa; even less frequently, two fibril-
lation potentials are time-locked.37 The ampli-
tude of a fibrillation potential is variable and is
proportional to the muscle fiber diameter. In
diseases with muscle fiber atrophy, fibrillation
potentials may have a low amplitude whereas
in hypertrophic muscle fibers, the amplitude
may be high.38 As a result of the range of
sizes of fibrillation potentials, the configuration
alone cannot be used to identify them. Both
spike and positive waveform fibrillation poten-
tials are recognized as fibrillation potentials by
their slow, regular firing pattern, which sounds
like the “ticking or tocking of a clock.” Both
Figure 26–13. Spike form and positive waveform fibril- forms have the same significance, indicating a
lation potentials. (From Daube JR. AAEM minimonogra- denervated muscle fiber.
phy #11: needle examination in clinical electromyography. Fibrillation potentials occur in any muscle
Muscle & Nerve. 1991;14:685–700. Used with permission
of the publisher.)
fiber that is not innervated, due to neuro-
genic or myopathic processes (Table 26–3).
These potentials may occur in muscle fibers
an initial sharp positivity followed by a long- that (1) have lost their innervation, (2) have
duration negative phase. The morphologic dif- been sectioned transversely or divided longitu-
ference of the two forms reflects the site of dinally, (3) are regenerating, and (4) have never
the initiation of the fibrillation potential along been innervated. In neurogenic disorders,
the muscle fiber relative to the site of the such as radiculopathies, mononeuropathies, or
needle electrode. The positive waveforms are motor neuron disease, loss or degeneration of
muscle fiber action potentials recorded from axons leads to denervated muscle fibers. In
an injured portion of the muscle fiber, when contrast, in myopathic diseases that produce
Table 26–3 Diseases Associated with Fibrillation Potentials

Lower motor neuron diseases Anterior horn cell diseases (e.g., ALS)
Polyradiculopathies
Radiculopathies
Plexopathies
Peripheral neuropathies, especially axonal
Mononeuropathies
Neuromuscular junction diseases Myasthenia gravis, severe
Botulinum intoxication
Lambert–Eaton myasthenic syndrome, severe
Myopathies Inflammatory (e.g., polymyositis, dermatomyositis, inclusion body
myositis)
Infiltrative (e.g., sarcoidosis, amyloid)
Muscular dystrophies (e.g., Duchenne, Becker, limb-girdle)
Myotonic dystrophy
Toxic myopathies (e.g., lipid-lowering agents, chloroquine)
Metabolic myopathies (e.g., acid maltase)
Congenital myopathies (e.g., myotubular, late onset rod myopathy)
Infectious myopathy (e.g., viral myositis, trichinosis)
Other Muscle trauma and rhabdomyolysis
pathologic changes of muscle fiber necrosis, graded from 1+ (few fibrillation potentials in
fiber splitting, functional denervation of indi- most areas of the muscle) to 4+ (profuse fib-
vidual or segments of muscle fibers occurs as rillations filling the free-running baseline in all
the fiber becomes separated from the end plate areas) (Fig. 26–14).
zone. In myopathies, fibrillation potentials are Other forms of electric activity could poten-
often of low amplitude and have a slow firing tially be mistaken for fibrillation potentials.
rate (e.g., 0.5 Hz). The density of fibrillation These include the spontaneous activity in the
potentials is a rough estimate of the number region of the end plate (end plate noise and
of denervated muscle fibers and is commonly end plate spikes), short-duration MUPs, and
Figure 26–14. Fibrillation potentials in denervated muscle. Grades of activity: 1, fibrillation potentials persistent in at
least two areas; 2, moderate number of persistent fibrillation potentials in three or more areas; 3, large number of persistent
discharges in all areas; and 4, profuse, widespread, persistent discharges that fill the baseline. (From Daube, J. R. 1991.
AAEM minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission
of John Wiley & Sons.)
MUPs with a positive configuration. While demonstrate a more rapid rate of change in fir-
the configuration of these waveforms may be ing frequency and amplitude, may also occur.39
identical to fibrillation potentials, all of them Myotonic discharges occur as brief spikes
are distinguished from fibrillation potentials by or positive waveforms, depending on the rela-
their firing patterns, none of which fire in a tion of the recording electrode to the muscle
regular pattern like a fibrillation potential. fiber. When initiated by insertion of the nee-
dle, myotonic potentials have the configuration
Key Points of a positive wave, with an initial sharp positiv-
ity followed by a long-duration negative com-
• Fibrillation potentials are the action ponent. Both amplitude and frequency may
potentials of single muscle fibers that are increase or decrease as the discharge con-
twitching spontaneously in the absence of tinues (Fig. 26–15). Myotonic discharges that
innervation. occur after a voluntary contraction are brief,
• Fibrillation potentials fire in a regular pat- biphasic or triphasic, initially positive spikes of
tern at rates of 0.5–15 Hz. 20–300 μV that resemble the spikes of fibril-
• There are two forms of fibrillation poten- lation potentials. They wax and wane, similar
tials: spike form and positive waveform. to mechanically induced myotonic discharges.
Both forms have the same significance. This afterdischarge corresponds to the clini-
• Fibrillation potentials occur in neurogenic cally evident poor relaxation. The degree of
disorders and myopathies, in which the waxing and waning has been shown to dif-
muscle fibers have lost their innerva- fer between different forms of myotonic dys-
tion, have been sectioned transversely or trophy (DM1 and DM2). In DM1, myotonic
divided longitudinally, are regenerating, discharges typically wax and wane (increase
or have never been innervated. and then decrease in firing rate), whereas in
• The density of fibrillation potentials DM2 (previously known as proximal myotonic
reflects the number of denervated muscle myopathy or PROMM), the discharges more
fibers. commonly wane in frequency.40
Disorders associated with myotonic dis-
charges are listed in Table 26–4. Myotonic
Myotonic Discharges discharges may occur in disorders with or
without associated clinical myotonia. In those
Myotonic discharges are the action potentials with clinical myotonia, the myotonic discharges
of single muscle fibers that are firing spon- are often prominent and frequent. Most com-
taneously in a prolonged fashion after exter- monly, these occur in myotonic dystrophy
nal excitation. The potentials wax and wane types 1 and 2 (DM1 and DM2) or myoto-
in amplitude and frequency because of an nia congenita. The severity of myotonic dis-
abnormality in the membrane of the mus- charges and the presence of waxing and
cle fiber. Myotonic discharges are regular in waning discharges have been shown to be
rhythm, but the firing rates vary exponen- correlated with muscle weakness in DM1,
tially in frequency between 40 and 100 Hz, but not in DM2.40 In a study compar-
which makes them sound like a dive-bomber. ing the abundance of myotonic discharges
Slowly firing myotonic discharges, which bear in patients with sodium and chloride chan-
some resemblance to fibrillation potentials but nelopathies, including myotonia congenita,
Figure 26–15. Myotonic discharge. (From Daube JR. AAEM minimonography #11: needle examination in clinical electro
myography. Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
Table 26–4 Diseases Associated with Myotonic Discharges

Myopathies with clinical myotonia Myotonic dystrophy type 1 and 2 (DM1, DM2)
Myotonia congenita
Paramyotonia congenita
Myopathies without clinical myotonia Hyperkalemic periodic paralysis
Polymyositis
Acid maltase deficiency
Statin-associated myopathy
Toxic myopathies (e.g., colchicine myopathy)
Potassium sensitive myotonia
Centronuclear myopathy
Hypothyroid myopathy
Amyloid
Neurogenic disorders Severe axonal disorders (e.g., peripheral neuropathies,
radiculopathies)
paramyotonia congenita, and hyperkalemic high-frequency potentials, or pseudomyotonic

periodic paralysis, no difference in the degree discharges, is the action potential of groups
of myotonic discharges was found between the of muscle fibers discharging spontaneously in
diseases.41 Rarely, briefer and less prominent near synchrony in a regular, repetitive fashion.
myotonic discharges may occur with fibrilla- The groups of muscle fibers arise from several
tion potentials in chronic denervating disor- different neighboring motor units rather than
ders and with some medications. They are less from the same motor unit. Standard and sin-
readily elicited in a muscle that has just been gle fiber EMG recordings suggest that they are
active than a resting muscle, which is equiva- the result of ephaptic activation of groups of
lent to the warm-up phenomenon that occurs adjacent muscle fibers.42 A CRD is initiated by
in patients with myotonic myopathies. the spontaneous firing of a single muscle fiber
action potential; however, that action potential
Key Points ephaptically spreads and depolarizes a neigh-
• Myotonic discharges are the action poten- boring muscle fiber. Subsequently, a variable
tials of single muscle fibers that are firing number of neighboring muscle fibers may be
spontaneously in a prolonged fashion after depolarized in sequence until the circuit is
external excitation. complete, whereby the initial muscle fiber dis-
• Myotonic discharges fire in a regular pat- charges again. Therefore, each spike within a
tern with exponentially varying firing rates group in a CRD is composed of individual
of 40 and 100 Hz, which makes them muscle fibers action potentials from fibers that
sound like a dive-bomber. may be part of a different motor unit, but lie
• Myotonic discharges are most commonly adjacent to one another.
seen in myotonic myopathies, such as CRDs fire in a regular pattern, character-
myotonic dystrophy or myotonia con- istically with an abrupt onset and cessation.
genita, but can also be seen less promi- During the discharge, they may have sudden
nently in chronic neurogenic disorders. changes in their configuration or firing rates.
• In DM1, myotonic discharges typically The frequency is uniform, ranging from as slow
wax and wane (increase and then decrease as 3 to up to 40 Hz (Fig. 26–16). Although their
in firing rate), whereas in DM2 the form is variable, it typically is polyphasic, with
discharges more commonly wane in 3–10 spike components with amplitudes from
frequency. 50 to 500 μV and durations of up to 50 ms.
CRDs sound like a motor boat that misfires or
a jackhammer.
Complex Repetitive Discharges CRDs are nonspecific in significance but
occur in neurogenic and myopathic disor-
Complex repetitive discharge (CRD), previ- ders that are chronic or longstanding in
ously referred to as bizarre repetitive potentials, nature (Table 26–5). Commonly, these include
Figure 26–16. Two examples of CRDs recurring at 30–40 per second. (From Daube, J. R. 1991. AAEM minimonography
#11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of John Wiley & Sons.)
Table 26–5 Disorders Associated with Complex Repetitive Discharges

Neurogenic disorders Chronic anterior horn cell diseases (e.g., ALS, spinal muscular
atrophy, polio)
Chronic radiculopathies
Chronic axonal neuropathies (e.g., CMT)
Myopathies Chronic inflammatory myopathies (e.g., polymyositis, inclusion
body myositis)
Muscular dystrophies
Hypothyroid myopathy
Schwartz–Jampel syndrome
Normal muscles Iliopsoas
Biceps
old or chronic radiculopathies, peripheral • CRDs fire in a regular pattern with an

neuropathies, or slowly progressive myopathies. abrupt onset and cessation, at rates of
In rare cases of patients with chronic S1 radicu- 3–40 Hz.
lopathies associated with pain and calf hyper- • CRDs occur in chronic neurogenic or
trophy, CRDs are seen in the gastrocnemius in myopathic disorders.
approximately 50%, raising the possibility that
CRDs may contribute to neurogenic hypertro-
phy in these cases.43 Rarely, CRDs occur in Fasciculation Potentials
otherwise normal muscles, such as the iliopsoas
or biceps. Fasciculation potentials are randomly discharg-
CRDs may be confused with other repeti- ing action potentials of a group of muscle
tive discharges, such as myokymic discharges, fibers innervated by the same anterior horn cell
cramps, neuromyotonia, tremor, and synkine- (motor unit) (Fig. 26–17). These spontaneously
sis. However, each of these has a characteristic firing MUPs may be generated anywhere along
pattern of firing best recognized by its sound the lower motor neuron, from the anterior
and distinct from that of CRDs. horn cell to the nerve terminal, but usually
from spontaneous firing of the nerve terminal.
Key Points The rates of discharge of an individual poten-
tial may vary from a few per second to fewer
• CRDs are the action potentials of groups than 1 per minute. The sum of all fascicula-
of adjacent muscle fibers discharging tions in a muscle may reach 500 per minute.
spontaneously in near synchrony in a reg- These potentials may be of any size and shape,
ular, repetitive fashion. depending on the character of the motor unit
• CRDs occur as a result of ephaptic spread from which they arise and their relation with
of single muscle fiber action potentials the recording electrode, and they may have
along a circuit of neighboring muscle the appearance of normal or abnormal MUPs.
fibers. They are identified by their irregular firing
Figure 26–17. Fasciculation potentials recurring in an irregular pattern. A, Slow sweep speed, continuous. B, Fast sweep
speed, raster. (From Daube, J. R. 1991. AAEM minimonography #11: Needle examination in clinical electromyography.
Muscle & Nerve 14:685–700. By permission of John Wiley & Sons.) (From Daube JR. AAEM minimonography #11: needle
examination in clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
pattern and may sound like “large raindrops on diseases. In normal persons, fasciculations
a tin roof.” occur more rapidly, on the average, and are
Fasciculation potentials may occur in nor- more stable.45 The presence of fasciculations
mal individuals as well as in individuals with alone on EMG, without fibrillation potentials
many different neuromuscular diseases. They or changes in voluntary MUPs, is not suffi-
are especially common in chronic neurogenic cient to make a diagnosis of progressive motor
disorders but have been found in all neuromus- neuron disease, such as ALS.
cular disorders (Table 26–6). Fasciculations Patients who have large motor units caused
usually occur in an overworked muscle, espe- by chronic neurogenic diseases may have vis-
cially if there is underlying neurogenic disease. ible twitching during voluntary contractions.
Fasciculation potentials have not been shown Such contraction fasciculations must be differ-
to occur more often in patients with myopathy entiated from true fasciculations by the pattern
than in normal persons. of firing.
Electrodiagnostic testing, using surface
EMG, detects fasciculations more frequently
than clinical observation or muscle palpation,
Key Points
and therefore EMG is useful in assessing for
fasciculations and other changes in patients • Fasciculation potentials are spontaneously
with suspected amyotrophic lateral sclerosis firing MUPs, which are generated any-
(ALS).44 However, neither surface nor needle where along the lower motor neuron.
EMG can reliably distinguish between benign • Fasciculation potentials fire in an irregular
fasciculations and those associated with specific pattern, at fast or slow rates.
Table 26–6 Disorders Associated with Fasciculation Potentials

Normal Benign fasciculation syndrome
Peripheral nerve Cramp fasciculation syndrome
hyperexcitability syndrome
Neurogenic disorders Anterior horn cell diseases (e.g., ALS,
Kennedy disease, spinal muscular atrophy)
Peripheral neuropathies, axonal
Radiculopathies
Metabolic disorders Hyperthyroidism
Medications Anticholinesterase agents
• Fasciculation potentials may be of any size or with a regularly changing rate of discharge
or shape, and the configuration reflects may have similar mechanisms, they are bet-
the motor units from which they arise. ter classified with the broad group of itera-
• Fasciculations are nonspecific in signifi- tive discharges. Some investigators consider
cance; they may be a benign phenomenon iterative discharges and myokymic discharges
or seen with neurogenic disorders. to be forms of fasciculation because they
arise in the lower motor neuron or axon.
However, it is best to separate these dis-
Myokymic Discharges charges from fasciculation potentials because
of their distinct patterns and different clinical
Myokymic discharges are groups of recurring, significance.
spontaneously firing MUPs that fire in a repet- Myokymic discharges may or may not
itive burst pattern. The individual potentials be associated with clinical myokymia, which
within each burst often have the appearance appears as fine, worm-like quivering of the
of normal MUPs, although they may also be of muscles. Although myokymic discharges are
long duration and high amplitude. Each burst more commonly found in limb muscles, clini-
may be composed of few or many potentials cal myokymia is more often observed in facial
(2–10) and the rate of firing of potentials within muscles, probably due to the smaller degree
each burst is typically 40–60 Hz. Each burst of overlying subcutaneous tissue, than in limb
fires with a regular or semirhythmic pattern at muscles. Diseases associated with myokymic
intervals of 0.1–10 seconds (Fig. 26–18). The discharges are listed in Table 26–7. Most
firing pattern is unaffected by voluntary activ- commonly, myokymic discharges are found
ity, and simultaneously occurring myokymic with radiation-induced nerve injury, chronic
discharges may vary in burst duration or fir- compressive neuropathies, or polyradiculo-
ing rates. Some myokymic discharges sound pathies. The myokymic discharges seen in
similar to groups of marching soldiers. chronic compressive neuropathies, such as
Although discharges that have regular pat- carpal tunnel syndrome, are often composed of
terns of recurrence but fire at different rates a single or few potentials.
Figure 26–18. Examples of recurrent bursts of myokymic discharges at a slow (left) and fast (right) sweep speed. Firing
rate is 20–30 per second within bursts, with variable recurrence rates of the bursts. (From Daube JR. AAEM minimonog-
raphy #11: needle examination in clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of
the publisher.)
Table 26–7 Disorders Associated with Myokymic Discharges

Facial muscles Multiple sclerosis
Brain stem neoplasms
Polyradiculopathy (e.g., acute inflammatory
demyelinating polyneuropathy)
Facial neuropathy (e.g., Bell’s palsy)
Radiation to head and neck
Extremity muscles Radiation (plexopathy, mononeuropathy)
Chronic nerve compression (e.g., chronic
carpal tunnel syndrome)
Syndrome of peripheral nerve
hyperexcitability (Isaac’s syndrome)
Morvan’s syndrome
Key Points • Myokymic discharges are most commonly

seen with radiation-induced nerve injury,
• Myokymic discharges are groups of recur-
but can occur in other disorders such as
ring, spontaneously firing MUPs that fire chronic nerve compression or polyradicu-
in a repetitive burst pattern. lopathies.
• Each burst may be composed of few or
many potentials (2–10) and the rate of
firing of potentials within each burst is
typically 40–60 Hz. Neuromyotonic Discharges
• Each burst fires with a regular or (Neuromyotonia)
semirhythmic pattern at intervals of
0.1–10 seconds. Neuromyotonic discharges, or neuromyoto-
• The firing pattern is not affected by volun- nia, are rare, spontaneously firing MUPs that
tary activity. are associated with some forms of continu-
• Myokymic discharges may or may not be ous muscle fiber activity (Isaac’s syndrome).
associated with clinical myokymia, which Neuromyotonic discharges fire at very high
is seen more often in facial than limb frequencies of 100–300 Hz (Figs. 26–19 and
muscles. 26–20). These potentials may decrease in
Figure 26–19. Examples of neuromyotonic (neurotonic) discharges in Isaac’s syndrome. (From Daube, J. R. 1991. AAEM
minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of
John Wiley & Sons.)
Figure 26–20. Two examples of neuromyotonic discharges in spinal muscular atrophy firing at over 200 per second.
(From Daube, J. R. 1991. AAEM minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve
14:685–700. By permission of John Wiley & Sons.) (From Daube JR. AAEM minimonography #11: needle examination in
clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
Table 26–8 Disorders Associated with Neuromyotonic Discharges

Hyperexcitable nerve syndromes Syndrome of peripheral nerve hyperexcitability
(Isaac’s syndrome)
Tetany
Morvan’s syndrome
Neurogenic Chronic spinal muscular atrophy
Hereditary motor neuropathy
Other Anticholinesterase poisoning
Intraoperative nerve irritation
amplitude because of the inability of mus- frequencies often ranging from 40–350 Hz,
cle fibers to maintain discharges at rates which may appear similar to myokymic dis-
greater than 100 Hz. The discharges may charges.47, 48 Neuromyotonia may also occur
be continuous for long intervals or recur with tetany, where they may be precipitated
in bursts. They are unaffected by voluntary by or augmented with ischemia, and Morvan’s
activity. syndrome.49
Neuromyotonic discharges are seen in dis- A form of neuromyotonic discharges called
orders of peripheral nerve hyperexcitability, neurotonic discharges occur intraoperatively
such as Isaac’s syndrome, and may occur as a with the mechanical irritation of cranial or
result of a defect in potassium channels in the peripheral nerves. These discharges are brief
nerve membrane (Table 26–8).46 Some forms bursts of MUPs discharging at very high rates,
of syndromes of peripheral nerve hyperex- similar to the rates of spontaneously occurring
citability are associated with bursts of doublet, neuromyotonic discharges. The identification
triplet, or multiplet discharges, with intraburst of neurotonic discharges intraoperatively is
Figure 26–21. Neurotonic discharges in facial muscles during acoustic neuroma surgery. The times of recordings were
at 1:10 pm, 1:50 pm, and 3:10 pm. (From Daube, J. R. 1991. AAEM minimonography #11: Needle examination in clin-
ical electromyography. Muscle & Nerve 14:685–700. By permission of John Wiley & Sons.) (From Daube JR. AAEM
minimonography #11: needle examination in clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with
permission of the publisher.)
valuable in alerting surgeons to possible nerve of MUPs that fire in a unique firing pattern,
damage (Fig. 26–21). which distinguishes them from other spon-
taneous activity and normal strong voluntary
Key Points activation. The configuration of the individual
• Neuromyotonic discharges are rare, spon- potentials resembles MUPs. However, in con-
taneously occurring MUPs that fire at trast to the pattern of activation that occurs
rates of 100–300 Hz. with voluntary contraction, potentials in cramp
• Neuromyotonic discharges may fire as discharges usually have an abrupt onset, rapid
long continuous discharges or brief bursts. buildup and addition of subsequent poten-
• They are often seen in disorders of periph- tials, and a rapid or sputtering cessation. The
eral nerve hyperexcitability, such as Isaac’s potentials fire rapidly (40–60 Hz) and during
syndrome, as well as tetany or other neu- their discharge they may fire irregularly in a
rogenic disorders. sputtering fashion, especially just before ter-
• A form of neuromyotonic discharges mination (Fig. 26–22). Typically, an increas-
called neurotonic discharges occur intra- ing number of potentials that fire at simi-
operatively with the mechanical irritation lar rates are recruited as the cramp develops
of cranial or peripheral nerves. and these potentials stop firing as the cramp
subsides.
Cramps are a common phenomenon in nor-
mal persons, usually when a muscle is activated
Cramp Potentials (Cramp strongly in a shortened position. In addition,
Discharge) cramps may occur with any chronic neuro-
genic disorder, in metabolic or electrolyte dis-
Cramps are painful, involuntary contractions of orders, or in disorders of peripheral nerve
muscle. The discharges associated with a mus- hyperexcitability (such as cramp fasciculation
cle cramp (cramp discharges) are composed syndrome) (Table 26–9).
Figure 26–22. Muscle cramp with MUPs firing at 30–50 per second. (From Daube, J. R. 1991. AAEM minimonogra-
phy #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of John Wiley &
Sons.) (From Daube JR. AAEM minimonography #11: needle examination in clinical electromyography. Muscle & Nerve.
1991;14:685–700. Used with permission of the publisher.)
Table 26–9 Disorders Associated with Cramp

Discharges
Neurogenic disorders Chronic radiculopathies
Peripheral neuropathy
Motor neuron disorders
Metabolic or electrolyte disorders Salt depletion
Pregnancy
Hypothyroidism
Uremia (dialysis)
Peripheral nerve hyperexcitability Cramp fasciculation syndrome
disorders
Other Benign nocturnal cramps
Key Points disorders, metabolic or electrolyte distur-

bances, or in peripheral nerve hyperex-
• Cramp discharges are composed of MUPs
citability syndromes.
that fire with an abrupt onset, rapid
buildup and addition of subsequent poten-
tials, and a rapid or sputtering cessation.
• Cramp discharges can be identified by the Synkinesis
rapid recruitment and irregular firing pat-
tern of MUPs, often associated with the The aberrant regeneration of axons after nerve
patient’s experience of a painful muscle injury may result in two different muscles
cramp. being innervated by the same axon called
• The potentials fire rapidly (40–60 Hz) and synkinesis. In such cases, voluntary potentials
during their discharge they may fire irreg- may be mistaken for spontaneous activity.
ularly in a sputtering fashion, especially Groups of MUPs fire in bursts in response
just before termination. to voluntary activation of a distant mus-
• Cramp discharges may occur in nor- cle. With synkinesis, MUPs may be normal
mal individuals, in chronic neurogenic or abnormal and, when abnormal, they are
Figure 26–23. Respiratory synkinesis. Spontaneously firing MUPs in the deltoid with long breaths. The rate and number
of potentials increase and then decrease with each inspiration. (From Daube JR. AAEM minimonography #11: needle
examination in clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
typically of long duration due to reinnervation diseases, both MUP firing rates and config-
from a neurogenic lesion. A common exam- urations may be altered. The types of these
ple of this is facial synkinesis, in which facial alterations, in conjunction with the identifica-
muscles such as the orbicularis oris sponta- tion of spontaneous discharges, help to identify
neously fire MUPs in association with blink- the underlying type, temporal profile of dis-
ing after facial reinnervation from facial neu- ease duration, and severity of neuromuscular
ropathy (Bells’ palsy). Another less common disorder.
example is arm–diaphragm synkinesis (also
referred to as the breathing arm or hand) in
which potentials in shoulder girdle or hand
muscles fire in association with respiration Abnormal Recruitment
as a result of aberrant regeneration of the
phrenic nerve following brachial plexus injuries In a normal muscle, increasing voluntary effort
(Fig. 26–23).50–52 causes an increase in the rate of firing of
individual MUPs and initiates the discharge
of additional MUPs. The relationship between
Key Points the rate of firing of individual potentials to the
• Synkinesis is the aberrant regeneration of number of potentials firing is constant for a
particular muscle and is called the recruitment
axons after nerve injury which may result
pattern.
in two different muscles being innervated
In disorders in which there is a loss of
by the same axon.
• Synkinesis is seen as groups of normal MUPs, the rate of firing of the remaining
individual potentials will be disproportionately
or abnormal MUPs that fire in bursts in
high compared to the number of potentials
response to voluntary activation of a dis-
firing; this is referred to as reduced recruit-
tant muscle.
• Synkinesis may occur in facial muscles fol- ment. Reduced recruitment may be found in
any disease process that destroys or blocks con-
lowing facial nerve palsy or arm muscles
duction in the axons innervating the muscle
following brachial plexus injuries.
or destroys a sufficient proportion of the mus-
cle so that muscle fibers of entire motor units
are lost. This pattern occurs in association with
all neurogenic disorders associated with axonal
ABNORMAL ELECTRICAL loss and may be the only finding in a neu-
ACTIVITY—VOLUNTARY MUPs rapraxic lesion in which the sole abnormality
is a focal conduction block. Reduced recruit-
The characteristic features of normal volun- ment may be the earliest finding in an acute
tary MUPs have been detailed in the Normal axonal lesion in which fibrillation potentials
Motor Unit Potentials section of this chap- or other MUP changes have not yet devel-
ter. The majority of normal MUPs in limb oped. Although a hallmark of neurogenic disor-
muscles are triphasic with durations of 8– ders, reduced recruitment may also be seen in
10 ms, stable appearing, and initially fire at severe or end-stage myopathies, where entire
rates of 6–8 Hz with an orderly increase in fir- motor units are lost due to primary muscle
ing rate associated with the firing of additional fiber degeneration, such as in muscular dystro-
units (normal recruitment). In neuromuscular phies.
Rapid recruitment of MUPs occurs in dis- can generate is decreased due to loss of
orders in which the force that a single motor muscle fibers within the motor unit.
unit can generate is decreased due to loss • Identification of rapid recruitment requires
of muscle fibers within the motor unit. As a assessment of patient effort.
result, more motor units are activated than
would be expected for the force exerted by
the patient. The recruitment frequency and
rate of firing in relation to number are normal Long-Duration MUPs
with rapid recruitment; however, the number
of motor units that fire is increased relative to MUP duration is measured as the time from
force. Rapid recruitment occurs primarily in the initial baseline deflection to the time of the
myopathies. While in many cases, abnormali- return to baseline, and reflects the density and
ties in MUP configuration will occur along with area of fibers within a motor unit, as well as the
abnormal recruitment, this is not always the synchrony of firing of those fibers. The size of
case and rapid recruitment may be the only MUPs in a muscle is dependent on the level of
abnormality identified on needle examination, activation, with larger MUPs becoming active
particularly in early or mild myopathies. at a stronger force.53 Normal values for MUP
duration have been published.54
Individual MUPs that are longer than the
Key Points normal range for a particular muscle or groups
of MUPs that have a mean duration greater
• Recruitment refers to the relationship than the normal range for the same muscle
between the rate of firing of individual in a patient of the same age are called long-
potentials and the number of potentials duration MUPs (Fig. 26–24). These occur in
firing which is constant for a particular diseases in which there is increased fiber den-
muscle. sity in a motor unit, an increased number of
• Reduced recruitment occurs when the fibers in a motor unit, or loss of synchronous
rate of firing of MUPs is increased relative firing of fibers in a motor unit, typically due
to the number of potentials firing. to collateral sprouting and reinnervation of
• Reduced recruitment is a hallmark of neu- a motor unit. Long-duration MUPs generally
rogenic disorders and occurs with axonal have high amplitude and show reduced recruit-
loss or conduction block. ment, but since the spike amplitude reflects
• Reduced recruitment can be seen in only the few muscle fibers closest to the nee-
severe or end-stage myopathies. dle recording tip they may have normal or
• Rapid recruitment occurs in disorders in low amplitude. When assessing MUP duration,
which the force that a single motor unit those MUPs recorded from damaged muscle
Free-running sweep - One second (9 per second)
Triggered sweep, 5 ms/div
Single trace
Superimposed
traces
20 ms duration
Figure 26–24. Single long-duration voluntary MUP displayed on a free-running and triggered sweep. Semirhythmic
firing rate of 9 per second without recruitment of other potentials is abnormal for this muscle.
Table 26–10 Disorders Associated with Long-Duration MUPs

Neurogenic disorders Motor neuron diseases (e.g., ALS, poliomyelitis, spinal muscular
atrophy)
Chronic axonal neuropathies (e.g., hereditary motor sensory
neuropathy type 2, diabetic neuropathy)
Chronic radiculopathies or the residua of an old radiculopathy
Chronic mononeuropathies or the residua of an old
mononeuropathy
Myopathies Chronic myopathies (e.g., inclusion body myositis)
fibers that are preponderantly positive with age are called short-duration MUPs. Short-
a long late negativity, which is a recording duration MUPs occur in diseases in which
artifact, should not be measured or interpreted there is (1) physiologic or anatomical loss of
as long duration. muscle fibers from the motor unit or (2) atro-
MUP duration is an important parame- phy of component muscle fibers. In these situ-
ter used to distinguish neurogenic disorders ations the number of innervated muscle fibers
from primary muscle diseases.55 Long-duration within the recording region of the electrode
MUPs typically occur in chronic neurogenic is decreased, thereby leading to a decrease in
disorders. Following an acute nerve injury, the area of that motor unit. Commonly, these
long-duration MUPs may be seen within sev- potentials also have low amplitude and show
eral weeks or months after reinnervation rapid recruitment with minimal effort, but they
has begun. Long-duration MUPs may also may have normal or reduced recruitment and
be seen in conjunction with short-duration normal amplitudes. The actual duration that
MUPs in chronic myopathies, such as inclusion identifies a potential as short duration varies
body myositis or long-standing polymyositis with the muscle and age of the patient. Some
(Table 26–10). short-duration MUPs may be as short as 1–3 ms
if only a single muscle fiber is in the recording
Key Points area. This may appear identical to a fibril-
lation potential or end plate spike, and only
• Long-duration MUPs occur in diseases in the semirhythmic firing pattern may allow for
which there is increased fiber density in a correct identification.
motor unit, an increased number of fibers Short-duration MUPs are most characteris-
in a motor unit, or loss of synchronous fir- tic and are often seen in primary muscle dis-
ing of fibers in a motor unit, typically due eases in which loss of muscle fibers from necro-
to collateral sprouting and reinnervation sis or degeneration occurs (Table 26–11).56
of a motor unit. Some myopathies, such as metabolic and
• Long-duration MUPs typically occur in endocrine disorders, show no or few short-
chronic neurogenic disorders. duration MUPs. In rare circumstances, short-
• Long-duration MUPs may also be seen duration MUPs can occur due to techni-
in conjunction with short-duration MUPs cal problems, such as incorrect filter settings
in chronic myopathies, such as inclusion (e.g., low-frequency filter increased from 20 to
body myositis or long-standing polym- 500 Hz) or an electrical short in the recording
yositis. electrode or connecting cables. When short-
duration MUPs occur when not expected,
these technical problems should be considered
Short-Duration MUPs and checked.
In addition to myopathies, short-duration
Single MUPs that are shorter than the nor- MUPs may occur in severe neuromuscular
mal range or groups of MUPs that have a junction disorders or in newly reinnervated
mean duration less than the normal range for motor units following severe nerve injury.
the same muscle in a patient of the same These nascent MUPs are composed of only a
Table 26–11 Disorders Associated with Short-Duration MUPs

Myopathies Muscular dystrophies
Inflammatory myopathies (e.g., polymyositis, inclusion body
myositis)
Infiltrative myopathies (e.g., sarcoidosis, amyloid)
Toxic myopathies (e.g., lipid-lowering agents, chloroquine)
Congenital myopathies
Endocrine myopathies (e.g., hypothyroid)
Neuromuscular junction disorders Myasthenia gravis
Lambert–Eaton myasthenic syndrome
Botulinum intoxication
Neurogenic disorders Early reinnervation after nerve damage (“nascent MUP”)
Late-stage neurogenic atrophy
Disorders of muscle membrane Periodic paralysis
few muscle fiber action potentials, are typically Polyphasic MUPs

polyphasic, and fire at a very high rate with
reduced recruitment. A phase of an MUP is defined as the area of
a potential on either side of the baseline and
is equal to the number of baseline crossings
Key Points plus one. Most normal MUPs contain three or
four phases, and less than 15% will have over
• Short-duration MUPs occur in diseases four phases. When an MUP consists of five
in which there is physiologic or anatom- or more phases, it is called a polyphasic MUP
ical loss of muscle fibers from the motor (Figs. 26–25 and 26–26). The individual com-
unit or atrophy of component muscle ponents of a polyphasic potential are action
fibers. potentials recorded from a single or a few
• Short-duration MUPs are most character- muscle fibers. The degree of phases reflects
istic of primary muscle diseases in which the synchrony of firing of the action potentials
loss of muscle fibers from necrosis or of muscle fibers within the MUP and when
degeneration occurs. the fibers fire asynchronously, the number of
• Short-duration MUPs may also be seen in phases (or turns) increases. This may occur as
severe neuromuscular junction disorders. a result of collateral sprouting, reinnervation,
• Nascent MUPs are short-duration, low- and an increase in fiber density (in neurogenic
amplitude, polyphasic MUPs that fire disorders) or due to relative asynchrony from
with reduced recruitment, and are signs drop-out of muscle fibers or differences in
of early reinnervation following a severe muscle fiber conduction velocities in the motor
nerve lesion. unit (in myopathies).
Figure 26–25. Polyphasic, long-duration MUP displayed on a free-running sweep at 10 ms per division. (From Daube JR.
AAEM minimonography #11: needle examination in clinical electromyography. Muscle & Nerve. 1991;14:685–700. Used
with permission of the publisher.)
Polyphasic potentials may be of any (Figs. 26–26 and 26–27). However, isolated
duration—normal, long, or short. Some may satellite potentials should not be included
have late, satellite components, sometimes in the duration measurement of the MUPs
called linked potentials or satellite potentials, when comparing to normative data. Polypha-
which give the total unit a long duration57 sic MUPs may occur in any of the myopathies
–5000
–4000
–3000
μV
–2000
–1000
–500
+500
ms
Normal MUP Highly Fibrillation
MUP in primary polyphasic potential
in slight muscular MUP
contraction disorders
MUP in
lesions of
anterior
horn cells
Figure 26–26. Relative average durations and amplitudes of some electric potentials observed in EMG of human muscle.
(From Daube, J.R. 1991. AAEM minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve
14:685–700. By permission of John Wiley & Sons.)
Figure 26–27. Long-duration polyphasic MUP with satellite potential. (From Daube JR. AAEM minimonography
publisher.)
Figure 26–28. Patient with inclusion body myositis. Quantitation of MUPs shows a bimodal distribution. (From Daube,
J. R. 1991. AAEM minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700.
By permission of John Wiley & Sons.) (From Daube JR. AAEM minimonography #11: needle examination in clinical
electromyography. Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
or the neurogenic disorders and are graded by or short. Rarely, the distribution of durations
the percentage of MUPs in the muscle that are may be bimodal (Fig. 26–28). These combina-
polyphasic. tions commonly occur in chronic myositis or in
rapidly progressing motor neuron disease.
Key Points
• A phase of an MUP is defined as the area
Varying or Unstable MUPs
of a potential on either side of the base-
line and is equal to the number of baseline MUPs fire repetitively under voluntary control,
crossings plus one. and they normally have the same amplitude,
• When an MUP consists of five or more
duration, and configuration each time they fire.
phases, it is called a polyphasic MUP. Fluctuation of any of these variables during
• In a normal muscle, no more than 15% of
repeated discharge of an MUP is abnormal and
MUPs are polyphasic. produces varying or unstable MUPs. Varying
• The degree of phases reflects the syn-
MUPs are caused by blocking of the discharge
chrony of firing of the action potentials of of action potentials of one or a few of the indi-
muscle fibers. vidual muscle fibers comprising the motor unit.
• Polyphasic MUPs may be seen in myopathies
The disorders in which MUPs fluctuate from
or in neurogenic disorders. moment to moment (Fig. 26–29) are listed in
Table 26–12. Varying MUPs are classically seen
in disorders of neuromuscular transmission,
Mixed Patterns: Long-Duration such as myasthenia gravis (MG) or Lambert–
and Short-Duration MUPs Eaton myasthenic syndrome (LEMS), but may
also be seen in reinnervating neurogenic disor-
Occasionally, patients have a combination of ders and occasionally in myopathies. In disor-
the abnormalities described for short, long, ders of muscle membrane, such as myotonia,
and polyphasic MUPs, but instead of having there may be a slower progressive decrease or
the usual pattern of an excess of either long- increase in an MUP (Fig. 26–30). In MG or
duration or short-duration potentials, both in cases of active reinnervation, the amplitude
types occur. The quantitative distribution initially may decline, but in the myasthenic
becomes broad rather than shifting to long syndrome, it may increase (Fig. 26–31).
Normal
500 μV
–
+
0.01′′
Decline of amplitude
Variation of amplitude
Figure 26–29. Top, Normal voluntary MUPs. Middle and bottom, Motor unit instability in MG. (From Daube, J. R. 1991.
AAEM minimonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission
of John Wiley & Sons.) (From Daube JR. AAEM minimonography #11: needle examination in clinical electromyography.
Muscle & Nerve. 1991;14:685–700. Used with permission of the publisher.)
Table 26–12 Disorders Associated with Varying (Unstable) MUPs

Neuromuscular junction disorders Myasthenia gravis
Botulism
Congenital myasthenic syndromes
Neurogenic disorders Reinnervation after nerve injury
Progressing neurogenic disorders (e.g.,
ALS)
Myopathies Inflammatory myopathies
Figure 26–30. Schema of motor unit potential showing characteristics that can be measured. Rise time defines the dis-
tance from the generator. Spikes and turns reflect the number of fibers; duration is determined by fiber size and synchrony.
Stability changes with disorders of the neuromuscular junction or nerve terminal. (From Daube, J. R. 1991. AAEM min-
imonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of John
Wiley & Sons.) (From Daube JR. AAEM minimonography #11: needle examination in clinical electromyography. Muscle
& Nerve. 1991;14:685–700. Used with permission of the publisher.)
443
Figure 26–31. MUP variation with gradual increase in amplitude in the LEMS. (From Daube, J. R. 1991. AAEM min-
imonography #11: Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of John
Wiley & Sons.) (From Daube JR. AAEM minimonography #11: needle examination in clinical electromyography. Muscle
& Nerve. 1991;14:685–700. Used with permission of the publisher.)
Key Points channel antibodies.47 In these patients, the

• Varying MUPs occur when there is block- doublets and multiplets have been reported
to occur more commonly in distal muscles
ing of the discharge of action potentials of and the intraburst frequency ranges from
one or a few of the muscle fibers within 40 to 350 Hz.47
the motor unit.
• MUP variation may occur in disorders
Key Points
of neuromuscular transmission or reinner-
vating neurogenic disorders. • Doublets or multiplets are the repetitive
firing of the same MUP two or more
times at short intervals (10–30 ms) in a
semirhythmic pattern.
Doublets (Multiplets) • Doublets or multiplets may be seen with
hyperventilation, hypocalcemia, ischemia,
Motor units under voluntary control normally or hyperexcitable nerve syndromes.
discharge as single potentials in a semirhyth-
mic fashion. In some disorders or occasionally
in otherwise normal individuals, they fire two Table 26–13 Disorders
or more times at short intervals of 10–30 ms
(Table 26–13) (Fig. 26–32). These are called
Associated with Doublets or
doublets, triplets, or multiplets. The bursts of Multiplets
two or more potentials recur in a semirhyth-
Hyperventilation
mic pattern under voluntary control. They are Tetany
often increased by hyperventilation, hypocal- Motor neuron disease (infrequent)
cemia, or ischemia. Additionally, doublets or Syndrome of peripheral nerve hyperexcitability
multiplets may be seen in patients with dis- (Isaac’s syndrome)
orders of peripheral nerve hyperexcitability, Other metabolic diseases
often associated with voltage-gated potassium
Figure 26–32. Voluntary MUPs. A, Doublets. B, Multiplets. (From Daube, J. R. 1991. AAEM minimonography #11:
Needle examination in clinical electromyography. Muscle & Nerve 14:685–700. By permission of John Wiley & Sons.)
ABNORMAL ELECTRICAL voluntary control. In upper motor neuron

ACTIVITY—DISORDERS OF weakness, patients cannot maintain motor unit
CENTRAL CONTROL firing.
Normal MUPs may fire spontaneously in dif- Key Points

ferent patterns as a result of central nervous
system motor disorders. When this occurs, the • Tremor is characterized by MUPs firing
MUP configuration is normal, since there is no in groups but not in a fixed relation; the
dysfunction of the motor unit. potentials of these motor units are super-
imposed and may resemble polyphasic,
complex, or long-duration MUPs.
Tremor • MUP firing patterns in dystonia, rigidity,
spasticity, and stiff-man syndrome are nor-
Tremor is the most common pattern of motor mal MUP and resemble normal patterns
unit firing that is caused by disorders of but with loss of voluntary control.
the central nervous system. Tremor must be
recognized because the discharge may be con-
fused with the changes seen with lower motor PATTERNS OF ABNORMALITIES
neuron disease, such as polyphasic MUPs or
myokymic discharges. While the electrical dis- The types of needle EMG abnormalities
charge in tremor is often associated with a described above may occur in different com-
clinical tremor, this is not always the case. In binations. Only through knowledge of these
muscle tremor, MUPs fire in groups but not in combinations can reliable interpretations be
a fixed relation. The potentials of these motor made. No single finding allows the identi-
units are superimposed and may resemble fication of a specific disease. The combi-
polyphasic, complex, or long-duration MUPs nations of particular forms of spontaneous
(Fig. 26–33). They are recognized by their activity and changes in MUPs in neuro-
rhythmic (often regular) pattern and changing muscular diseases are too varied to be
appearance. Minimal activation, with slightly included in this review, but some general
increasing and decreasing effort, often allows comments about patterns of abnormality of
single MUPs to be resolved and characterized. MUPs can be made. MUP changes have been
divided broadly into neuropathic and myo-
pathic.58 The concept that MUP changes must
Dystonia, Rigidity, Spasticity, be either one or the other of these two
Stiff-man syndrome types is incorrect and can lead to misinter-
pretations.
MUP firing patterns in dystonia, rigidity, Each of the variables—recruitment, dura-
spasticity, and stiff-man syndrome are nor- tion, amplitude, configuration, and stability—
mal and resemble normal patterns with less of MUPs may be altered separately or in
Figure 26–33. Superimposed MUPs with tremor that resemble polyphasic potentials.
judged individually, quantified if necessary,

and compared with normal values. The result
should then be interpreted on the basis of
known pathophysiologic mechanisms or by
common association with known disorders.
Recruitment, duration, and stability are the
important features of MUPs in determining
the underlying pathologic factors. With these
three criteria, it is possible to distinguish most
patterns of MUP abnormality. Each pattern
of abnormality changes with the severity and
duration of the disease (Fig. 26–34). Careful
attention to the independent changes of the
Figure 26–34. Temporal profile of MUPs changes in variables of MUPs can allow an electromyo-
neurogenic diseases. grapher to comment on the severity, dura-
tion, and prognosis of a disease.60 Because
combination with one or more of the oth- of the various patterns that may be found, a
ers in different disorders.59 Each must be description of the abnormalities should always
Table 26–14 Patterns of Abnormalities Seen With Needle EMG

Recruitment MUP appearance Variation Disorder
Normal Normal No Normal

Some metabolic or endocrine myopathies
Yes Neuromuscular junction disorders (e.g.,
MG, LEMS)
Short duration, polyphasic No Primary myopathies
Yes Severe neuromuscular junction disorders
(e.g., MG, LEMS, botulism)
Primary myopathies (occasionally)
Mixed short and long duration No or Yes Chronic myopathies (e.g., inclusion body
myositis)
Reduced Normal No Acute neurogenic lesion
Yes Subacute neurogenic lesion
Long duration, polyphasic No Chronic neurogenic lesion
Yes Chronic, progressing neurogenic lesion
Short duration, polyphasic No Severe myopathy
End-stage neurogenic disorder
Yes Early reinnervation after severe nerve
damage
Severe, reinnervating neurogenic
disorders
Mixed short and long duration No or Yes Rapidly progressing neurogenic disorders
(e.g., ALS)
Rapid Normal No Mild myopathies
Yes —
Short duration No Myopathies
Yes Myopathies
Long duration No Chronic myopathies
Yes Chronic myopathies (occasionally)
include comments about each of the variables. disease, its duration or stage of evolution, and
The findings then can be interpreted most the likely anatomical location of the pathologic
reliably by listing the disorders that may be process.
seen with the pattern of abnormality found
(Table 26–14).
REFERENCES
SUMMARY 1. Caruso, G., A. Eisen, E. Stålberg, et al. 1999.
Clinical EMG and glossary of terms most com-
Virtually all primary neuromuscular diseases monly used by clinical electromyographers. The
International Federation of Clinical Neurophysiology.
result in changes in the electric activity Electroencephalography and Clinical Neurophysiol-
recorded from muscle fibers. These changes ogy 52(Suppl):189–98.
can best be depicted using fine needle elec- 2. Stålberg, E., B. Falck, A. Gilai, J. Jabre, M. Sonoo,
trodes inserted into the muscle to record spon- and K. Todnem. 1999. Standards for quantification of
EMG and neurography. The International Federation
taneous and voluntary EMG. Thus, EMG can of Clinical Neurophysiology. Electroencephalog-
be used to distinguish among lower motor raphy and Clinical Neurophysiology 52(Suppl):
neuron, peripheral nerve, neuromuscular junc- 213–20.
tion, and muscle disease with great sensitivity 3. Strommen, J. A., and J. R. Daube. 2001. Determi-
and some specificity. The sensitivity is usu- nants of pain in needle electromyography. Clinical
ally greater than clinical measures; specificity 4. Fuglsang-Frederiksen, A. 1987. Quantitative elec-
in identifying the cause of the disease often tromyography. I. Comparison of different meth-
requires muscle biopsy or other clinical mea- ods. Electromyography and Clinical Neurophysiology
sures. Although EMG is somewhat uncom- 27:327–33.
5. Fuglsang-Frederiksen, A. 1987. Quantitative elec-
fortable for patients because needles need to tromyography. II. Modifications of the turns analysis.
be inserted into the muscles, it generally is Electromyography and Clinical Neurophysiology
well tolerated by patients and provides a rapid, 27:335–8.
efficient means of testing the motor unit. 6. Fuglsang-Frederiksen, A., and J. Ronager. 1990. EMG
The application of techniques of clinical power spectrum, turns-amplitude analysis and motor
unit potential duration in neuromuscular disorders.
neurophysiology in the evaluation of peripheral Journal of the Neurological Sciences 97:81–91.
neuromuscular disorders relies heavily on nee- 7. Pfeiffer, G., and K. Kunze. 1992. Turn and phase
dle EMG. It was the first of the electrophysio- counts of individual motor unit potentials: Correla-
logic techniques to be applied in this way, and tion and reliability. Electromyography and Clinical
it has remained a mainstay of electrodiagno- 8. Al-Shekhlee, A., B. E. Shapiro, and D. C. Preston.
sis. The collection of data with needle EMG 2003. Iatrogenic complications and risks of nerve con-
and the interpretation for clinical purposes duction studies and needle electromyography. Muscle
require a firm understanding of the physiol- & Nerve 27:517–26.
ogy of muscle fibers and motor units. It is 9. Gruis, K. L., A. A. Little, V. A. Zebarah, and J. W.
Albers. 2006. Survey of electrodiagnostic laborato-
heavily dependent on controlling technical fac- ries regarding hemorrhagic complications from needle
tors, mastering the skills of data collection, electromyography. Muscle & Nerve 34:356–8.
and understanding the changes that occur with 10. Rosioreanu, A., A. Dickson, S. Lypen, and D. S. Katz.
the many disorders that may affect peripheral 2005. Pseudoaneurysm of the calf after electromyo-
graphy: Sonographic and CT angiographic diagnosis.
nerves, neuromuscular junctions, and muscle. American Journal of Roentgenology 185:282–3.
The essence of quality needle EMG rests 11. Butler, M. L., and R. W. Dewan. 1984. Subcuta-
with the ability to isolate, recognize, and inter- neous hemorrhage in a patient receiving anticoagulant
pret the wide range of specific waveforms therapy: An unusual EMG complication. Archives of
and their variation that occur in normal and Physical Medicine and Rehabilitation 65:733–4.
12. Lynch, S., A. Boon, J. Smith, and C. Harper.
diseased muscle. The nature and meaning 2008. Complications of needle electromyography:
of fibrillation potentials and the alteration in Hematoma risk and correlation with anticoagulation
appearance and firing pattern of MUPs in each and antiplatelet therapy. Muscle & Nerve, 38:1225–30.
muscle tested are the data on which a clinical 13. AANEM Position Statement. 2005. Needle EMG in
certain uncommon clinical contexts. Muscle & Nerve
interpretation is based. The extent and distri- 31:398–9.
bution of these abnormalities allow conclusions 14. Dumitru, D., J. C. King, and S. D. Nandedkar. 1997.
to be drawn about the type and severity of Motor unit action potential duration recorded by
monopolar and concentric needle electrodes. Phys- 31. Bischoff, C., J. Machetanz, and B. Conrad. 1991.
iologic implications. American Journal of Physical Is there an age-dependent continuous increase in
Medicine & Rehabilitation 76:488–93. the duration of the motor unit action poten-
15. Chan, R. C., and T. C. Hsu. 1991. Quantitative com- tial? Electromyography and Clinical Neurophysiology
parison of motor unit potential parameters between 81:304–11.
monopolar and concentric needles. Muscle & Nerve 32. Buchthal, F., P. Pinelli, and P. Rosenfalck. 1954. Action
14:1028–32. potential parameters in normal human muscle and
16. Walker, W. C., L. A. Keyser-Marcus, J. S. Johns, and R. their physiological determinants. Acta Physiologica
T. Seel. 2001. Relation of electromyography-induced Scandinavica 32:219–29.
pain to type of recording electrodes. Muscle & Nerve 33. Dimitru, D., J. C. King, and D. F. Stegman. 1998.
24: 417–20. Normal needle electromyographic insertional activity
17. Stålberg, E. 1980. Macro EMG, a new recording morphology: A clinical and simulation study. Muscle &
technique. Journal of Neurology, Neurosurgery, and Nerve 21:910–20.
Psychiatry 43:475–82. 34. Mitchell, C. W., and T. E. Bertorini. 2007. Dif-
18. Finsterer, J., and A. Fuglsang-Frederiksen. 2000. Con- fusely increased insertional activity: “EMG Disease”
centric needle EMG versus macro EMG I. Rela- or asymptomatic myotonia congenita? A report of
tion in healthy subjects. Clinical Neurophysiology 2 cases. Archives of Physical Medicine and Rehabili-
111:1211–15. tation 88:1212–13.
19. Gan, R., and J. F. Jabre. 1992. The spectrum of con- 35. Dumitru, D. 2000. Configuration of normal and abnor-
centric macro EMG correlations. Part II. Patients mal non-volitional single muscle fiber discharges. Clin-
with diseases of muscle and nerve. Muscle & Nerve ical Neurophysiology 111:1400–10.
15:1085–8. 36. Dumitru, D., and J. C. King. 2000. Hybrid fibrillation
20. Schulte-Mattler, W. J., D. Georgiadis, and S. Zierz. potentials and positive sharp waves. Muscle & Nerve
2001. Discharge patterns of spontaneous activity and 23:1234–42.
motor units on concentric needle electromyography. 37. Nandedkar, S. D., P. E. Barkhaus, D. B. Sanders, and
Muscle & Nerve 24:123–6. E. V. Stålberg. 2000. Some observations on fibrilla-
21. Stålberg, E., and L. Karlsson. 2001. Simulation of tions and positive sharp waves. Muscle & Nerve 23:
the normal concentric needle electromyogram by 888–94.
using a muscle model. Clinical Neurophysiology 112: 38. Kraft, G. H. 1990. Fibrillation potential amplitude
464–71. and muscle atrophy following peripheral nerve injury.
22. Stålberg, E., and L. Karlsson. 2001. Simulation of Muscle & Nerve 13:814–21.
EMG in pathological situations. Clinical Neurophys- 39. Barkhaus, P. E., and S. D. Nandedkar. 2006. “Slow”
iology 112:869–78. myotonic discharges. Muscle & Nerve 34:799–800.
23. Lorente de Nó, R. 1947. A Study of Nerve Physiol- 40. Logigian, E. L., E. Ciafaloni, C. Quinn, et al. 2007.
ogy. Studies from the Rockefeller Institute for Medical Severity, type, and distribution of myotonic discharges
Research, Vol. 132, 466–70. New York: The Rocke- are different in type 1 and type 2 myotonic dystrophy.
feller Institute for Medical Research. Muscle & Nerve 35:479–85.
24. Okajima, Y., Y. Tomita, R. Ushijima, and N. Chino. 41. Fournier, E., M. Arzel, D. Sternberg, et al. 2004.
2000. Motor unit sound in needle electromyography: Electromyography guides toward subgroups of muta-
Assessing normal and neuropathic units. Muscle & tions in muscle channelopathies. Annals of Neurology
Nerve 23:1076–83. 56:650–61.
25. Jiang, G. L., L. Y. Zhang, L. Y. Shen, J. G. Xu, 42. Roth, G. 1994. Repetitive discharge due to self-
and Y. D. Gu. 2000. Fibrillation potential amplitude ephaptic excitation of a motor unit. Electroencephalog-
to quantitatively assess denervation muscle atrophy. raphy and Clinical Neurophysiology 93:1–6.
Neuromuscular Disorders 10:85–91. 43. Costa, J., P. Graca, T. Evangelista, and M. de Carvalho.
26. Dumitru, D., King J. C., and D. F. Stegeman. 1998. 2005. Pain and calf hypertrophy associated with spon-
Endplate spike morphology: A clinical and simulation taneous repetitive discharges treated with botulinum
study. Archives of Physical Medicine and Rehabilita- toxin. Clinical Neurophysiology 116:2847–52.
tion 79:634–40. 44. Mateen, F. J., E. J. Sorenson, and J. R. Daube.
27. Conwit, R. A., B. Tracy, A. Cowl, et al. 1998. Firing 2007. Comparison of clinical methods for fascicula-
rate analysis using decomposition-enhanced spike trig- tion detection in amyotrophic lateral sclerosis. Muscle
gered averaging in the quadriceps femoris. Muscle & & Nerve 36:404–5.
Nerve 21:1338–40. 45. de Carvalho, M., and M. Swash. 1998. Fascicula-
28. Gunreben, G., and W. Schulte-Mattler. 1992. Evalu- tion potentials: A study of amyotrophic lateral sclero-
ation of motor unit firing rates by standard concen- sis and other neurogenic disorders. Muscle & Nerve
tric needle electromyography. Electromyography and 21:336–44.
Clinical Neurophysiology 32:103–11. 46. Hart, I. K. 2000. Acquired neuromyotonia: A new
29. Schulte-Mattler, W. J., D. Georgiadis, K. Tietze, and autoantibody-mediated neuronal potassium chan-
S. Zierz. 2000. Relation between maximum discharge nelopathy. American Journal of Medicine and Sciences
rates on electromyography and motor unit number 319:209–16.
estimates. Muscle & Nerve 23:231–8. 47. Maddison, P., K. R. Mills, and J. Newsom-Davis. 2006.
30. Sun, T. Y., J. J. Chen, and T. S. Lin. 2000. Analysis of Clinical electrophysiological characterization of the
motor unit firing patterns in patients with central or acquired neuromyotonia phenotype of autoimmune
peripheral lesions using singular-value decomposition. peripheral nerve hyperexcitability. Muscle & Nerve
Muscle & Nerve 23:1057–68. 33:801–8.
48. Hart, I. K., P. Maddison, J. Newsom-Davis, A. Vin- ology, ed. A. Fuglsang-Frederiksen, Vol. 2, Number 2.
cent, and K. R. Mills. 2002. Phenotypic variants of Skovlunde, Denmark: DANTEC Elektronik.
autoimmune peripheral nerve hyperexcitability. Brain 55. Pfeiffer, G. 1999. The diagnostic power of motor unit
125:1887–95. potential analysis: An objective bayesian approach.
49. Loscher, W. N., J. Wanschitz, K. Reiners, and S. Muscle & Nerve 22:584–91.
Quasthoff. 2004. Morvan’s syndrome: Clinical, labora- 56. Liguori, R., A. Fuglsang-Frederiksen, W. Nix, P. R.
tory, and in vitro electrophysiological studies. Muscle Fawcett, and K. Andersen. 1997. Electromyogra-
& Nerve 30:157–63. phy in myopathy. Neurophysiologie Clinique 27:
50. Friedenberg, S. M., and R. C. Hermann. 2004. 200–3.
The breathing hand: Obstetric brachial plexopathy 57. Lang, A. H., and V. S. Partanen. 1976. “Satellite poten-
reinnervation from thoracic roots? Journal of Neurol- tials” and the duration of motor unit potentials in
ogy, Neurosurgery, and Psychiatry 75:158–60. normal, neuropathic and myopathic muscles. Journal
51. Swift, T. R. 1994. The breathing arm. Muscle & of Neurological Sciences 27:513–24.
Nerve.17:125–9. 58. Buchthal, F. 1977. Electrophysiological signs of
52. Swift, T. R., R. T. Leshner, and J. A. Gross. 1980. myopathy as related with muscle biopsy. Acta Neurol
Arm-diaphragm synkinesis: Electrodiagnostic studies (Napoli) 32:1–29.
of aberrant regeneration of phrenic motor neurons. 59. Sonoo, M., and E. Stålberg. 1993. The ability of MUP
Neurology 30:339–44. parameters to discriminate between normal and neu-
53. Akaboshi, K., Y. Masakado, and N. Chino. 2000. Quan- rogenic MUPs in concentric EMG: Analysis of the
titative EMG and motor unit recruitment threshold MUP “thickness” and the proposal of “size index.”
using a concentric needle with quadrifilar electrode. Electroencephalography and Clinical Neurophysiol-
Muscle & Nerve 23:361–7. ogy 89:291–303.
54. Buchthal, F. 1991. Electromyography in the evaluation 60. Johnston, S. C. 2000. Prognostication matters. Muscle
of muscle diseases. In Methods in clinical neurophysi- & Nerve 23:839–42.
Chapter 27
Quantitative Electromyography
Benn E. Smith
INTRODUCTION Automated Analysis of Single MUPs

CHARACTERISTICS OF THE Automated Decompostion
MOTOR UNIT POTENTIAL EMG
CHARACTERISTICS OF THE Multiple Motor Unit Action Potential
RECORDING EQUIPMENT Decomposition-based
PROPERTIES OF MUPs quantitative EMG
EVALUATED USING
PROPERTIES OF INTERFERENCE
STANDARD ELECTRODES
PATTERN AND METHODS OF
Duration
INTERFERENCE PATTERN
Area
ANALYSIS
Amplitude
Method
Rise Time
Analysis
Spike Duration
Utility
Phases and Turns
Stability and Variation
TURNS AND AMPLITUDE
PROPERTIES OF MUPs ANALYSIS OF THE
MEASURABLE ONLY WITH INTERFERENCE PATTERN
SPECIAL ELECTRODES POWER-SPECTRUM
QUANTITATIVE ANALYSIS OF ANALYSIS
SINGLE MUPs AUTOMATED METHODS OF
Manual Analysis ANALYSIS OF SPONTANEOUS
Computer-Assisted Quantitative ACTIVITY
Analysis of User Selected MUP SUMMARY
INTRODUCTION This approach is fast, efficient, and reasonably

accurate if the electromyographer has a solid
In most clinical electromyography (EMG) lab- foundation in quantitative EMG and is experi-
oratories motor unit potential (MUP) activity enced from studying large numbers of patients
is analyzed subjectively or at most by semi- across the spectrum of age and disease. Inexpe-
quantitative EMG methods (see Chapter 26). rienced electromyographers are more prone to
451
error than their seasoned counterparts. What produce reproducible results when performed
are the principal types of error in this situation? by the same examiner at different times and
Examiner bias toward a particular diagnosis is by different examiners in the same or different
easily introduced during subjective MUP anal- institutions. Provided that the same techniques
ysis. A number of cases encountered regularly are used, comparisons of the results from dif-
in the practice of EMG are not readily solved ferent laboratories should be more robust, and
by the strategy of semiquantitative analysis. reporting of normal and abnormal results in the
For instance, it is common to encounter chal- literature should be more amenable to statis-
lenging cases with mild or equivocal disease tical analysis and critical assessment by others
in which the examiner, after the usual subjec- in the field. What would be the characteris-
tive analysis of MUP data, is in doubt about tics of an ideal method of quantitative EMG?
whether the findings are minimally abnormal Compared to conventional subjective or semi-
or normal. Another frequent situation faced quantitative EMG, such a technique would
by the electromyographer is borderline results be less time-consuming, be relatively easy to
in patients who may have mild myopathy or perform, be available on all existing EMG plat-
possible radiculopathy and questionably abnor- forms (without the need to acquire additional
mal MUPs in a few muscles. In addition, in hardware), be inexpensive, improve accuracy
some neuromuscular diseases EMG findings and reliability, be applicable to all muscles, and
are mixed, with both large and small MUPs. In come with rigorously collected normal control
these kinds of cases, more objective measures values across a broad range of ages and muscles
of MUP variables might help to resolve the in males and females. In addition, the tech-
uncertainties the examiner may have regarding nique should sample a large number of motor
the EMG findings. units in the muscle being studied. Ideally, it
These issues have led to efforts to use quan- would be helpful to the profession to have
titation to analyze electrophysiologic results. access to a variety of such techniques which
The quantitative approach focuses on measur- would assess EMG activity in different ways
ing electric activity from muscle fibers and and allow the extraction of data that have not
motor units as accurately as possible and to been available with previous methods.
record numerical values derived from these Although quantitative EMG methods have
precise measurements. These numerical data brought the field closer to many of these aspi-
can be statistically analyzed and graphed to rations, a number of disadvantages remain to
illustrate findings. Data from groups of normal be solved. These include increased costs for
subjects can be collected and compared with software and hardware, in some cases the need
data from patients suspected of having disease. for special equipment, and, depending on the
The advantages that quantitation holds over technique, extra time to learn and perform
subjective or semiquantitative EMG meth- the examinations. The introduction of any new
ods are many. The main aim of quantita- techniques, including quantitative EMG meth-
tion is to enable separation of patients into ods, requires the establishment of new normal
major diagnostic categories; for example, those control data, a process requiring considerable
with neuromuscular disease and those with- time and expense. Also, novel techniques must
out neuromuscular disease, and patients with be shown to be superior to the best methods
myopathies from those with neurogenic disor- currently available for routine clinical prac-
ders or defects of neuromuscular transmission. tice and be convincingly demonstrated to be
Similarly quantitation should markedly reduce applicable even in difficult cases with sub-
examiner bias, and therefore false-negative and tle abnormalities. Signals acquired from nerve
false-positive EMG studies resulting from this conduction studies, evoked potential testing,
systematically faulty approach. By virtue of the and needle examination recordings have all
ease of sorting and comparing sets of numeri- been quantified. Spontaneous activity and vol-
cal data, quantitation should allow distinctions untary activity on needle EMG are among the
to be made regarding severity of disease in most difficult to quantify due to the complexity
different patients with the same disorder or and rapid firing rates of the constituent wave-
in a single patient at different times during forms. The development of analog-to-digital
the course of the disease and its treatment. In conversion of needle examination data and the
terms of reliability, quantitative EMG should availability of low-cost, fast digital processors
Quantitative Electromyography 453
are leading to the development of automated volume of the cross-sectional diameter of the
techniques of evaluating MUPs and the inter- muscle. This is called the motor unit territory.
ference pattern. This chapter considers only The number of muscle fibers per given cross-
quantitative analysis of the findings on needle sectional area in the motor unit territory is
EMG. the fiber density. In a given muscle, motor
unit territories from different motor neurons
Purpose and Role of Quantitative EMG overlap and interdigitate, meaning that directly
adjacent muscle fibers may belong to separate
• Quantitative EMG measurements of MUPs
motor units.
record numerical values derived from pre- The synchrony of firing of muscle fibers in
cise measurements. a motor unit influences how the MUP appears
• Quantitative EMG generates normative
on the EMG screen. The synchrony of firing
data and compares with data from patients is determined by the (1) length, diameter, and
with suspected neuromuscular diseases. conduction velocity of the motor nerve termi-
• Quantitative EMG allows for repro-
nals; (2) the location and function of the neuro-
ducible results that can be compared at muscular junctions; (3) the diameter, conduc-
different times by different examiners and tion velocity, and membrane characteristics of
in different labs. the muscle fibers; (4) temperature; and (5) the
• Quantitative EMG allows accurate assess-
physical arrangement of muscle fibers in the
ment of improvement or deterioration in motor unit.
disease severity over time. In most muscles, individual muscle fibers
have only one neuromuscular junction, or end
plate. Exceptions are the extraocular mus-
CHARACTERISTICS OF THE cles and the extensor digitorum muscle in the
MOTOR UNIT POTENTIAL forearm.4 The location of the neuromuscular
junction is termed the end plate zone; this
The MUP, sometimes referred to as the motor varies from muscle to muscle. In the biceps
unit action potential, is the sum of individ- brachii muscle, for example, the end plate zone
ual potentials from muscle fibers within the is an irregular V-shaped band 5 mm wide, but
recording area of the electrode which are in the deltoid muscle, it forms an irregular
innervated by a single anterior horn cell. The sinusoidal pattern across the muscle. In the
distinguishing features of the MUP are the anterior tibial muscle, the end plate zone is at
result of several complex factors which reflect the periphery and is cone-shaped.
the characteristics of the motor unit. The num- The characteristics of the action potentials
ber of motor units in a given muscle varies by generated by individual muscle fibers are fun-
the anatomic site and size of the muscle. Esti- damental to the attributes of the MUP. Muscle
mates range from 100–500 motor units in a cell diameter and conduction velocity (typi-
typical muscle in a human extremity. Different cally 1.5–6.5 m/second) affect the muscle-fiber
types of lower motor neurons (LMNs) (types action potentials that summate to form the
I and II) comprise the 100–500 motor units. MUP. The amount and properties of the tissue
In general, type I LMNs are activated at low interposed between the measuring electrode
levels of force, have smaller nerve cell bodies and the discharging muscle fibers, including
and motor axons, and generate smaller MUPs. connective tissue, blood vessels, and fat, affect
Type II LMNs are larger, have larger axons, the MUP. These intervening tissues act as
produce larger MUPs, and are activated at a high-frequency filter and as such diminish
higher levels of force.1, 2 high-frequency signals and relatively enhance
The number of muscle fibers innervated by low-frequency activity.
each LMN is termed the innervation ratio.
This value varies by the LMN pool. In the Key Points
extraocular muscles, there are as few as 9
muscle fibers per motor unit, compared with • The MUP is the sum of potentials from
1900 muscle fibers per motor unit in the the muscle fibers innervated by a single
gastrocnemius muscle.3 The muscle fibers of anterior horn cell (a single motor unit)
one motor unit are distributed over 5–10 mm within the recording area of the electrode.
• The number of motor units in a muscle the exposed tip of the central wire is connected
ranges from 100 to 500. to the G1 input of the differential amplifier,
• Type I LMNs generate smaller MUPs at and the recording surface of the cannula is con-
lower force levels than type II LMNs. nected to the G2 input. This has the effect of
• The innervation ratio (number of mus- canceling out undesired electrical activity from
cle fibers innervated by each LMN) varies the surrounding muscle and other tissues. The
from a few to over 1000. concentric needle recording volume is roughly
• Motor unit territory is the cross section of hemispheric in shape. Commonly available
muscle in which an MUP is recorded with sizes are 25 mm (30–26 gauge), 50 mm (26
a rise time of less than 1.0 ms (overlaps gauge), and 75 mm (22–20 gauge). The wire
territories of other MUPs). is separated from the cannula by an insulator.
• The characteristics of the MUP are The tip of the electrode is machined to a 15◦
directly related to the electrical proper- angle, producing an exposed surface of the cen-
ties of tissues near the recording electrode ter wire that is a 150 μm × 580 μm ellipse.
surface. The 0.07 mm recording surface of the central
wire provides a very stable recording surface.
The recording area is smaller and directional
CHARACTERISTICS OF THE compared with the monopolar electrode. MUP
RECORDING EQUIPMENT amplitudes are smaller and shorter in dura-
tion than those measured with a monopolar
The construction of the recording electrode electrode. Because the recording territory is
determines many of the properties of the smaller, fewer fibrillations, fasciculations, and
recorded activity. Several different types of complex repetitive discharges are recorded.
recording electrodes are available (Fig. 27–1). There is less recording noise and the record-
Concentric and monopolar needle electrodes ing electrode surface area is more constant
are used most commonly. The concentric nee- than that of a monopolar needle electrode.
dle electrode is a bare 20- to 30-gauge hol- MUPs recorded from deeper in the muscle
low needle electrode with a thin wire core of appear larger than those recorded from near
approximately 150 μm diameter inserted down the surface.
the center of a hollow cannula, beveled at its This type of electrode holds a number
tip to expose an active oval recording surface of advantages: (1) recording EMG activity
measuring 580 μm × 125 μm. The activity at minimizing interference from surrounding
Figure 27–1. Various types of needle electrodes used in quantitative electromyography.

muscles, (2) the lack of need of a separate ref- activity from motion and the surrounding dis-
erence electrode, (3) the large number of mus- tant MUP activity. This is particularly promi-
cles with well-defined MUP size data, and (4) nent with low-frequency filter settings as low
widespread availability of inexpensive dispos- as 2 Hz. Increasing the low-frequency filter set-
able versions of the electrode for most EMG ting to 30 Hz reduces much of this noise, and
systems. settings of 500 Hz eliminate almost all of it.
The monopolar EMG electrode is a nee- However, increasing the low-frequency filter
dle insulated down to its 0.56–0.80 mm2 con- setting from 2 to 500 Hz drastically alters diag-
ical tip that serves as the recording surface nostic MUP variables. Increasing the amount
and is connected to the G1 input of the dif- of low-frequency filtering may add extra com-
ferential amplifier of the EMG machine. The ponents to the MUP such as the terminal neg-
G2 electrode is supplied by a surface record- ative afterpotential (Fig. 27–2). Stalberg et al.3
ing electrode. The active recording surface noted that the negative afterwave that follows
of the monopolar needle is larger than that the return to baseline of the positive afterwave
of the concentric needle electrode, but this is an artifact generated by the capacitance of
recording area may be more variable due to the low linear frequency filter, and that this
imprecise manufacturing or if the insulation is artifact can be minimized using a low linear
peeled back from use.. The monopolar elec- frequency filter setting of 2 Hz and should be
trode’s pick-up area is multidirectional and ignored in measurements of MUP duration.
spherical, which is why MUPs recorded with a The sensitivity settings of the amplifier can
monopolar needle electrode have longer dura- also have a marked effect on MUP variables,
tions and higher amplitudes than the same and in particular on measurements of dura-
potentials recorded with a concentric nee- tion (Fig. 27–3). The greater the sensitivity, the
dle electrode. In addition, monopolar tracings longer the apparent duration of the MUP. This
also record more noise from distant motor phenomenon secondarily affects other mea-
units. Frequently, monopolar EMG recordings sures, such as area, thickness, and size index.
are contaminated by activity from the sur- If the sensitivity is too great, the MUP over-
face G2 electrode. Monopolar electrodes are loads the amplifier and the amplitude cannot
less expensive and somewhat less painful for be measured. If the sensitivity is set too low,
patients. Because of their larger pick-up area, the MUP may not be detected at all.
monopolar electrodes may detect more fibril- Amplifier characteristics such as input
lation potentials, fasciculation potentials, and impedance, inherent noise level, amplifier
other spontaneous activity. recovery time, analog-to-digital sampling rate,
Single fiber EMG (SFEMG) electrodes are signal-to-noise ratio, and common mode rejec-
constructed of a very thin wire in a hollow tion affect the characteristics of the recorded
cannula. The wire is insulated from the can- EMG activity. Such variables limit the reliabil-
nula and exposed through a hole in the shaft ity, reproducibility, and ability to compare data
of the cannula. The recording surface is small obtained from different EMG equipment.5
(25 μm). The small recording area in com-
bination with higher low-frequency filtering Key Points
(500 Hz) allows individual muscle fibers of the
motor unit to be recorded in relative isola- • MUPs recorded with concentric needle
tion. Disposable concentric needle electrodes electrodes are of shorter duration and
are increasingly being used for SFEMG, ben- lower amplitude than those recorded with
efiting from yet a higher low-frequency filter a monopolar electrode.
setting (1 kHz) which is needed because of the • Isolated recordings of single muscle fibers
larger recording volume of the concentric elec- in SFEMG are facilitated by a small
trode, to diminish the effect of neighboring (25 μm) recording surface and a low-
muscle fibers which are firing with the fibers of frequency filter setting of 500 Hz.
interest. Single fiber EMG is discussed in more • Increasing the low-frequency filter setting
detail in Chapter 28. reduces the amplitude and duration of
The filter settings of the recording device MUPs.
can alter the appearance of the MUP. EMG • Higher amplifier sensitivity settings incre-
signals are contaminated by low-frequency ase the measured duration of an MUP.
Figure 27–2. Effects of varying low-frequency filter setting on recorded MUP morphology. The waveforms on the left
are from a single MUP recorded with a low-frequency filter of 2 Hz. The top tracing is an average of the triggered super-
imposition of 5 recurrences of the same MUP in the lower half. The waveforms on the right are the same MUP recorded at
the same position with a low-frequency filter setting of 30 Hz. Note changes in configuration, particularly the introduction
of a new phase, the negative afterpotential.
Figure 27–3. Effects of different sensitivities on measured duration of the same MUP. At higher sensitivities, measured
duration is longer.
PROPERTIES OF MUPs terminal components from the baseline is often

EVALUATED USING STANDARD very gradual and difficult to define. In distal
ELECTRODES lower limb muscles, MUP duration, amplitude,
and number of turns increase with age. In prox-
imal muscles and distal upper limb muscles,
Duration MUP variables are affected less significantly by
age. The greatest changes in duration occur
The duration of the MUP is the time between after 60 years of age. Duration also increases
the starting point and ending point of the by 5%–10% per degree centigrade decrease in
slow component of the MUP. Duration is temperature.
measured in milliseconds and includes the The duration of the MUP reflects the activ-
main spike and the initial and terminal parts ity of muscle fibers of the motor unit that
(Fig. 27–4). The deviation of the initial and are within 2.5 mm of the recording electrode
Figure 27–4. Commonly measured variables of the MUP.
Figure 27–5. Relative size of the territory of muscle fibers of a single motor unit compared with the pick-up area of a
concentric needle electrode. The amplitude of the MUP is determined by muscle fibers of the motor unit within 0.5 mm of
the recording surface. The area is determined by muscle fibers within 1.5–2 mm and duration by muscle fibers of 2.5 mm.
(Fig. 27–5). If the muscle-fiber density or a high-frequency filter with capacitance and
the territory of the motor unit changes, the resistance. Because slower frequency activity is
duration of the MUP will change. In gen- transmitted further through connective tissue,
eral, an increase in muscle-fiber density or muscle fibers at the periphery of the pick-
territory of the motor unit will increase dura- up area of the recording electrode primarily
tion. Diminished muscle-fiber density or motor generate the low-frequency, slow initial and
unit territory will decrease MUP duration. terminal components of the MUP.
MUP duration rises if there is increased vari-
ation in the diameter, length, or conduc-
tion velocity of the nerve terminal or muscle Area
fiber. A greater distance between the record-
ing electrode and contributing neuromuscu- The area of the MUP is the two-dimensional
lar junctions also increases MUP duration. territory under the curve of the waveform
The tissue intervening between the muscle (Figs. 27–4 and 27–5). It reflects the amount
fibers and recording electrodes functions as of functioning muscle fibers near the electrode
better than does the amplitude or other vari- Spike Duration

ables. Digital analysis allows accurate measure-
ment of the area. Simulation studies suggest Spike duration is the time from the first posi-
that the activity of 15–20 muscle fibers (from tive peak to the last positive peak of the main
a single motor unit) within a volume of radius component of the MUP. The spike is the sum
1.5–2 mm of the core of the concentric needle of the action potentials, of usually fewer than
electrode contributes to the area of the MUP 15 (and as few as 1–8) muscle fibers, closest to
being measured. the recording electrode.2 The spike may have
linked or satellite potentials separate from the
main spike. The terminal component of the
MUP is measured from the end of the main
Amplitude spike to the return to baseline (Fig. 27–4).
The terminal component is generated by action
Amplitude is determined by the contributions potentials traveling away from the recording
of action potentials within 500 mm of the elec- electrode along muscle fibers. Marking the end
trode (Fig. 27–4). This is typically 2–12 muscle of the slow return of the terminal compo-
fibers, but sometimes only 1 or 2. Although nent to the baseline is the most inconsistent
the diameter of the muscle fibers and the fiber measurement in semiquantitative and quanti-
density affect amplitude, this attribute is more tative EMG (Fig. 27–6). The imprecision in
dependent on the proximity of the electrode marking the end of the terminal component
to the muscle fibers than are duration or area. results in major variations in quantitative anal-
The amplitude is measured from the maxi- ysis of MUP morphology. Large amounts of
mum positive peak to the maximum negative low-frequency filtering (10–30 Hz) may add an
peak. Thickness is an MUP property derived extra component to the MUP, called the termi-
by dividing the area by the amplitude. Some nal negative afterpotential, which follows the
investigators have found that for the detec- return to baseline of the positive afterwave
tion of myopathies, decreased thickness is one (Fig. 27–2). This is an artifact generated by the
of the most sensitive variables of motor unit capacitance of the low linear frequency filter
morphology. which can be minimized using a low-frequency
filter setting of 2 Hz and should be ignored
in measuring MUP duration. If the terminal
negative afterpotential is included in measure-
ments, duration will be artificially prolonged.
Rise Time
Rise time is the duration of the rising phase of
the spike from the positive peak to the nega- Phases and Turns
tive peak (Fig. 27–4). Rise time, or rise rate,
is the best indicator of proximity of the source Most commonly, the overall shape of the MUP
(generator) of the potential to the recording waveform is triphasic, with an initial slow
electrode. downward deflection followed by a dominant,
Figure 27–6. Illustration of very gradual return to baseline of the terminal component of an MUP. The end is determined
more easily when the baseline is displayed, as in this illustration.
Figure 27–7. Example of a single large-amplitude, long-duration MUP firing at very rapid rates and randomly varying in
amplitude from one spike to the next.
rapid, upward deflection or spike, and then MUP, separated from the main component by
a slow downward deflection called the termi- a segment of flat baseline (Fig. 27–4). Many
nal component (Fig. 27–4). If the recording different terms have been used for satellite
electrode is close to the end plate zone, the potentials, including coupling discharges, par-
initial slow component is lost and the potential asite potentials, and linked potentials. These
becomes biphasic, beginning with the upward usually follow the main component, but may
spike component. Monophasic MUPs, with also precede it. Such satellite potentials are
only a single upward or negative phase, are typically excluded from measurements of MUP
much less common. An MUP may have the duration. Satellite potentials can be seen in
configuration of a “positive” wave, with an ini- 1%–3% of MUPs of normal muscles. These
tial large downward “positive” component fol- potentials occur in 10% of cases of normal
lowed by a low, long, late “negativity.” Such muscles, 12% of cases of neuropathy, 60% of
“positive” MUPs are recorded from damaged cases of old poliomyelitis, and 45% of cases of
muscle fibers, from the ends of the fibers or the myopathy.
tendon, or from the cannula (G2 electrode) of
the concentric needle electrode. Such positive
potentials should be excluded from the analysis Stability and Variation
of MUP properties.
The number of components of the waveform Motor unit potential variation is a general term
above or below the baseline are considered that refers to constantly fluctuating changes in
phases (Fig. 27–4). If there are more than amplitude or configuration of the MUP with
four phases, the MUP is considered polypha- successive discharges (Fig. 27–7). Jiggle refers
sic. The number of phases can be determined to variation in the position of phases or turns
by counting the portions of the waveform in the MUP relative to each other from one
above or below the baseline or by deter- discharge of the MUP to another (Fig. 27–8).6
mining the number of baseline crossings and These two terms reflect an instability of con-
adding one. Normal muscles may have 5%– duction along nerve terminals, across the neu-
15% polyphasic motor units. An increase in romuscular junction, or along muscle fibers.
the percentage of polyphasic motor units is Although classically thought of as a feature of
not only a very sensitive indicator but also a disordered neuromuscular transmission, MUP
very nonspecific indicator of neuromuscular variation may be less specific. Firing rate and
disease. recruitment are important factors for mea-
A turn is defined as a peak in the waveform surement in quantitative analysis; these are
of the MUP (Fig. 27–4). The number of turns is discussed in detail in Chapter 26.
determined by counting the number of positive
and negative peaks separated from the preced- Key Points
ing potential by some arbitrary amount, usually
50 μV or more. If the MUP contains more than • MUP duration reflects the activity of mus-
five turns, it is termed complex, or serrated. cle fibers that are within 2.5 mm of the
Satellite potentials are late components of the recording electrode.
Figure 27–8. Multiple firings of a single large-amplitude, long-duration, polyphasic complex MUP. Some of the
components block (second trace down) and others jiggle back and forth relative to each other.
• MUP amplitude is determined by the usually because of a neuromuscular junc-

action potentials of 2–12 muscle fibers tion disorder.
within 0.5 mm of the electrode.
• Area reflects the amount of functioning
muscle near the electrode better than
does the amplitude or other variables.
• Rise time is the best indicator of proximity PROPERTIES OF MUPs
of the recording electrode to the source of MEASURABLE ONLY WITH
the potential. SPECIAL ELECTRODES
• MUP duration, amplitude, and number of
turns increase with age. The size of the motor unit territory was mea-
• An increase in polyphasic motor units is sured in normal subjects and in patients with
a sensitive but nonspecific indicator of disease by Buchthal et al.7 using a multilead
neuromuscular disease. electrode made with multiple recording sites
• Satellite potentials are nonspecific find- along the length of a special needle electrode.
ings seen in normal muscles and in neu- The territory of motor units in limb muscles
romuscular disease. was found to be circular in cross section, with a
• Unstable MUPs vary in amplitude or diameter of 5.1 mm in the biceps brachii mus-
configuration with successive discharges, cle and 10 mm in the rectus femoris muscle.
Motor unit territory diameter and fiber den- the cannula insulated except for the terminal
sity tended to be reduced in patients with 15 mm (Fig. 27–1). This 15-mm length serves
myopathy. as the recording surface for the macro-EMG
The distribution of electric activity of the needle. In most limb muscles, the muscle
motor unit can be mapped in a cross-sectional fibers of a single motor unit are scattered
plane with the scanning EMG technique.8 This over a cross-sectional area of 5–10 mm. The
technique uses a standard concentric needle 15-mm recording area of the macro-EMG
electrode and a standard single fiber needle needle therefore encompasses the territory of
electrode. The activity from each of the two most motor units in limb muscles. There is
needles is recorded on a different channel also a 25-μm diameter wire electrode exposed
of the EMG machine using different filter on the shaft of the terminal part of the can-
and sensitivity settings. The single fiber nee- nula, 7.5 mm from the tip that enables record-
dle electrode is inserted into the muscle and ing of single muscle-fiber activity. Recordings
manipulated to a point where the activity of are made on two channels. The first channel
one muscle fiber is recorded. The concentric records activity from the 15-mm bare shaft
needle electrode is then inserted at a nearby (G1 electrode) and a surface electrode (G2).
point, perpendicular to the course of the mus- The second channel records single fiber activ-
cle fibers. The activity at the single fiber elec- ity from the 25-μm wire electrode (G1) and
trode triggers the sweep, and the MUP of the shaft (G2). The needle is inserted into the
the parent motor unit is recorded at the con- muscle and manipulated during minimal levels
centric needle. The concentric needle is then of contraction to a position at which the action
advanced through the muscle until no activ- potential of a single muscle fiber is recorded
ity is recorded from the motor unit under from channel 2. The activity in channel 2 acts as
consideration. Next, the concentric needle is a trigger, and the activity from channel 1 is then
connected to a mechanical motor drive that recorded and averaged over 60–80 ms. Next,
withdraws the concentric needle through the the needle is moved to a different site in the
muscle in small uniform steps. The MUP is muscle and the process is repeated. To obtain
recorded and averaged at each site and the an adequate sample from the muscle being
needle is withdrawn another step and the pro- examined, 20 different potentials are recorded
cess is repeated until the concentric needle from 20 different sites. Normal values have
is withdrawn from the territory of the motor been established.9
unit under study. Scanning EMG demonstrates In neurogenic disorders characterized by
that the shape of the MUP varies consider- denervation and reinnervation, the ampli-
ably across the motor unit territory. There tudes of the macro-EMG potentials are gen-
may be one, two, or more distinct areas of erally increased. In disorders of muscle or
activity, sometimes occurring with different myopathies, on the other hand, the ampli-
latencies. These have been called motor unit tudes are often low, particularly in suba-
fractions, and these are likely generated by cute myopathies. In chronic or long-standing
groups of muscle fibers, each innervated by a myopathies, amplitudes may be increased.
major intramuscular axonal branch of the par- Fiber density, blocking, and jitter are MUP
ent anterior horn cell. Scanning EMG allows attributes optimally evaluated with SFEMG.
measurement of the size of the motor unit Other special recording techniques can mea-
and assessment of the density and distribu- sure muscle-fiber conduction velocity, contrac-
tion of muscle fibers within the motor unit tion time, twitch time and tension, and the
territory. effects of fatigue, but these are not considered
Macro-EMG was developed in 1980 to in this chapter.
record the activity of the majority of muscle Akaboshi et al.2 have recently used a similar
fibers from a single motor unit. The concen- technique to decompose the interference pat-
tric needle records activity from a 2–3 mm tern at forces of up to 50% of maximal volun-
volume within the 5–10 mm diameter vol- tary contraction. They were able to determine
ume of the limb motor units. A macro-EMG the firing rates and recruitment frequencies
needle is a modified SFEMG needle, with of the motor units as well as isolate the units
for morphologic analysis. Unfortunately, this • Low-frequency filter for manual MUP
technique is too slow for routine clinical work. analysis is set at 2 Hz.
• Despite the high level of accuracy and
Key Points reliable normal control data, manual MUP
analysis is time-consuming.
• Scanning EMG measures the territory of
a motor unit and assesses the density and
distribution of muscle fibers within the
territory.
• Macro-EMG records a larger proportion Computer-Assisted Quantitative
of the muscle fibers in a single motor unit Analysis of User Selected MUP
than other needle recording electrodes.
• Fiber density can only be measured with The introduction of triggering and delay tech-
an SFEMG electrode. niques permitted the sampling of one MUP
over and over, even if other motor units were
active.15 This allowed more rapid collection of
QUANTITATIVE ANALYSIS OF data, with an improved signal-to-noise ratio.
Digitized signals allowed computer averaging
SINGLE MUPs and storage of waveforms and the measure-
ment of variables, such as area and thickness
Manual Analysis of the motor unit, not readily available with
the previous techniques. Such analyses are
Historically, quantitative EMG began with the not directly comparable to the results of the
measurement of MUP attributes by manual manual method of Buchthal et al.10–12, 14 and
analysis of paper tracings or photographs of Petersen and Kugelberg.13 The newer tech-
MUPs.10–14 This technique required minimal niques give a sampling bias which favors larger
activation of preferably only one but at the MUPs.
most three MUPs at a time. This limited the A commonly used technique for automated
evaluation to activity of type 1 muscle fibers analysis of single MUPs with computer-aided
and MUPs. The needle electrode was posi- methods is the quantitative EMG (QEMG)
tioned to obtain fast rise times. Photographs program (Viasys) (Fig. 27–9). To use the
were taken of several recurrences of the same original Buchthal normative data, 2–10 kHz
potential and measurements were made. The filter settings, degree of minimal activation,
amplitude was recorded as the maximal deflec- and standard sensitivity settings should be
tion on the screen possible to prevent the employed. Ideally, MUPs should have rise
loss of low-amplitude components and to pre- times less than 500 μs, although many experts
vent blocking of the largest component of think that a rise time of 1000 μs is accept-
the MUP. MUPs less than 50 μV in ampli- able. The examiner must be able to hold the
tude were excluded. The low-frequency filters needle immobile during data collection. The
were set at 2 Hz and high-frequency filters patient must be able to activate only one to
at 10,000 Hz. About 20–40 different record- three MUPs at a time and maintain a very
ing sites per muscle were evaluated, with the steady firing rate. The trigger must be adjusted
sites separated by at least 3 mm. The technique to isolate individual motor units. Some bias is
was accurate but quite time-consuming. Nor- introduced by triggering with the largest unit,
mal values have been published for a large causing the selection of large-amplitude long-
number of muscles. These data have been duration MUPs. This can be avoided by using
widely used for many years as the reference minimal activation, employing a dual trigger
standard. line to select the smaller units, and intention-
ally collecting a representative sample of the
MUP populations in the muscle being stud-
Key Points
ied. The dual trigger uses two lines that can be
• Manual MUP analysis assesses individual adjusted together to select peaks between the
MUPs at low levels of activation. two trigger lines, with sampling of the smaller
• Manual MUP analysis evaluates primarily MUPs. The MUP fires repetitively, and mul-
type I MUPs. tiple traces are superimposed (5–15 iterations
B C D
Figure 27–9. Computer-assisted quantitation of MUPs. A, Free running trace with four MUPs. Dashed line is trigger
level that selects MUPs for individual display; B, Two different MUP isolated by trigger; C, Third and fifth MUP have been
discarded manually; D, Remaining samples of a single MUP are superimposed to confirm their identity.
seem to work best). At least 20 MUPs are col- The analysis generates a report of the prop-
lected from different sites within the muscle. erties of all recorded MUPs and a sepa-
Different locations should be sampled so as rate listing of the quantitative attributes of
to avoid recording repeatedly from the same the simple and complex MUPs. These data
MUP and to increase the likelihood of find- include the duration, amplitude, turns, and
ing abnormalities localized to one part of the percentage of polyphasic MUPs as well as
muscle which may occur in some multifocal their mean, median, mode, variation, minimal–
diseases. maximal values, standard deviations, and confi-
An important issue with these systems is dence intervals.
how often the examiner must correct the mea- Despite the methodological and statisti-
surements made by the automated system cal advantages of quantitative EMG, the
(Fig. 27–10). The examiner must always check evaluation of motor unit attributes is subject to
the automated markers on every MUP collec- bias even with this more rigorous technique.
ted. As remarkable as the pattern recognition If a trigger line and delay are used, the sin-
capabilities of the QEMG software is, mark- gle trigger tends to bias the examination toward
ing the gradual onset and termination of the larger potentials. An excessive number of units
MUP is frequently subject to variability and firing at excessive force levels biases the exam-
error (Fig. 27–6). The duration marked is usu- ination toward the larger units. Because MUPs
ally greater at higher sensitivities, so the sen- appear larger when the needle is deeper in the
sitivity should be close to the 100 μV/division muscle, the depth of needle insertion can also
setting that is generally accepted when mark- skew the results. The slow initial and terminal
ing duration (Fig. 27–3). Background noise components of the MUP must be marked with
and other artifacts must be minimized. Relax- special care. As mentioned previously, the neg-
ation of neighboring muscles is particularly ative afterpotential should be excluded from
important. MUP measurements.
Figure 27–10. A, Display of sixteen of twenty MUPs that were isolated, selected and averaged as shown in Figure 27–9. B,
Mean duration, amplitude, and phases calculated for all isolated MUPs (MUP 1.20∗ column is not shown in the panel of 16).
Key Points (ADEMG) of Dorfman and McGill,16 multi-

• Computer-assisted quantitative EMG pro- ple motor unit action potential (MMUAP) of
Stalberg,17 and decomposition-based quanti-
grams focus on collecting a representative tative EMG (DQEMG) of Stashuk.18 Meth-
sample of the MUP populations in the
ods to further improve the QEMG con-
muscle being studied.
• Digitized signals with triggering and delay tinue to be developed.19 Using a variety of
approaches, such systems automatically detect
allow for storage and efficient analysis of
events that may be MUPs. Comparisons
multiple single MUP.
• Computer-assisted quantitative EMG among these methods have been reported.20–24
All current methods suffer from inability
programs typically include descriptive
to detect unstable MUPs and inability to
statistics of duration, amplitude, turns,
identify reduced numbers of axons (reduced
and percentage of polyphasic MUPs
recruitment).
as well as their mean, median, mode,
variation, minimal–maximal values, stan-
dard deviations, and confidence intervals.
• Comparison of quantitative EMG data Automated Decompostion EMG
with data collected by Buchthal requires
2–10 kHz filter settings and 100 μV sensi- Dorfman and McGill16 were the first to report
tivity. a clinically applied, automated program using
standard EMG needle electrodes called auto-
matic decomposition EMG or ADEMG. It
evaluates a single-channel interference pat-
Automated Analysis of Single tern at steady isometric contraction with forces
MUPs up to 30% of maximal voluntary contraction.
The analysis is performed in nearly real time
In recent years, a number of semiautomated (less than 1 minute), but requires a good
and automated methods have been developed deal of operator time to assure the quality of
that select and measure MUPs with minimal the recording. The program can extract up
human decision-making using standard nee- to 15 simultaneously active MUPs at a single
dle EMG electrodes. Some of these meth- site (practically 4–8). Later recruited MUPs
ods are now in common clinical use, while have different characteristics from the ini-
others continue to be developed. The major tially recruited MUPs. The mean MUP ampli-
methods are automated decomposition EMG tudes are significantly greater and increase
with increasing force. The mean MUP duration collected at each site. Graphs and reports of
is shorter and declines with increasing force. the data can be viewed or printed out. With
The number of turns increases with escalat- patient cooperation, a sample of MUPs can be
ing force. The mean firing rates are linearly collected quickly. The examiner should review
related to contractile force. The mean ampli- the individual MUPs accepted and their mark-
tude, duration, and number of turns increase ings to ensure accuracy. The collection of data
with advancing age, but mean MUP firing requires a quiet background and activation of
rates decrease with age. The results are pre- only a few MUPs at a time.
sented in numeric and graphic formats. The The electrode is inserted into the middle
results cannot readily be compared with the part of the muscle at different depths through
manual quantitative normative data. In its three separate skin insertions. No attempt is
current form, ADEMG is somewhat limited, made to position the electrode to record maxi-
in that MUPs smaller than 100 μV, unstable mal amplitude, but the electrode is positioned
MUPs, and MUPs with slow rise times can- so that at least some MUPs are “sharp” or crisp,
not be recorded. Normative data are available that is have short rise times. Force is varied
for only a few muscles. Studies demonstrating from slight to moderate muscle contraction,
the clinical usefulness of ADEMG are few and but no special equipment is required to mea-
focus on small numbers of subjects. For practi- sure force. The baseline should be clearly dis-
cal purposes, ADEMG is being used in only a cernible between signals. This analysis report-
very limited number of laboratories. edly requires fewer than 4–8 minutes per mus-
cle. Short segments (5 seconds) of activity
Key Points are recorded. At least 20 MUPs are recorded
from each muscle. MUPs are classified on
• ADEMG evaluates an interference pat-
the basis of shape variables by a multiple
tern at a steady isometric contraction and template matching technique. A minimum of
can extract up to 15 simultaneously active five recurrences of each MUP is averaged.
MUPs. About 2–5 MUPs can be recorded at each site.
• Low amplitude and unstable MUPs are
MUPs must be larger than 50 μV in ampli-
not recorded with ADEMG. tude and meet a rising phase criterion of less
than 30 μV/0.1 ms. Duration markers can be
adjusted manually and require manual correc-
Multiple Motor Unit Action tion in approximately 25% of the recorded
Potential MUPs. About 5%–15% of MUPs need to be
rejected because of background noise.
Bischoff et al.17 developed a technique of The program automatically measures ampli-
multi-motor unit action potential analysis tude, duration, spike duration, thickness,
(multi-MUAP analysis) that uses standard con- phases, and turns, and calculates mean val-
centric needle or monopolar needle electrodes. ues and standard deviations for each param-
It segments the record into epochs with and eter. Reference data for different age groups
without MUPs usually on the basis of a pre- are available for the deltoid, biceps brachii,
determined amplitude threshold. Presumptive first dorsal interosseous, vastus lateralis, and
MUPs are then sorted and classified by com- anterior tibial muscles. A good correlation
paring their wave shapes sequentially (tem- has been demonstrated between examina-
plate matching) using a priori match crite- tions performed by different examiners, repeat
ria. An MUP is accepted as an MUP when examinations by the same examiner, and side-
an arbitrary number of recurrences (2–10) to-side comparisons in the same subject. The
are confirmed. The recurrences may be aver- multi-MUAP analysis program takes a different
aged to produce a potential less affected by approach to defining the limits of normal by
random noise. The waveforms that do not using outliers, based on the assumption that
recur are considered to represent either noise in mild or early disease, abnormalities may
or superimposition of more than one MUP and be limited to a few MUPs rather than the
therefore are not accepted. The properties of entire MUP population. Such changes may be
the identified MUP are then measured and lost when averaged in with other MUPs. The
accumulated. Several different MUPs can be outlier limits were determined from the third
largest and third smallest value of a given vari- • Outlier analysis may provide another
able in normal subjects. The highest and lowest sensitive method to quantitate MUPs
values of these limits for the whole control which requires fewer than 20 MUPs to
group were chosen as the extreme outlier lim- assess whether the muscle in question is
its. The only outlier limit found to change with normal.
age was the amplitude of MUP in the ante-
rior tibial muscle, but not other muscles. Using
these criteria, none of the normal muscles
had more than two values outside the defined Decomposition-based quantitative
limits.
Multi-MUAP analysis detected abnormali-
EMG
ties in 25 of 31 cases of neuropathy.17 The size
Stashuk et al.18 have developed the most recent
index, amplitude, and duration were the most
automated MUP selection and measurement
frequently abnormal variables. The method
method, which is coming into more common
detected abnormalities in 6 of 8 cases of
clinical use as it becomes available on EMG
myopathy, with amplitude abnormalities more
machines. Waveforms are isolated and selected
common than duration. Outliers were as sensi-
by a multistep, mathematical algorithm from
tive as mean values in neuropathies and more
20 seconds of data recorded with a needle elec-
sensitive than mean values in myopathies.
trode from a single site in the muscle. The
Bischoff et al.17 pointed out that determining
user is provided ongoing measures of MUP
mean values may miss mild abnormalities of
rise time and MUP quality as the needle is
a few MUPs. An increased number of out-
moved to a new location. Up to eight simulta-
liers that indicate abnormality can be found
neously firing MUPs can be reliably recorded
only after evaluating a few MUPs, making it
from a single site. The electrode is moved to as
unnecessary to evaluate 20 or more units and,
many distinct locations in the muscle to obtain
thus, saving time. Podnar in studying a cohort
a satisfactory sampling of MUP, typically more
with fascioscapulohumeral muscular dystrophy
than 30. Measurements of the standard param-
pointed out the increased sensitivity of using
eters of MUP are available within 10 seconds
outlier analysis in addition to traditional mea-
of completing the recording at a site. The user
sures of MUP morphology.25
is given the opportunity to review and remark
There appear to be advantages of multi- or delete any MUP deemed unsatisfactory after
MUAP analysis. It allows sampling of a the primary data collection. Statistical parame-
large number of MUPs in a short time, is ters of the measurements are provided with the
reproducible, and allows MUP sampling at summary data.
levels of contraction greater than threshold. Studies of the reliability, reproducibility, and
It would reduce examiner bias. When edit- normal values have been published.18, 24, 26, 27
ing time is included, the analysis usually takes
They suggest that the method works as well
longer than 5 minutes per muscle to perform.
or better than other existing MUP quantitation
Multi-MUAP is available on a number of com-
programs. The program takes longer than an
mercially available EMG machines, and is in
EMG performed without formal quantitation
widespread use, particularly in Europe.
by an experienced electromyographer.
Key Points
Key Points • DQEMG rapidly and reliably isolates
• Multi-MUAP identifies MUPs on the multiple MUP from a moderate level con-
basis of a predetermined amplitude traction.
threshold and sorts the waveforms accord- • DQEMG provides live guidance to the
ing to their morphology (template match- electromyographer regarding the quality
ing). of the MUPs being recorded to allow opti-
• Multi-MUAP analysis is an automated mal selection of the data to be recorded.
technique that provides a relatively rapid • Prompt data display after each 20-second
and reliable method for analyzing MUPs recording allows quick assessment of the
which allows sampling of a large number quality of MUP selection and measure-
of MUPs in a short time. ment.
Figure 27–11. DQEMG data display from an anal sphincter in four columns named at the top of the figure. “Isolated
MUP” (first column) displays all the isolated MUPs superimposed to confirm their validity. “Averaged MUP” (second
column) is an average of all the isolated MUPs. “IDI Histogram” (third column) displays a histogram of the inter-discharge
intervals of each successive MUP to further assure their identity. MUP occurrence (fourth column) displays the time of
occurrence of each of the isolated MUP. The wiggly line plots the interdischarge intervals. Superimposed MUPs are not
measured (blank spaces).
PROPERTIES OF INTERFERENCE have been developed as efforts to quantify the

PATTERN AND METHODS OF interference pattern.
INTERFERENCE PATTERN
ANALYSIS Method
Individual MUPs can be identified and mea- The basic techniques of interference pattern
sured precisely at minimal levels of muscle analysis (IPA) are straightforward. A needle
contraction (Fig. 27–12). As the level of force electrode or surface recording electrode can
produced increases to the maximal voluntary be used, but the surface electrode techniques
contraction (MVC) force, the number and fir- are not clearly useful for diagnostic purposes.29
ing rates of active motor units increase. At the The needle electrode is inserted into an area of
same time, the size of the individual MUPs the muscle where MUPs have short rise times.
increases. As the number of MUPs increases, IPA is performed at different force levels in
it becomes difficult and ultimately impossible different systems. The test can be performed
for the unaided examiner to identify individual at variable or fixed percentages of MVC (usu-
potentials using standard electrodes and man- ally in the range of 30% of MVC), but this
ual quantitative methodologies. The observed requires special equipment. Analysis can be
activity in this situation is described as an inter- performed at fixed forces of 2–5 kg weights or
ference pattern in which individual MUPs are at forces varying from minimal contraction to
superimposed on others, resulting in a com- maximal contraction force. After the measure-
plex profile of summated and canceled wave- ments have been made at one site, the needle
forms from the activity of multiple different is moved to another distinct site and another
contributing MUPs.28 Automated techniques measurement is made. Ideally, a total of 20 to
Figure 27–12. The top tracing shows a partial interference pattern. The bottom tracing, an exploded view of a short
segment of the top tracing, illustrates some of the properties measured in quantitation of the interference pattern.
30 sites are measured through three or more the number of turns for a certain time inter-
needle tracks. val divided by the mean amplitude for that
same interval. Others have measured the max-
imal value of the turns/amplitude ratio for all
Analysis the sites tested and called this the peak ratio.
Peak ratio appears to be a useful measurement
Many different properties of the interference for distinguishing between normal subjects and
pattern have been measured. The number patients with neuromuscular disease.
of turns is measured frequently (Fig. 27–12).
Turns represent a change in signal direction of
at least 50 μV. Turns indirectly reflect the num- Utility
ber of active MUPs, the proportion of polypha-
sic MUPs, and the MUP firing rate. A turn may The advantages of IPA are that it incorpo-
reflect an MUP peak, an interaction between rates signals from more of the muscle under
overlapping MUPs, or noise. Baseline or zero study and samples activity from motor units
crossings are the number of voltage crossings of that are activated at higher force levels, usu-
the baseline per unit time. The time in millisec- ally type II motor units. The technique sam-
onds between turns or peaks can be measured ples a much larger amount of electric activity.
as time intervals or T’s (Fig. 27–12). Measure- Many of these techniques are available on com-
ment of the number of short time intervals mercial EMG machines. The evaluations are
or comparing the number of turns with short often rapid to perform. Most of the techniques
time intervals of 0–1.5 ms to those with longer are applicable to all muscles, and many can
time intervals of 1.5–5 ms and 5–20 ms seems be applied to uncooperative patients, such as
to have considerable clinical usefulness.30–32 small children.
Amplitude is measured as the potential dif- The disadvantages of IPA must be con-
ference between successive turns (Fig. 27–12). sidered. The variables measured cannot be
Cumulative amplitude is the total amplitude related in a simple and direct manner to
of turns over a certain time. Dividing the the properties of the constituent MUPs.
cumulative amplitude for a fixed time inter- The effects of summation and cancelation
val by the number of turns during that same of superimposed MUP activity are complex
interval defines mean amplitude. IPA data and difficult to understand. Large-amplitude
have also been expressed as a ratio called the potentials obscure activity from small MUPs,
turns/amplitude ratio, which is derived from and a few long-duration polyphasic MUPs may
give results similar to many short-duration at constant forces of 2 or 5 kg. In myopathies,

MUPs. Also, IPA systems often do not effec- these workers found that turns per second
tively measure activity less than 50–100 μV. increased and amplitude per turn decreased.
Technical factors such as system noise, fil- In neuronal or axonal loss and reinnervation,
ters, and analog-to-digital conversion rates can amplitude per turn increased and turns per
interfere. Normal values are available for a lim- second decreased.
ited number of muscles, but their usefulness Fuglsang–Frederiksen28 obtained more con-
in mild or borderline cases has not been eval- sistent results using a fractional contractile
uated carefully. The problem of the examiner force of 30% of the maximal voluntary force
entering data that are not well-analyzed into a and found it necessary to record from multiple
program that carries out complex analyses and sites (10 for each muscle). The most sensitive
provides data difficult to verify by other means variables were the ratio of the number of turns
introduces another element of uncertainty. to the mean turn amplitude in myopathies and
the decrease in the number of turns in neu-
Key Points ropathic disorders. IPA did not replace eval-
• IPA evaluates the properties of multi- uation of single MUPs but did increase the
ple, different, overlapping MUPs during a sensitivity and specificity of MUP analysis.
strong contraction. Hirose et al.33 evaluated the interference
• IPA incorporates and assesses both type I pattern at maximal contraction levels and
and type II motor units. found that the frequency of short-duration
• Of the variety of automated techniques turns of low amplitude was the most sensitive
that have been developed, the turns/ measure for differentiating myopathies from
amplitude system has gained the most neuropathies.
widespread use. Liguori et al.31 measured mean amplitude
• IPA is rapid to perform, useful in uncoop- (cumulative amplitude per time divided by
erative patients, and applicable to a num- number of turns per time), maximum ratio
ber of muscles for which normal data has (peak ratio, which is the maximal value for
been determined by age group. the ratio of the number of turns per time to
• Properties of the interference pattern the mean amplitude), and the incidence of
which have been measured include num- different time intervals between peaks (0 to
ber of turns, cumulative amplitude, and 1.5 ms intervals, 1.5 to 5 ms intervals, and 5 to
turns/amplitude ratio. 20 ms intervals) in normal subjects and patients
• IPA classifies EMG patterns into normal, with myopathies or neuropathies. They used
myopathic, or neurogenic, but is unable standard concentric needle electrodes inserted
to identify unstable MUPs, mixtures of at three different sites (proximal, medial, and
large and small MUPs, or abnormalities in distal) in the muscle and recorded activity at
recruitment without other changes. a total of 10 different recording sites at least
5 mm apart. The force of contraction gradu-
ally increased from 0 to maximum voluntary
contraction over 10 seconds, with the patient
TURNS AND AMPLITUDE resting 1 or 2 minutes between each contrac-
ANALYSIS OF THE tion. Of the patients with myopathy, 92% had
INTERFERENCE PATTERN abnormally high peak ratio values; the num-
ber of time intervals of short duration was
Probably the best-studied and most widely increased in 84%. All patients were classified
used method of automatic IPA was the turns correctly by using the peak ratio and time inter-
analysis technique developed by Willison.28 vals. In comparison, QEMG was diagnostic of
Initially, the number of turns and the mean myopathy in 72% of cases.
amplitude of the turns were measured from Of patients with neurogenic disorders, 86%
photographs of the interference pattern. Later, had low peak ratio values. The number
Fitch and Willison developed an electronic of short-time intervals was reduced in 48%
analyzer that could extract the data from the while the number of long-time intervals was
raw EMG signal. It counted the number increased. In all patients, the diagnosis was
of turns per unit time. The amplitude was made correctly using the combination of peak
measured between the turns. This was done ratio and time intervals, whereas QEMG was
diagnostic in 95%. This technique appears to 239 patients referred for quantitative MUP
be objective, fast, and reliable, but it takes at analysis. They found that for the detection of
least 20 minutes per muscle. myopathies or neuropathies, IPA with the Stal-
Stalberg et al.34 depicted IPA graphically berg technique was more sensitive and specific
in a scatter plot (Fig. 27–13) without careful than quantitative measurement of MUPs at
control of force. With a steady contraction for minimal effort (QEMG) or semiquantitative
1 second and rest for a few seconds between EMG. Other studies have shown that both IPA
epochs, force was varied from slight to near- and single MUP analysis of EMG recordings
maximal. Standard concentric or monopolar provide complementary data.20, 21
needles were used. The needle was moved to
a place in the muscle where a “spiky” pat-
tern was obtained. The filters were set at a Key Points
low linear frequency of 3.2 Hz and a high lin-
ear frequency of 8 kHz. The sensitivity was • Turns and amplitude analysis of the inter-
varied between 200 and 1000 μV/division to ference pattern compares the number of
allow adequate display of the activity without turns per unit time with the amplitude of
blocking. Twenty epochs were recorded in each the activity between the turns.
muscle. Turns per second were plotted against • In myopathies, turns per second increases
mean amplitude per turn. Using this tech- and amplitude per turn decreases; in neu-
nique in normal muscles, the data points fall rogenic processes the turns per second
within a so-called normal cloud. In myopathies, decreases and the amplitude per turn
the data points fall below the normal cloud, increases.
because of excessive turns and low ampli- • The parameters assessed with turns and
tude (Fig. 27–11). In neuropathic disorders, amplitude analysis can be displayed graph-
the data points fall above the normal cloud ically in a scatter plot, which eliminates
because of increased amplitude and a low turn the need to control for the force of
count. Nirkko et al.32 prospectively evaluated contraction.
Figure 27–13. Quantitative interference pattern analysis in a mild myopathy. MUP interference pattern (top) using mea-
surements of time intervals (bottom left), mean amplitude (middle), and the “cloud” (bottom right), note borderline low
amplitude and excess turns per second.
Figure 27–14. Examples of differences in frequency analysis of normal, myopathy, and neuropathy subjects. Note the
excess of high frequencies in myopathy and reduction of all frequencies in neuropathy (arrows).
POWER-SPECTRUM ANALYSIS the lower frequencies, probably because of the

long mean duration of the MUPs. Ronager
Another less commonly used method is the et al.35 examined power spectrum analysis in
application of fast Fourier transforms to the the biceps brachii muscle of normal subjects
interference pattern to obtain the power of the and patients with myopathies and neurogenic
interference pattern at different frequencies. disorders. They determined the mean power
This can be calculated and displayed across the frequency, the power at a variety of frequencies
entire frequency spectrum or given at specific (140 Hz, 1400 Hz, 2800 Hz, and 4200 Hz) rel-
frequencies (Fig 27–14).28, 35 Using fast Fourier ative to the total power, and the high/low
transformation, the IPA can be described frequency ratio (1400/140). At a force of
mathematically as a sum of sine waves of dif- 30% of maximal voluntary contraction, power-
ferent frequencies. The interference pattern spectrum analysis identified 55% of patients
can be characterized according to the density with myopathies and 64% of those with neu-
of power it contains at various frequencies; rogenic diseases. In myopathies, the relative
this is referred to as the power spectrum.36 power at 1400 Hz was increased in 50%. In
The precise shape of the interference pattern neurogenic disorders, there was a decrease
power spectrum depends on the geometry of in relative power at 1400 Hz and decrease
the recording electrodes and the characteris- in the high/low frequency ratio in 55% of
tics of the instruments. The shape is generally patients.
an inverted “U,” with a broad peak or plateau Frequency analysis of the EMG signal has
between 100 and 500 Hz, and the power falling been used as a tool for measuring muscle
off toward 0 at approximately 800–2000 Hz. fatigue.37 The mean and median frequency of
The lower frequency components (10–50 Hz) the power-density spectrum decay with fatigue
of the power spectrum tend to reflect the as the result of changes in muscle-fiber con-
firing rates of MUPs. The higher frequency duction velocity and motor unit activation.
components are more closely related to MUP The technique is, however, limited by poorly
morphology. defined normal values and wide ranges of val-
In myopathies, the interference pattern ues among normal subjects.
power spectrum contains relatively more
power in the higher frequencies, probably
Key Points
because of the short mean duration of the
MUPs. Frequency peaks are above 400 Hz. In • The power of the interference pattern can
neuropathies, the interference pattern power be expressed in component frequencies
spectrum contains relatively more power in for power-spectrum analysis.
• In myopathies, the interference pattern quantitatively grading fibrillation potentials in

power spectrum contains relatively more resting skeletal muscle as an EMG technique.
power in the higher frequencies while
in neuropathies there is generally more Key Points
power in the lower frequencies.
• While different methods to quantitate
MUP activity have been introduced at a
steady rate over the last several years, rig-
orous approaches to quantitate fibrillation
AUTOMATED METHODS OF potentials are yet to be developed.
ANALYSIS OF SPONTANEOUS • Non-EMG fields of clinical physiology
ACTIVITY such as electroencephalography, which
use spike detection algorithms, may hold
While quantitative methods for analyzing MUP promise for application in quantifying fib-
activity have steadily advanced over the last rillation potentials in denervated skeletal
several years, precise methods for assessing muscle.
spontaneous activity are in their early stages.
Drost et al. using multichannel high-density
arrays of surface recording electrodes have SUMMARY
developed a method of studying fasciculation
potentials. They found that both spatial and The quantitative evaluation of individual
temporal information about involuntary con- MUPs, decomposition of the interference pat-
tractions of different segments of the motor tern, IPA of turns and amplitude, and fre-
unit can be acquired noninvasively, yielding quency and power-spectrum analyses of the
new insights regarding the sources of individ- interference pattern are all useful techniques.
ual fasciculation potentials.38 They evaluate different features of the vol-
Methods for analyzing fibrillation poten- untary activity of skeletal muscle, with results
tials largely remain in the subjective realm. probably reflecting different properties of the
In considering what techniques being used in function and dysfunction of motor units. One
other fields might be considered for application method is not necessarily superior to the oth-
in quantifying fibrillation potentials in EMG, ers. It is likely that the more methods used, the
two existing methodologies come to mind: (1) closer the examiner will come to understand-
In electroencephalography, a number of algo- ing the functional structure of the muscle being
rithms have been developed for automated studied and whether or not a neuromuscular
detection of seizure spikes.39 Electrodes are disorder is present.
placed either on the scalp, on the dura, in The most critical limiting variable in using
subdural arrays, or in the substance of the these techniques is time. In a busy EMG prac-
brain. When regularly recurring spikes meet- tice, it is not possible to use all of these tech-
ing amplitude and rise-time criteria are identi- niques in one patient. It is the author’s prac-
fied, interventions can be instituted. (2) In car- tice to use semiquantitative EMG analysis first
diac electrophysiology, electrodes are placed because this can be performed on any muscle
on the chest wall or in proximity to the heart in and on a number of muscles relatively quickly.
order to detect, among other life threatening The clinical question is usually answered with
dysrhythmias, the regularly recurring narrow this technique. If doubt still remains, quanti-
complex pattern of ventricular tachycardia.40 tation of the MUP in the most suspect mus-
When this is identified in a cardiac inpatient, cle is performed with the QEMG program or
alarms sound to summon critical care person- the multi-MUAP analysis program. Cloud IPA
nel. In the case of patients with implanted with the Stalberg method, interpotential inter-
cardiac defibrillators, the identification of such vals, and the peak ratio can provide supportive
a perilous rhythm sets in motion an automated evidence.
direct current electrical impulse in an attempt The development of innovative methods of
to abort the dysrhythmia. Perhaps a variation signal analysis, including wavelet domain anal-
of one of these algorithms might be found ysis, and the use of artificial intelligence tech-
to have a diagnostic role in identifying and niques, including neural networks that can
learn, suggest that improved methods of anal- Journal of Neurology, Neurosurgery, and Psychiatry
ysis of EMG signals will become available as 12:124–8.
the field advances. These new approaches in 14. Sacco, G., F. Buchthal, and P. Rosenfalck. 1962. Motor
unit potentials at different ages. Archives of Neurology
combination with the availability of faster and (Chicago) 6:366–73.
cheaper digital microprocessors will probably 15. Stalberg, E., S. Andreassen, B. Falck, H. Lang, A.
lead to significant improvements in QEMG Rosenfalck, and W. Trojaborg. 1986. Quantitative
over the next few years. Methods to quantitate analysis of individual motor unit potentials: A propo-
sition for standardized terminology and criteria for
fibrillation potential activity in skeletal muscle measurement. Journal of Clinical Neurophysiology 3:
will likely remain beyond the horizon for some 313–48.
time. 16. Dorfman, L. J., and K. C. McGill. 1988. AAEE mini-
monograph #29: Automatic quantitative electromyog-
raphy. Muscle & Nerve 11:804–18.
17. Bischoff, C., E. Stalberg, B. Falck, and K. E. Eeg-
Olofsson. 1994. Reference values of motor unit action
potentials obtained with multi-MUAP analysis. Muscle
REFERENCES & Nerve 17:842–51.
18. Boe, S. G., D. W. Stashuk, W. F. Brown, and
1. Conwit, R. A., D. Stashuk, B. Tracy, M. McHugh, W. T. J. Doherty. 2005. Decomposition-based quantita-
F. Brown, and E. J. Metter. 1999. The relationship tive electromyography: Effect of force on motor unit
of motor unit size, firing rate and force. Clinical potentials and motor unit number estimates. Muscle
Neurophysiology 110:1270–5. & Nerve 31(3):365–73.
2. Akaboshi, K., Y. Masakado, and N. Chino. 2000. Quan- 19. Pattichis, C. S., C. N. Schizas, and L. T. Mid-
titative EMG and motor unit recruitment threshold dleton. 1995. Neural network models in EMG
using a concentric needle with quadrifilar electrode. diagnosis. IEEE Transactions on Bio-medical Engi-
Muscle & Nerve 23:361–7. neering 42:486–96.
3. Stalberg, E., S. D. Nandedkar, D. B. Sanders, and 20. Podnar, S., E. Zalewska, and I. Hausmanowa-
B. Falck. 1996. Quantitative motor unit potential Petrusewicz. 2005. Evaluation of the complexity of
analysis. Journal of Clinical Neurophysiology 13: motor unit potentials in anal sphincter electromyog-
401–22. raphy. Clinical Neurophysiology 116(4):948–56.
4. Trontelj, J. V., and E. V. Stalberg. 1995. Multiple inner- 21. Fuglsang-Frederiksen, A. 2006. The role of differ-
vation of muscle fibers in myasthenia gravis. Muscle & ent EMG methods in evaluating myopathy. Clinical
Nerve 18:224–8. Neurophysiology 117(6):1173–89.
5. Bromberg, M. B., A. G. Smith, and J. Bauerle. 1999. 22. Brownell, A. A., O. Ni, and M. B. Bromberg. 2006.
A comparison of two commercial quantitative elec- Comparison of three algorithms for multi-motor unit
tromyographic algorithms with manual analysis. Mus- detection and waveform marking. Muscle & Nerve
cle & Nerve 22:1244–8. 33(4):538–45.
6. Stalberg, E. V., and M. Sonoo. 1994. Assessment of 23. Calder, K. M., M. J. Agnew, D. W. Stashuk, and
variability in the shape of the motor unit action poten- L. McLean. 2008. Reliability of quantitative EMG
tial, the “jiggle,” at consecutive discharges. Muscle & analysis of the extensor carpi radialis muscle. Journal
Nerve 17:1135–44. of Neuroscience Methods 168(2):483–93.
7. Buchthal, F., C. Guld, and F. Rosenfalck. 1957. Mul- 24. Doherty, T. J., and D. W. Stashuk. 2003.
tielectrode study of the territory of a motor unit. Acta Decomposition-based quantitative electromyog-
Physiologica Scandinavica 39:83–104. raphy: Methods and initial normative data in five
8. Stalberg, E., and P. Dioszeghy. 1991. Scanning EMG muscles. Muscle & Nerve 28(2):204–11.
in normal muscle and in neuromuscular disorders. 25. Podnar, S., and J. Zidar. 2006. Sensitivity of motor unit
Electroencephalography and Clinical Neurophysiol- potential analysis in fascioscapulohumeral muscular
ogy 81:403–16. dystrophy. Muscle & Nerve 34:451–6.
9. Stalberg, E., and P. R. Fawcett. 1982. Macro EMG in 26. Boe, S. G., D. W. Stashuk, and T. J. Doherty. 2007.
healthy subjects of different ages. Journal of Neurol- Motor unit number estimates and quantitative motor
ogy, Neurosurgery, and Psychiatry 45:870–8. unit analysis in healthy subjects and patients with
10. Buchthal, F., C. Guld, and P. Rosenfalck. 1954. Action amyotrophic lateral sclerosis. Muscle & Nerve 36(1):
potential parameters in normal human muscle and 62–70.
their dependence on physical variables. Acta Physio- 27. Boe, S. G., D. W. Stashuk, and T. J. Doherty. 2006.
logica Scandinavica 32:200–18. Within-subject reliability of motor unit number esti-
11. Buchthal, F., P. Pinelli, and P. Rosenfalck. 1954. Action mate sand quantitative motor unit analysis in a distal
potential parameters in normal human muscle and and proximal upper limb muscle. Clinical Neurophys-
their physiological determinants. Acta Physiologica iology 117(3):596–603.
Scandinavica 32:219–29. 28. Fuglsang-Frederiksen, A. 2000. The utility of interfer-
12. Buchthal, F., and P. Rosenfalck. 1955. Action potential ence pattern analysis. Muscle & Nerve 23:18–36.
parameters in different human muscles. Acta Psychi- 29. Zwarts, M. J., G. Drost, and D. F. Stegeman. 2000.
atrica et Neurologica Scandinavica 30:125–31. Recent progress in the diagnostic use of surface EMG
13. Petersen, I., and E. Kugelberg. 1949. Duration and for neurological diseases. Journal of Electromyogra-
form of action potential in normal human muscle. phy and Kinesiology 10:287–91.
30. Liguori, R., K. Dahl, and A. Fuglsang-Frederiksen. 35. Ronager, J., H. Christensen, and A. Fuglsang-
1992. Turns-amplitude analysis of the electromyo- Frederiksen. 1989. Power spectrum analysis of the
graphic recruitment pattern disregarding force EMG pattern in normal and diseased muscles. Journal
measurement. I. Method and reference values in of the Neurological Sciences 94:283–94.
healthy subjects. Muscle & Nerve 15:1314–18. 36. Kamen, G., and G. E. Caldwell. 1996. Physiology
31. Liguori, R., K. Dahl, A. Fuglsang-Frederiksen, and and interpretation of the electromyogram. Journal of
W. Trojaborg. 1992. Turns-amplitude analysis of the Clinical Neurophysiology 13:366–84.
electromyographic recruitment pattern disregarding 37. Krivickas, L. S., A. Taylor, R. M. Maniar, E. Mascha,
force measurement. II. Findings in patients with neu- and S. S. Reisman. 1998. Is spectral analysis of the sur-
romuscular disorders. Muscle & Nerve 15:1319–24. face electromyographic signal a clinically useful tool
32. Nirkko, A. C., K. M. Rosler, and C. W. Hess. 1995. for evaluation of skeletal muscle fatigue? Journal of
Sensitivity and specificity of needle electromyography: Clinical Neurophysiology 15:138–45.
A prospective study comparing automated interfer- 38. Drost, G., B. U. Kleine, D. F. Stegeman, et al.
ence pattern analysis with single motor unit potential 2007. Fasciculation potentials in high-density sur-
analysis. Electroencephalography and Clinical Neuro- face EMG. Journal of Clinical Neurophysiology 24:
physiology 97:1–10. 301–7.
33. Hirose, K., M. Uono, I. Sobue, and M. Uono. 1975. 39. Gardner, A. B., G. A. Worrell, E. Marsh, et al. 2007.
Quantitative electromyography. Difference between Human and automated detection of high-frequency
myopathic findings and neuropathic ones. Electromyo- oscillations in clinical intracranial EEG recordings.
graphy and Clinical Neurophysiology. 15(5–6): Clinical Neurophysiology 118:1134–43.
431–49. 40. Singh, J. P., W. J. Hall, S. McNitt, et al. 2005. Fac-
34. Sanders, D. B., E. V. Stalberg, and S. D. Nanded- tors influencing appropriate firing of the implanted
kar. 1996. Analysis of the electromyographic inter- defibrillator for ventricular tachycardia/fibrillation.
ference pattern. Journal of Clinical Neurophysiology Journal of the American College of Cardiology 46:
13:385–400. 1712–20.
Chapter 28
Single Fiber Electromyography

INTRODUCTION Damaged Fiber

TECHNIQUE Split Fiber or Ephaptic Activation
Hardware Neurogenic Blocking
Software Pitfalls Unique to Stimulated SFEMG
Method of Activation
CLINICAL APPLICATIONS OF SFEMG
Measurement
Primary Disorders of
PITFALLS OF SFEMG Neuromuscular Transmission
General Primary Disorders of Muscle
Unstable Trigger Primary Neurogenic Disorders
False Trigger
SUMMARY
Incorrect Measurement Position
INTRODUCTION reinnervation associated with muscle or nerve

disease also impair neuromuscular transmis-
Single fiber electromyography (SFEMG) selec- sion. Thus, SFEMG results should be inter-
tively evaluates individual components of the preted carefully in the context of clinical and
motor unit. SFEMG measurements that are laboratory data as well as the results of nerve
useful clinically include fiber density, jitter, conduction studies and concentric needle elec-
and blocking. Fiber density is increased in a tromyography (EMG).
nonspecific manner in most neurogenic and
myopathic disorders, but it can be used in con- Purpose and Role of SFEMG
junction with jitter and blocking to make infer-
ences about the time course and progression • Identify the presence of a disorder of neu-
of disease. Jitter and blocking are quantita- romuscular transmission when other elec-
tive and sensitive measures of neuromuscular trophysiologic assessments are normal.
transmission efficiency. These measures iden- • Follow the course of mild defects of neu-
tify abnormalities early in the course of dis- romuscular transmission.
ease when standard electrodiagnostic studies • Identify the presence of signs of regen-
are normal. Although SFEMG is highly sensi- eration in neurogenic and myopathic
tive, it is nonspecific because denervation and disorders.
475
SFEMG is the most selective technique avail- EMG electrode to 500–1000 μm and, in many
able in clinical neurophysiology to study the cases, allows recording of potentials from sin-
motor unit. It is more selective than concentric gle muscle fibers or potentials summated from
needle EMG, which in turn is more selective two or three fibers. Thus, a reasonable esti-
than surface recordings. Selectivity refers to mation of jitter is obtained using a concen-
the ability to resolve individual generators of tric needle EMG electrode by increasing the
electrical activity within a volume conductor. low-frequency filter to 500 Hz.1, 2
In EMG, the generator is the action potential The SFEMG electrode has a circular record-
of a single muscle fiber. Selectivity depends on ing surface with a diameter of 25 μm, about the
three main factors: same as individual muscle fibers. When used
with a 500-Hz low-frequency filter, the effec-
1. The size of the electrode in relation to the tive recording distance is limited to 200 μm.
size of the action potential generator. The combination of small electrode size, low-
2. The filtering characteristics of the con- pass filter characteristics of muscle tissue, and
ducting medium. use of a 500-Hz low-frequency filter provide
3. The filter settings used in the recording the selectivity required to record single muscle
process. fiber action potentials. The standard SFEMG
needle has the recording electrode located
Surface electrodes record summated activity along the shaft of the cannula 3 mm proximal to
from many different motor units. A com- the nonbeveled side of the tip. This minimizes
pound muscle action potential (CMAP) is the chance of recording activity from muscle
recorded when motor units discharge syn- fibers that are damaged by the tip and fur-
chronously after supramaximal electrical nerve ther enhances selectivity by recording activity
stimulation. Assuming the electrodes are large directly adjacent to the recording electrode.
enough, the amplitude and area of the CMAP SFEMG can be performed with minimal
reflect the summated activity of the entire mus- voluntary activation or electric stimulation. The
cle. When the muscle is activated voluntarily, single muscle fiber action potential recorded
surface electrodes record activity from a large during SFEMG is typically biphasic, with an
number of motor units. These types of record- initial positive phase followed by a major neg-
ings provide information about firing patterns ative spike. When the electrode is close to the
of large motor unit groups but do not per- muscle fiber (rise time, 500 μs), the amplitude
mit selective recording of individual motor unit ranges from 500 μV to 10 mV with a duration
potentials. The lack of selectivity results from of 1–1.5 ms. The amplitude varies greatly with
the large size of the electrode relative to the minor changes in distance because of the small
size of individual muscle fibers and the ten- size of the recording electrode. The power
dency of the intervening tissue of the volume spectrum of the single fiber action potential
conductor to act as a high-frequency filter. ranges from 100 to 5000 Hz, with a peak from
The concentric needle EMG electrode, 1 to 2 kHz. Four types of measurement can be
with a recording surface of 150 × 580 μm, made during SFEMG:
is able to record selectively from individual
motor unit potentials containing an average 1. Fiber density reflects the packing den-
of several hundred muscle fibers. Due to the sity of muscle fibers within the recording
high-frequency filter characteristics of muscle area of the single fiber electrode. It corre-
tissue, most of the motor unit potential wave- lates with the degree of motor unit poten-
forms are generated from the 10 to 20 muscle tial polyphasia in concentric needle EMG
fibers located within several millimeters of the recordings. Fiber density is increased in
electrode. The selectivity of concentric needle neurogenic and myopathic disease.
EMG is limited by the large electrode size rel- 2. Jitter measures the latency variability of
ative to the diameter of a single muscle fiber muscle fiber action potentials within the
(25–100 μm). This can be overcome to some same motor unit. It reflects the variabil-
extent by increasing the low-frequency filter to ity in rise time of the end plate potential,
500 Hz, which attenuates low-frequency activ- providing a sensitive indicator of a mild
ity from distant muscle fibers. This narrows defect of neuromuscular transmission. Jit-
the recording area of the concentric needle ter is increased in disorders associated
Single Fiber Electromyography 477
with denervation and reinnervation as adds value only when routine nerve con-
well as primary neuromuscular junction duction studies and concentric needle
diseases. EMG are normal.
3. Blocking measures the intermittent loss • SFEMG can be performed with voluntary
of a regularly firing muscle fiber action activation or electric stimulation of motor
potential within a motor unit. This typi- units.
cally reflects the failure of the end plate • The most clinically useful measure of
potential to reach threshold in disorders SFEMG is jitter, the latency variability of
of neuromuscular transmission, but can muscle fiber action potentials within the
also occur in neurogenic disorders when same motor unit.
the impulse is blocked along a terminal
branch of the motor axon. Blocking is
present in moderate to severe disorders of
neuromuscular transmission, in disorders TECHNIQUE
associated with denervation and reinner-
vation of muscle, and in neuropathies
associated with impulse blocking in the Hardware
nerve terminal.
4. Duration measures the time between the NEEDLE ELECTRODE
first and the last muscle fiber action For standard SFEMG recordings, the nee-
potentials in a motor unit within record- dle consists of a stainless steel shaft (0.5 mm
ing distance of the electrode. This reflects diameter), with a single platinum wire down
differences in conduction time along the the center, opening onto a side port opposite
terminal axonal branch and muscle fiber. the beveled edge, and 3 mm proximal to the
Duration correlates with the duration of electrode tip (Fig. 28–1). The active record-
motor unit potentials recorded in con- ing surface is circular and 25 μm in diameter.
centric needle EMG and is increased in The shaft of the needle serves as the reference
neurogenic disease and in some chronic electrode. Single fiber electromyographic elec-
myopathies. trodes are expensive and, thus, are sterilized
and reused. The electrode should be inspected
Key Points under a dissecting microscope after being used
every 5–10 times and sharpened as needed.
• Selective isolation of a single muscle fiber Electrolyte treatment may be required if single
action potential depends on fiber amplitudes are low or noise is excessive.
◦ Recording electrode size down to Due to the high expense of SFEMG needles
25 μm and safety concerns related to reusability, many
◦ Optimal signal filtering at 500 Hz low-
frequency settings.
• SFEMG can be reasonably approxi-
mated with a standard concentric nee-
dle electrode using 500-Hz low-frequency
settings.
• SFEMG is very sensitive but not spe-
cific for mild disorders of neuromuscular
transmission.
• SFEMG is abnormal in myopathies, neu-
ropathies, and anterior horn cell disorders.
• SFEMG assists in the assessment of the
temporal profile and activity of neuromus-
cular diseases.
• SFEMG complements nerve conduction
studies (repetitive stimulation) and con- Figure 28–1. SFEMG electrode with active electrode
centric needle EMG in the evaluation (arrow) located along needle shaft proximal to tip of
of neuromuscular junction diseases, and needle.
laboratories now use disposable concentric Method of Activation

needle electrodes that have 5 μsec shorter
jitter MCD.1, 2 : but cannot measure fiber VOLUNTARY SFEMG
density.
Muscle fiber action potentials are isolated with
minimal voluntary muscle contraction. Approx-
AMPLIFIER imately 60% of the time, a single potential is
recorded. For the other 40%, from one to five
The high impedance of the SFEMG electrode potentials that are time-locked to the triggered
requires that the recording amplifier has high potential are recorded on the same sweep.
input impedance and low input capacitance, The level of activation should be adjusted to
which maintains the frequency response and maintain the triggered potential at a firing rate
prevents distortion of the single fiber potential. of 10–15 Hz. When activation is too vigorous,
Most standard clinical EMG preamplifiers and various technical problems can arise, includ-
amplifiers are sufficient for SFEMG recording ing overestimation of jitter (caused by unstable
as long as the low-frequency filter can be set to trigger and variation in amplitude of measured
500 Hz. potential) and false blocking (caused by alter-
nation of the trigger between a time-locked
and a single potential). The position of the
TRIGGER, DELAY, AND DISPLAY
needle is adjusted to maximize the rise time
An amplitude trigger placed on the negative of the triggered and time-locked potentials.
phase of the single fiber action potential initi- Minor rotational movements of the needle help
ates the sweep, and an analog or digital delay reduce noise from distant potentials and sepa-
allows the triggered potential and all potentials rate time-locked potentials that are fused with
that are time-locked to the triggered potential the triggered potential.
to be displayed. The trigger can be used with
voluntary, reflex, or electric activation. The
potential should be displayed at a sweep speed STIMULATED SFEMG
of 0.2–1 ms and a sensitivity of 20 μV–2 mV Muscle fiber action potentials are recorded
for accurate resolution of single fiber poten- after electric stimulation in the study of disor-
tials and measurement of fiber density, jitter, ders of neuromuscular transmission or inves-
blocking, and duration. When analog equip- tigation of certain reflexes (F waves, H reflex,
ment and manual measurement are used, a blink reflex, etc.). Electric stimulation is useful
counter and filming system that provide print- when patients are unable to cooperate with vol-
outs of five consecutive groups of 10 superim- untary SFEMG or when the effect of a change
posed sweeps are required to measure and to in firing rate on jitter and blocking needs to be
quantitate jitter. quantitated. Stimulated SFEMG has several
disadvantages. Fiber density cannot be mea-
sured accurately, and careful attention to tech-
Software nical problems is necessary to ensure accurate
and reliable measurement of jitter and block-
The majority of commercially available systems ing. In particular, minor movements in the
used to perform SFEMG are digital. Digital stimulating needle or “perithreshold” stimulus
conversion of the signal affords the advan- intensities can falsely elevate jitter while direct
tages of flexible display, automated measure- muscle stimulation can falsely depress jitter.
ment and calculation, storage, and reanalysis of In addition, different normal values are used
data. Each system has features that differ with for jitter in stimulated and voluntary SFEMG
regard to ease of use, degree of automation, because jitter reflects neuromuscular transmis-
display, and price. Key features that are essen- sion from a single end plate in stimulated
tial include automated jitter (mean consecutive SFEMG and from two end plates in volun-
difference, MCD) measurement of a minimum tary SFEMG (i.e., both the triggered and the
of 50 consecutive sweeps, storage capacity for measured potentials). Finally, electric stimula-
at least 30 pairs, and the ability to review and tion tends to selectively activate large diameter
reanalyze all the raw data. axons with low activation thresholds, unlike
voluntary SFEMG in which smaller diameter procedure is repeated for 30 separate triggered
axons are recruited initially and recorded pref- potentials to obtain an average fiber density for
erentially. The relative ease of obtaining data the entire muscle. In normal subjects, a sin-
in stimulated SFEMG may lull the inexperi- gle potential is isolated 60% of the time, two
enced examiner into a false sense of security, potentials 35%, and three or more potentials
but the technical issues described above make 5% of the time. Average fiber density ranges
stimulated SFEMG less accurate than volun- from 1.3 to 1.8 in normal persons younger than
tary SFEMG, even in the hands of experienced 70 years. Fiber density reflects the density of
electromyographers. muscle fibers in one motor unit within the
Electric stimulation can be applied to a recording area and corresponds most directly
branch of the nerve located outside the mus- to the number of turns seen on standard con-
cle or to an intramuscular motor branch. centric needle EMG. This feature of the motor
A monopolar needle is used as a cathode with unit potential is sometimes called complexity
another needle or surface electrode as the because each of these turns typically represents
anode. Very small currents (1–10 mA, 0.5 ms a separate fiber that contributes to the motor
duration) are used to activate a small num- unit potential. Fiber density has a less direct
ber of muscle fibers. The current and posi- relationship to the percentage of polyphasic
tion of the SFEMG recording electrode are motor unit potentials because in a polypha-
adjusted until a single muscle action potential sic potential a phase may include more than
with a rise time less than 500 μs and time- one turn. Satellite potentials seen in standard
locked to the stimulus is recorded. Jitter is recordings are also recorded as separate single
measured as the latency variability of the mus- fiber potentials in SFEMG.
cle fiber action potential in relation to the Fiber density is increased in disorders that
stimulus. When increased jitter or blocking is produce denervation and reinnervation. Thus,
observed, the stimulus is increased slightly. If increased fiber density is observed in most
the blocking disappears or the jitter lessens, motor neuron diseases and peripheral neu-
the abnormalities likely were caused by slight ropathies. The finding of increased fiber den-
variation in current strength above and below sity is particularly striking and out of propor-
the threshold of activation. This is a techni- tion to other changes on SFEMG early in
cal problem unique to stimulated SFEMG. the course of reinnervation, when differences
Reduced excitability resulting in latency pro- in conduction along regenerating nerve ter-
longation, increased jitter, and blocking can minals cause marked asynchrony of firing of
also be seen with prolonged stimulation at rates newly reinnervated muscle fibers and disper-
of stimulation greater than 20 Hz. This can be sion of single muscle fiber action potentials.
avoided by keeping stimulation rates less than Chronic disorders with minimal active den-
20 Hz except for brief 5- to 10-second inter- ervation but considerable compensated rein-
vals of stimulation at higher rates. Very low nervation have increased fiber density, with
jitter (MCD, 10 μs) is related to direct muscle only minimal increase in jitter and blocking. In
stimulation and should be ignored. contrast, subacute progressive disorders associ-
ated with ongoing reinnervation have a marked
increase in jitter and blocking and only a mild
increase in fiber density. Fiber density is also
Measurement increased in myopathies that are associated
with fiber splitting, degeneration, and regen-
FIBER DENSITY
eration. Therefore, fiber density can be used
Fiber density is defined as the average num- to quantitate the severity and time course of
ber or density of muscle fiber action potentials some neuromuscular disorders, but it cannot
within the recording area of the single fiber distinguish between neurogenic and myopathic
needle electrode. The needle is adjusted until disorders.
a single fiber action potential with an ade-
quate rise time (i.e., 500 μs) is isolated. The JITTER
number of muscle fiber action potentials with
an amplitude greater than 200 μV time-locked In voluntary SFEMG, jitter is defined as varia-
to the triggered potential are counted. This tion in the interpotential interval between two
single muscle fiber action potentials recorded (e.g., amyotrophic lateral sclerosis, some peri-
simultaneously from a single motor unit. Jit- pheral neuropathies, and some myopathies).
ter typically results from variation in the rise Jitter can be measured as the standard devi-
time and amplitude of the end plate potential ation of the interpotential interval, but because
at the neuromuscular junction. Jitter can also of the occasional occurrence of a gradual
result from variability of conduction along the change in the mean interpotential interval over
muscle membrane, but these factors produce time, it is more reliable to use the MCD:
negligible jitter at regular firing rates and when
the interpotential interval is less than 1 ms. In MCD = (IPI1 − IPI2 ) + (IPI2 − IPI3 )
normal subjects, there are small variations in −IPIN
the size of the end plate potential caused by + . . . IPIN−1
N−1
variations in the number of quanta of acetyl-

choline released from the nerve terminal. A where IPI is interpotential interval, and N is
smaller end plate potential has a slower rise the number of intervals measured. For maxi-
time and reaches threshold later than a larger mal reliability, the MCD should be calculated
end plate potential, so that the time from the from 50 or more consecutive interpotential
action potential in the nerve terminal to the intervals. The MCD is calculated for each pair
action potential in the muscle fiber varies by as and then the mean of 20 pairs is calculated and
much as 50 μs. The presence of this variation compared with normal values for the muscle
is evidence of a synapse between the activa- and the age of the patient.
tion site and the recording site. Two single fiber When the firing interval is variable or the
potentials with little or no jitter are either time- interpotential interval is large (i.e., 1 ms), the
locked by ephaptic (electric) activation of each variation in the conduction velocity of the mus-
other or recorded from a single muscle fiber cle fiber can affect jitter. This is described by
that has been split or otherwise distorted. The the velocity recovery function, which demon-
amplitude and the rise time of the end plate strates that when the preceding discharge
potential are a direct reflection of the safety interval is short the conduction velocity of
margin of neuromuscular transmission. Any the muscle fiber is faster on the subsequent
disorder of neuromuscular transmission that discharge. If there is variability in the dis-
decreases the safety margin will increase jitter charge frequency and the interpotential inter-
(Fig. 28–2). This includes primary disorders of val is long (i.e., if the velocity recovery function
neuromuscular transmission (e.g., congenital is beginning to affect jitter), then the mean
and autoimmune myasthenia gravis, Lambert– sorted difference (MSD) is a better representa-
Eaton myasthenic syndrome, and botulism) tion of jitter than the MCD. The MSD is calcu-
or secondary disorders that reduce the safety lated by first sorting the potentials in ascending
margin by producing denervation and rein- order of size of the interpotential interval and
nervation with immature nerve terminals then calculating the MCD. If the MCD:MSD
ratio is greater than 1.25, the variability of the
discharge frequency is affecting jitter and the
MSD should be used.
The best point to measure jitter is on the
steep rising phase of the potential close to the
baseline crossing. Movement of the triggered
potential (amplitude jitter) or contamination of
the baseline with other potentials or one poten-
tial riding upon another will artificially increase
jitter.
In stimulated SFEMG, jitter is measured
as the MCD of the interpotential difference
between the stimulus artifact and the sin-
Figure 28–2. SFEMG displaying multiple consecutive gle muscle fiber potential. The MCD reflects
traces from a single pair superimposed. The sweep is trig-
gered on potential “A.” The variability of the latency of
the jitter from a single end plate rather than
potential “B” measured between “1” and “2” in microsec- the pair of end plates measured in voluntary
onds represents jitter. SFEMG. The velocity recovery function of
muscle does not affect jitter measurement dur- plate potential produces increased jitter. This
ing stimulated SFEMG because there is no occurs not only in disorders of neuromuscu-
random variation in discharge frequency. lar transmission, such as myasthenia gravis,
Several software applications are available to but also in disorders with ongoing reinnerva-
automate the measurement of jitter and block- tion or regeneration of muscle fibers, such as
ing. Each program provides a graphic display amyotrophic lateral sclerosis and polymyositis.
of the calculated MCD for 50–100 consecu- Thus, abnormalities of jitter are not diagnos-
tive sweeps and the ability to store, review, tic of a specific disease of the neuromuscular
and reanalyze each individual sweep collected junction but must be considered in relation
in order to ensure accuracy of the data col- to findings obtained with standard electro-
lected. Manual calculation of MCD requires a physiologic recordings. Abnormalities of jitter
counter that captures and displays 50 consecu- can occur without clinical weakness in the
tive sweeps in five groups of 10 superimposed muscle.
images. The variation of the interpotential dif- Jitter is a function of the variation in synaptic
ference can be measured directly from each potential size; therefore, it is present in record-
of the five groups and the MCD calculated ings that include other synapses. F waves are
directly with a conversion factor. a result of antidromic activation of the anterior
Normal values for MCD vary with age and horn without a central synapse so that F-wave
the muscle. In the Mayo EMG laboratory, the jitter is approximately the same magnitude as
normal jitter in the extensor digitorum commu- with voluntary motor unit potentials. In con-
nis, the most commonly recorded limb mus- trast, H reflexes, which include a synapse in
cle, is between 16 and 34 μs (upper limit of the spinal cord in addition to the neuromus-
normal for a single pair is 55 μs) for persons cular junction, have normal jitter of two to
younger than 60 years. For persons older than three times that of voluntary motor unit poten-
60 years, the upper limit for MCD increases to tials. Other more complex reflex phenomena,
43 μs, and it is normal to have up to two pairs such as the blink and flexion reflexes, have
with an MCD greater than 55 μs. The jitter correspondingly larger amounts of jitter.
is smaller in facial muscles than in limb mus- Reliable jitter measurement depends on the
cles. Facial muscles are also less susceptible to presence of steep rising phases of both of the
local trauma, which can increase jitter indefi- potentials from which to measure the interpo-
nitely. The MCD for the frontalis muscle is 23– tential interval. If the two single fibers have a
31 μs in normal subjects younger than 60 years short interpotential interval they will overlap,
and 23–35 μs for those 60 years and older. obscuring the steep rising phases. An alterna-
Similar normal values have been defined for tive measure of the variation in the waveform
these and other muscles.3 Normal MCD val- resulting from the overlap of the two fiber
ues for stimulated SFEMG are approximately potentials is called jiggle. Jiggle measures the
80% of the value obtained with voluntary change in shape of the summated fiber poten-
SFEMG of the same muscle in the same age tials that results from variation in their interpo-
group. tential intervals. Amplitude measurements at
In normal muscle, jitter is not identifiable each point in time of the potential are used
on standard concentric or monopolar needle to calculate consecutive amplitude differences
EMG. However, if a motor unit potential is and cross-correlation coefficient of consecutive
recorded with a standard concentric needle discharges. Jiggle measurements are particu-
electrode at a low-frequency filter of 500 Hz larly useful for recordings from concentric nee-
and a sweep speed of 1 or 2 ms/cm, jitter can dle electrodes where the single fiber potentials
be identified. Quantitative measurements of are less well distinguished than with SFEMG
jitter with a standard concentric needle elec- electrodes.4, 5
trode are somewhat larger than those recorded
with a single fiber electrode, making the study
BLOCKING
less specific in the detection of mild defects of
neuromuscular transmission. In a normal muscle, the end plate potential
Because jitter is the result of fluctuations always reaches threshold and initiates a single
in the amplitude of the end plate poten- fiber action potential. Therefore, when mul-
tial, any disorder that decreases the end tiple single fiber potentials are found, they
Normal elderly subjects display occasional

blocking in some muscles. In fact, in a study
of 20 pairs of single fiber potentials, if a sin-
gle pair exceeds the limit of normal jitter or
displays blocking (or both), many electromyog-
raphers do not use those changes alone to inter-
pret the study as abnormal. Blocking begins to
occur when the jitter in a pair has increased
to 80–100 μs. In disorders of neuromuscular
transmission, amplitude decrement of a CMAP
with 2-Hz repetitive stimulation is caused by
blocking, as is moment-to-moment variation
observed on standard concentric or monopolar
needle EMG.
Blocking is observed in disorders of neu-
romuscular transmission such as myasthenia
gravis, and, when present, it provides evidence
of a defect severe enough to produce weakness
either at rest or with exertion. As with jitter,
however, blocking can occur in other disor-
ders in which neuromuscular transmission may
be impaired, such as amyotrophic lateral scle-
rosis, polymyositis, and ongoing reinnervation.
For blocking to be considered evidence of a
disorder such as myasthenia gravis, it should
Figure 28–3. Five consecutive traces of three muscle be found in the absence of other electro-
fiber action potentials isolated during voluntary SFEMG
recording. The sweep is triggered on the first potential (T).
physiologic signs of neurogenic or myopathic
The later potentials (1 and 2) are in the same unit as the disease.
triggered potential. The latency of potentials 1 and 2 with
respect to potential “T” vary slightly with each trace (jitter)
until the fourth trace when potentials 1 and 2 disappear DURATION
completely (blocking).
The interval between the first and the last
potential of multiple single fiber potentials
occur with each discharge of the motor unit recorded from a motor unit has been mea-
potential. Blocking occurs when an end plate sured as duration. Duration is the total time
potential does not reach threshold or when from the first to the last potential averaged
conduction fails in the nerve terminal, result- for all multiple potentials recorded. An alterna-
ing in a loss of a single fiber potential dur- tive measure is mean interspike interval, which
ing one or more discharges of the motor unit divides the duration by the number of sin-
(Fig. 28–3). Blocking is measured as the per- gle fiber potentials in each discharge. Mean
centage of discharges of a motor unit in which interspike interval and duration increase when
a single fiber potential is missing. A motor unit the activation of individual single fiber poten-
in which a single fiber potential did not fire tials is dispersed in time. Factors that may
half the time would have 50% blocking. Block- contribute to dispersion include reduced syn-
ing for a particular muscle is expressed either chrony of firing, anatomical dispersion of end
as the percentage of 20 fiber pairs that show plates along muscle fibers, and differences in
any blocking or as the total percentage of dis- conduction along the terminal axon or mus-
charges in the 20 pairs that displayed blocking. cle fiber. Duration is used less frequently than
For example, if 20 potential pairs each dis- other SFEMG measurements because similar
charge 50 times and blocking occurs a total of information is obtained from the duration of
20 times in two of the pairs, 10% of the pairs motor unit potentials and the examination for
show blocking and 2% of all discharges have satellite potentials during standard concentric
blocking. or monopolar needle EMG.
Key Points increases jitter. Changes in firing rates affect

the velocity recovery function of the muscle
• A special needle electrode is used to
fiber, which also increases jitter.
record SFEMG although jitter can be
accurately recorded with a concentric
needle electrode with a shorter MCD.
Unstable Trigger
• A high input impedance amplifier is
Because all measurements are taken in refer-
required to obtain high-quality SFEMG
ence to the triggered potential, stability of this
recordings.
• SFEMG can be done with either volun- potential forms the cornerstone of voluntary
SFEMG. Gross instability results in intermit-
tary activation or electrical stimulation of
tent loss of the triggered potential and inter-
the target muscle.
• Fiber density can only be recorded with ruption of the SFEMG recording. When this
occurs occasionally, the sweep can be deleted
a single fiber needle and estimates the
from analysis, but frequent loss of the triggered
packing density of muscle fibers.
• Jitter represents time variation of the rise potential makes it impossible to record jitter
or blocking accurately. Even minor fluctuation
time of the end plate potential and repre-
of the amplitude of the triggered potential can
sents a sensitive measure of neuromuscu-
add as much as 10 μs to the MCD. Superimpo-
lar transmission efficiency.
• Blocking occurs when end plate potentials sition of consecutive sweeps is an effective way
to check for stability of the triggered poten-
intermittently fail to reach threshold for
tial (Fig. 28–4). Electrode movement, exces-
firing of a muscle action potential.
• Blocking reflects more severe neuromus- sive activation, or variation in firing rates can
produce an unstable trigger.
cular transmission failure than increased
jitter alone.
False Trigger
PITFALLS OF SFEMG Occasionally in voluntary SFEMG, two mus-

cle fiber action potentials close to the electrode
have similar firing thresholds, amplitude, and
General configuration. When this occurs, the trigger
Most errors in SFEMG result from needle
movement, excessive activation, or variability
in the firing rate of the muscle action poten-
tial. The selectivity that affords the single fiber
electrode the ability to focus the recording on
one or two muscle fibers also renders SFEMG
extremely sensitive to small movements of the
needle electrode. Minor movement along the
long axis as well as angulation or rotation of the
cannula often produces marked changes in the
amplitude or configuration of the muscle fiber
action potential. Variation in amplitude and
configuration leads to errors in jitter and block-
ing measurements. The ability to hold the elec-
trode motionless is the greatest technical skill
required to produce reliable SFEMG record-
ings. The second great technical challenge in
voluntary SFEMG is to maintain minimal acti-
vation at relatively stable firing rates. Excessive
noise produced by nearby muscle fiber action Figure 28–4. Superimposition of sweeps during
potentials may distort the baseline or the sin- SFEMG recordings. Poor superimposition traces in pair 2
gle fiber potentials of interest, which artificially demonstrates an unstable triggered potential.
may “jump” from one potential to the other. If

one of the potentials is a double—i.e., it has a
second potential time-locked to the first—and
the other potential is a single, alternation of
the trigger between the two potentials gives the
false impression of blocking (Fig. 28–5).
Incorrect Measurement Position

Jitter is best measured from the inflection point
of the second potential, which typically occurs
as the rise time of the potential crosses the
baseline. In voluntary SFEMG, this position
requires minimal activation because any noise
that moves the baseline will affect the mea-
surement of jitter. The farther the measure-
ment point is moved away from the inflection
point, the more jitter is affected by changes in
amplitude of the potential produced by needle
movement. Some automatic algorithms mea-
sure jitter from the peak rather than the ris-
ing phase of the potential (Fig. 28–6). This
method is also useful when multiple poten- Figure 28–5. False trigger. The trigger alternates
tials are superimposed on one another. The between two potentials of similar size and shape. Potential
#1 is a single potential, whereas potential #2 is time-locked
accuracy of the peak jitter method depends on to a second potential (i.e., part of a double). Alternation
the ability of the examiner and the software to of the trigger between potentials #1 and #2 gives the false
clearly define stable peaks of the waveforms. impression of blocking.
Figure 28–6. Jitter measurement. Left, Jitter measurement from the negative slope of the potential. Right, Jitter
measurement from the peak of the potential.
Damaged Fiber which occurs at the neuromuscular junction,

is caused by immaturity of recently reinner-
Recording from a damaged fiber often pro- vated end plates. The second, which occurs
duces a broad positivity, followed by a negative within the terminal branch of the motor axon,
phase of shorter duration and lower amplitude. is caused by a prolonged refractory period
The negative phase sometimes has the appear- related to branching and demyelination of the
ance of a second potential, which usually has nerve terminals. Axonal blocking can be recog-
low jitter with the parent potential (Fig. 28–7). nized when more than one muscle fiber action
This has been referred to as a false double and potential disappears or blocks simultaneously
should not be included in jitter calculations. during either voluntary or stimulated SFEMG
(Fig. 28–9). This phenomenon has also been
referred to as concomitant blocking.
Split Fiber or Ephaptic Activation
Jitter less than 10 μs generally indicates the Pitfalls Unique to Stimulated
absence of a neuromuscular junction between SFEMG
the two recorded potentials (Fig. 28–8). This
occurs when the potentials are generated from Stimulated SFEMG is useful when evaluat-
the same muscle fiber that has split into two ing poorly cooperative patients or when it is
sections or when one fiber is activated by important to measure the effect of discharge
an adjacent fiber by ephaptic transmission. In frequency on jitter and blocking. A disadvan-
stimulated SFEMG, low jitter occurs when tage of stimulated SFEMG is the inability
direct muscle stimulation occurs. These values to measure fiber density. Other problems are
should be eliminated from jitter calculations. technical in nature and must be accounted
for to ensure accuracy of jitter and blocking
Neurogenic Blocking
There are two mechanisms by which block-
ing occurs in neurogenic disorders. The first,
Figure 28–7. Damaged fiber. Jitter from the sec-

ond component should not be included in calculations
of MCD.
Figure 28–8. Low jitter indicates recording from a split

fiber or one activated by ephaptic transmission. MCD, Figure 28–9. Neurogenic blocking. The second and
mean consecutive difference. third potentials block concomitantly.
measurements. The first technical pitfall is stimulation studies, it can be learned read-
related to direct muscle stimulation. This is ily and applied with a minimum of special-
fairly easy to recognize because of the result- ized equipment. However, in uncooperative or
ing small jitter. A more subtle but equally tremulous patients, reliable voluntary SFEMG
important technical problem is related to a can be time-consuming, so the selection of
false increase in jitter and blocking that occurs patients and muscles for SFEMG requires
when the stimulus intensity is close to the consideration.
axonal threshold of the single fiber poten-
tial being examined. Increasing the stimulus
intensity slightly identifies this problem, which
AUTOIMMUNE MYASTHENIA
should eliminate blocking and reduce jitter if
GRAVIS
the problem is technical. The axonal threshold
may also increase with prolonged stimulation The abnormalities found on SFEMG in
or stimulation at rates in excess of 20 Hz. Thus, patients with myasthenia gravis were demon-
the effect of small increases in stimulus inten- strated clearly in the early studies of Ekstedt
sity should always be determined before jitter and Stålberg and are reviewed in the text-
and blocking are measured during stimulated book by Stålberg and Trontelj,12 Single Fiber
SFEMG. Electromyography. Both jitter and blocking are
increased in proportion to the severity of clin-
ical involvement, with greater abnormality in
Key Points weaker muscles. However, in a given muscle,
• Most errors in SFEMG result from needle and even among the end plates of a single
movement, excessive activation, or vari- motor unit, there is marked variation in the
ability in the firing rate of the muscle amount of jitter in different fiber pairs. In a
action potential. single muscle, some fiber pairs may be entirely
• Technical errors that result in falsely normal and others may be grossly abnormal,
increased jitter or blocking include an with frequent blocking.
unstable trigger, a false trigger, neuro- Most important among the features that
genic blocking, and an incorrect position Stålberg, Ekstedt, and Broman9 noted was
of measurement marker. the presence of abnormal jitter even in clini-
• Technical errors that produce a falsely cally normal muscles. In their early experience
reduced jitter include recording from a with 70 patients with myasthenia gravis, jit-
split fiber, a damaged fiber, or direct mus- ter was always abnormal. They concluded that
cle stimulation (with stimulated SFEMG if jitter was normal in the presence of weak-
only). ness, the weakness was not caused by myas-
thenia gravis. They also noted that jitter was
increased with motor unit potential firing rate
and muscle activity and decreased with rest or
CLINICAL APPLICATIONS OF edrophonium. Frequent blocking in patients
SFEMG with severe myasthenia gravis made SFEMG
recordings difficult to perform.
The work of Stålberg et al.9 was confirmed
Primary Disorders of and amplified by other investigators.6–8, 13, 14
Neuromuscular Transmission Even though the patient populations studied
by these authors differed, all of them found
SFEMG is so sensitive that abnormalities of that 77%–95% of patients with myasthenia
neuromuscular transmission are recognized in gravis have abnormal SFEMG findings, with
the absence of clinical weakness or abnor- increased jitter or blocking (or both). Patients
malities on other physiologic tests.6–8 The with generalized myasthenia gravis have a
clinical usefulness of SFEMG in identify- higher proportion (98%–100%) of abnormal
ing and quantitating defects of neuromuscular jitter, and if there is weakness caused by myas-
transmission has been demonstrated repeat- thenia gravis in the muscle being tested, all
edly.7–11 Although the method is more com- authors agreed that the jitter in that muscle is
plicated and time-consuming than repetitive abnormal.
The frequency of SFEMG abnormalities stimulation is a reliable, readily performed

in patients with ocular myasthenia gravis was method of obtaining a quantitative measure
less consistent among the authors mentioned of a defect of neuromuscular transmission.
above, ranging from 20% to 70% abnormal The presence of a reproducible decrement
in the extensor digitorum communis mus- with slow rates of stimulation, which is par-
cle and from 60% to 100% in proximal tially repaired after exercise and is enhanced
and facial muscles.7, 8, 15, 16 The presence of late after exercise, provides enough evidence
increased jitter in the extensor digitorum com- of abnormality that SFEMG is not required
munis muscle in patients with clinical disease unless further quantitation of the defect is
restricted to ocular muscles increases the risk needed. Moreover, most patients who have
of the development of clinically generalized a decrement with repetitive stimulation have
myasthenia.17 marked SFEMG abnormalities that make mea-
Slight increases in fiber density have been surement of the jitter and blocking more diffi-
reported for up to 25% of patients with myas- cult and less reliable.
thenia gravis; however, this usually is not found, In addition to repetitive stimulation, each
and if it is present, it is minimal and of little patient undergoes standard needle EMG of
clinical significance. SFEMG is often abnormal the proximal and distal muscles, including
when myasthenia gravis is in clinical remission clinically weak muscles, paraspinal muscles,
(66%–83%) and in patients with thymoma and and bulbar muscles. The needle examination
no clinical weakness. can provide evidence of other diseases that
SFEMG findings have been compared with may be associated with abnormal SFEMG,
those of other diagnostic studies. All the com- such as myopathies and anterior horn cell dis-
parisons have shown a much greater frequency ease. These diseases must be identified before
of abnormality on SFEMG than on repetitive SFEMG is performed, to save the time and
stimulation.18, 19 All the studies found a slightly effort of performing SFEMG and to ensure
higher percentage of abnormality for SFEMG that they are not mistaken for myasthenia
than for acetylcholine receptor antibody levels. gravis. Also, standard needle EMG can sample
However, the studies of both Stålberg et al.10 rapidly several muscles on which it may be dif-
and Kelly et al.14 demonstrated that SFEMG ficult to perform SFEMG. In these muscles,
and acetylcholine receptor antibodies are com- single motor unit potentials are studied for
plementary tests: one test identifies abnormal- the variation in size or shape that can provide
ity in some patients in whom the results of the clear evidence of a defect of neuromuscular
other test are normal. transmission. The presence of abnormal motor
SFEMG has been used in the Mayo Clinic unit potentials or motor unit potential varia-
EMG Laboratory since 1976 as a useful tion on standard needle EMG makes SFEMG
adjunct to standard EMG and nerve conduc- unnecessary unless further quantitation is
tion studies in patients thought to have a needed.
defect of neuromuscular transmission. Patients Therefore, in our practice, SFEMG is lim-
with other disorders have been evaluated with ited to patients with normal findings on stan-
SFEMG only as part of special studies and dard studies. SFEMG studies begin on the
not routinely. A large number of patients are extensor digitorum communis muscle because
evaluated, primarily for diagnostic purposes, it is easy to examine and is well defined,
in the Mayo Clinic laboratory, and the effi- and age-controlled normal values are available.
cient evaluation of individual patients becomes Twenty or more fiber pairs are measured for
a major determinant in the selection of studies. MCD, blocking, and mean duration of the total
SFEMG is applied regularly only to patients interpotential interval. This muscle is tested
in whom defects of neuromuscular transmis- even with no symptoms or signs of weakness
sion are suspected. Furthermore, it is limited in the extremities because a high proportion of
to patients who have no evidence of disease on them show abnormalities. If the extensor dig-
other electrophysiologic testing. itorum communis muscle is normal, SFEMG
Each patient in whom myasthenia gravis or is performed on other muscles selected on the
a similar disorder is suspected first undergoes basis of clinical weakness. The frontalis muscle
repetitive stimulation of distal and proximal is commonly used for patients with ocular or
muscles before and after exercise. Repetitive bulbar symptoms.20
Key Points the effect of firing rate on jitter and blocking,

• In myasthenia gravis, both jitter and SFEMG has not been shown to add substantial
information to repetitive stimulation and stan-
blocking are increased in proportion to
dard concentric needle EMG in patients with
the severity of clinical involvement, with either Lambert–Eaton myasthenic syndrome
greater abnormality in weaker muscles.
• There is marked variation in the amount or botulism.
of jitter in different fiber pairs in patients
with myasthenia gravis. Key Points
• Abnormal jitter may be present in clini- • Abnormalities on standard EMG and
cally normal muscles. SFEMG are seen in disorders such as
• The frequency of SFEMG abnormalities the Lambert–Eaton myasthenic syndrome
in patients with ocular myasthenia gravis and botulism.
ranges from 20% to 70% abnormal in the • Increased jitter and blocking occur at
extensor digitorum communis muscle and higher discharge rates in Lambert–Eaton
from 60% to 100% in proximal and facial myasthenic syndrome and botulism.
muscles.
CONGENITAL MYASTHENIC Primary Disorders of Muscle

SYNDROMES
Most muscle disorders that cause abnormal-
Use of voluntary SFEMG to study patients ities on SFEMG do so by causing mus-
with congenital myasthenic syndromes is often cle necrosis, with secondary regeneration and
limited by the difficulty in obtaining cooper- reinnervation. However, muscle fiber mem-
ation in young patients. Stimulated SFEMG brane abnormalities that cause electrophys-
provides a good alternative in these circum- iologic failure can also lead to changes.
stances.21 There are few reports of the use of Decreased fiber density and blocking without
stimulated SFEMG in congenital myasthenic much increase in jitter were observed during a
syndromes. We have used this technique in paralytic attack in a patient with hypokalemic
selected patients but have yet to demonstrate periodic paralysis.28
that it provides additional information to repet- Muscular dystrophies and inflammatory
itive stimulation and standard concentric nee- myopathies are the muscle diseases that
dle EMG. Determination of the mechanism show the most marked abnormalities on
of the neuromuscular transmission defect in SFEMG, but abnormalities have been noted
congenital myasthenic syndromes requires a in metabolic and congenital myopathies as
combination of microelectrode and morpho- well.29, 30 Foote et al.31 found markedly increa-
logical studies performed on an intercostal or sed jitter, blocking, duration of the single fiber
anconeus muscle biopsy specimen.22 complex, and fiber density in 18 patients with
inflammatory myopathy. A significant nega-
tive correlation existed between the motor
LAMBERT–EATON MYASTHENIC
unit potential duration (measured with con-
SYNDROME AND BOTULISM
centric needle electrode) and the degree of
Presynaptic disorders of neuromuscular trans- jitter and blocking. This was believed to pro-
mission have also been studied with single fiber vide evidence that the myopathic process was
techniques. Abnormalities, both with standard involved in the genesis of both electrophysio-
EMG and SFEMG, are readily seen in disor- logic abnormalities. Duchenne muscular dys-
ders such as the Lambert–Eaton myasthenic trophy causes increased fiber density and a
syndrome and botulism.23, 24 Increased jitter markedly increased duration as well as mod-
and blocking that decrease at higher inner- erately increased jitter and blocking. Limb gir-
vation rates are characteristic of presynaptic dle muscular dystrophy demonstrates quantita-
disorders.25, 26 tively less marked changes.
Postsynaptic disorders typically demonstrate SFEMG is not an effective method for pri-
increased jitter and blocking at higher dis- mary diagnosis of myopathies. However, it adds
charge rates.27 With the possible exception of complementary information to the concentric
needle examination about the degree and tim- alcohol abuse. Findings of increased fiber den-
ing of associated reinnervation and may detect sity, jitter, and blocking were seen in alcoholic
early mild abnormalities not recognized on patients who had only mildly slower nerve con-
standard EMG. The usefulness of SFEMG duction velocities but evidence of denervation
for serial study of myopathies during treat- on concentric needle examination. Patients
ment trials and for investigation of membrane with diabetic or uremic polyneuropathies had
abnormalities has not been determined. slower nerve conduction velocities, relatively
normal concentric needle examinations, and
mild abnormalities on SFEMG; that is, mildly
Key Points
increased jitter without blocking and normal
• Muscle disorders that cause abnormali- fiber density and duration. SFEMG corrob-
ties on SFEMG do so by causing muscle orated standard EMG and pathologic data
necrosis, with secondary regeneration and about the neuropathies of these patients. Simi-
reinnervation. lar findings have been demonstrated by other
• Muscular dystrophies and inflammatory investigators in critical illness neuropathy33
myopathies are the muscle diseases that and length-dependent diabetic polyneuropa-
show the most marked abnormalities on thy.34 SFEMG may also help detect conduction
SFEMG. block in focal neuropathy.35
• SFEMG is not an effective method for
primary diagnosis of myopathies. Key Points
• Increases in fiber density are seen as early
as 3–4 weeks after peripheral nerve injury.
Primary Neurogenic Disorders • Increased jitter and blocking are seen for
3–6 months after the injury but rarely
Because reinnervating nerve terminals have longer.
immature neuromuscular junctions, SFEMG
demonstrates abnormalities. The specific type
and degree of abnormality depend on the mag- ANTERIOR HORN CELL
nitude and rate of progression of the neu- DISORDERS
rogenic process. Abnormalities on standard Stålberg et al.36 reported the SFEMG find-
nerve conduction studies and EMG differenti- ings in 21 patients with anterior horn cell dis-
ate these from disorders of the neuromuscular ease and in 3 with syringomyelia. All patients
junction. had increased fiber density. The increase was
greatest in anterior horn cell disorders that
PERIPHERAL NEUROPATHY were slowly progressive (fiber density, 5.4).
The increase (fiber density, 3.3) was less in
In studies of severed nerves, increased fiber rapidly progressive amyotrophic lateral sclero-
density is the first sign of reinnervation. sis. Increased jitter and blocking were observed
Increases in fiber density are seen as early as in all these conditions: the largest increase
3–4 weeks after nerve injury, usually before was in amyotrophic lateral sclerosis, and the
changes can be detected on muscle biopsy. increase was less in the spinal muscular atro-
Fiber density increases rapidly for the first phies and syringomyelia. In the chronic con-
3 months after injury and slowly thereafter. ditions, the complexes (particularly the initial
Increased jitter and blocking are seen for part) were more stable. Duration of single
3–6 months after the injury but rarely longer fiber potentials varied considerably. However,
than that. Most clinical neuropathic disease the longest durations were seen in the more
presents with more complex findings because chronic conditions. The authors concluded that
the process is progressive rather than a single the dual findings of moderately increased fiber
insult and the disease may affect the ability to density and the unstable complexes of varying
reinnervate. configuration represent a rapidly progressive
Thiele and Stålberg32 reported SFEMG process with active reinnervation, such as amy-
findings in 54 patients with polyneuropathy otrophic lateral sclerosis. Markedly increased
associated with uremia, diabetes mellitus, or fiber density and relatively stable complexes
(particularly of the initial part) indicate a slowly initial part) indicate a slowly progressive
progressive disease or burned-out process with neurogenic disease or burned-out process
long-standing reinnervation. The combination with long-standing reinnervation.
of markedly increased fiber density and unsta- • In neurogenic disorders of all types, the
ble potentials was believed to reflect reactiva- abnormalities on SFEMG complement
tion of a long-standing process.37 those seen during conventional needle
Schwartz et al.38 reported similar con- EMG.
clusions in 10 patients with long-standing
syringomyelia. SFEMG abnormalities (and
clinical changes) were maximum in muscles SUMMARY
innervated by spinal segments C8 and T1.
In the first dorsal interosseous muscle, mean SFEMG is a highly selective technique that
fiber density was 4.1, with 21% of potential permits recording of individual components
pairs demonstrating increased jitter and 7% of the motor unit. The selectivity of SFEMG
demonstrating blocking. The distribution of depends on the use of a low-frequency filter
abnormalities, rather than the type, differen- of 500 Hz and a small electrode size. Volun-
tiated these patients from those with anterior tary activation or electric stimulation is used to
horn cell disease. Patients with chronic non- activate the muscle fiber in SFEMG. Voluntary
progressive clinical conditions demonstrated activation allows measurement of fiber density,
complex, but stable, motor unit action poten- whereas jitter and blocking are recorded with
tials and increased fiber density. Patients with both voluntary and stimulated SFEMG.
recent clinical progression demonstrated more SFEMG is technically demanding and
blocking. requires specialized recording equipment. A
Daube and Mulder39 reported mildly increa- variety of modern digital equipment is avail-
sed fiber density and more markedly increased able that assists in the collection, display, anal-
jitter and blocking in 31 unselected patients ysis, reporting, and archiving of SFEMG data.
with amyotrophic lateral sclerosis. The patient’s SFEMG is the most sensitive clinical elec-
age, clinical severity, CMAP amplitude, and trophysiologic technique for the detection of
presence of a decrement to slow repeti- a defect of neuromuscular transmission. The
tive stimulation were valuable in predicting findings are not specific or diagnostic for
longevity. However, SFEMG findings did not individual diseases. SFEMG findings are also
add to the prognostic accuracy. Single fiber abnormal in disorders associated with denerva-
study of spontaneously recorded fasciculations tion and reinnervation of muscle, such as cer-
in patients with amyotrophic lateral sclerosis tain myopathies, motor neuron diseases, and
has documented increased jitter and blocking peripheral neuropathies. Correlation of fiber
in those discharges. density and jitter and blocking analysis may
In neurogenic disorders of all types, the help to determine disease chronicity and rate
abnormalities on SFEMG complement those of progression.
seen during conventional needle EMG. The
single fiber profile can delineate the rate of
progression and longevity of the disease pro-
cess. Whether SFEMG will become clinically REFERENCES
useful in establishing prognosis in certain dis-
eases, serially following diseases during clini- 1. Kouyoumdjian, J. A., E. V. Stalberg. 2008. Refer-
ence jitter values for concentric needle electrodes
cal treatment trials, or evaluating for evidence in voluntarily activated extensor digitorum commu-
of reinnervation in neurogenic diseases is not nis and orbicularis oculi muscles. Muscle & Nerve
known. 37(6):694–9.
2. Kouyoumdjian, J. A., E. V. Stalberg. 2008. Concentric
Key Points needle single fiber electromyography: Comparative
jitter on voluntary-activated and stimulated Exten-
• Increased jitter and blocking and fiber sor Digitorum Communis. Clinical Neurophysiology
density are observed in amyotrophic lat- 119(7):1614–8.
3. Bromberg, M. B., and D. M. Scott. 1994. Single
eral sclerosis. fiber EMG reference values: Reformatted in tabular
• Markedly increased fiber density and rela- form. AD HOC Committee of the AAEM Single Fiber
tively stable complexes (particularly of the Special Interest Group. Muscle & Nerve 17:820–1.
4. Stalberg, E. V., and M. Sonoo. 1994. Assessment of with myasthenia gravis. Electromyography and Clin-
variability in the shape of the motor unit action poten- ical Neurophysiology 47:93–6.
tial, the “jiggle,” at consecutive discharges. Muscle & 20. Ukachoke, C., P. Ashby, A. Basinski, and J. A. Sharpe.
Nerve 17(10):1135–44. 1994. Usefulness of single fiber EMG for distinguish-
5. Campos, C., A. Malanda, L. Gila, V. Segura, ing neuromuscular from other causes of ocular mus-
I. Lasanta, and J. Artieda. 2000. Quantification of jig- cle weakness. The Canadian Journal of Neurological
gle in real electromyographic signals. Muscle & Nerve Sciences 21:125–8.
23(7):1022–34. 21. Tidswell, T., and M. C. Pitt. 2007. A new analytical
6. Sanders, D. B., J. F. Howard Jr., and T. R. Johns. 1979. method to diagnose congenital myasthenia with stim-
Single-fiber electromyography in myasthenia gravis. ulated single-fiber electromyography. Muscle & Nerve
Neurology 29:68–76. 35:107–10.
7. Cui, L. Y., Y. Z. Guan, H. Wang, and X. F. Tang. 22. Engel, A. G., K. Ohno, M. Milone, and S. M.
2004. Single fiber electromyography in the diagnosis of Sine. 1998. Congenital myasthenic syndromes. New
ocular myasthenia gravis: Report of 90 cases. Chinese insights from molecular genetic and patch-clamp stud-
Medical Journal 117:848–51. ies. Annals of New York Academy Sciences 841:
8. Valls-Canals, J., M. Povedano, J. Montero, and 140–56.
J. Pradas. 2003. Stimulated single-fiber EMG of the 23. Chaudhry, V., and T. O. Crawford. 1999. Stimulation
frontalis and orbicularis oculi muscles in ocular myas- single-fiber EMG in infant botulism. Muscle & Nerve
thenia gravis. Muscle & Nerve 28:501–3. 22:1698–703.
9. Stålberg, E., J. Ekstedt, and A. Broman. 1971. The 24. Padua, L., I. Aprile, M. L. Monaco, et al. 1999. Neuro-
electromyographic jitter in normal human muscles. physiological assessment in the diagnosis of botulism:
Electroencephalography and Clinical Neurophysiol- Usefulness of single-fiber EMG. Muscle & Nerve
ogy 31:429–38. 22:1388–92.
10. Stålberg, E., J. V. Trontelj, and M. S. Schwartz. 25. Trontelj, J. V., E. Stålberg, M. Mihelin, and A. Khurai-
1976. Single-muscle-fiber recording of the jit- bet. 1992. Jitter of the stimulated motor axon. Muscle
ter phenomenon in patients with myasthenia & Nerve 15:449–54.
gravis and in members of their families. Annals 26. Mandler, R. N., and R. A. Maselli. 1996. Stimulated
of the New York Academy of Sciences 274: single-fiber electromyography in wound botulism.
189–202. Muscle & Nerve 19:1171–3.
11. Murga, L., F. Sanchez, C. Menendez, and J. M. 27. Lin, T. S., and H. C. Chiu. 1998. Motor end-plate jitter
Castilla. 1998. Diagnostic yield of stimulation and in myasthenia gravis at different firing rates. Journal of
voluntary single-fiber electromyography in myasthenia Clinical Neurophysiology 15:262–7.
gravis. Muscle & Nerve 21:1081–3. 28. De Grandis, D., A. Fiaschi, G. Tomelleri, and
12. Stålberg, E., and J. Trontelj, eds. 1994. Single fiber D. Orrico. 1978. Hypokalemic periodic paralysis.
electromyography: Studies in healthy and diseased A single fiber electromyographic study. Journal of
muscle, 2nd ed. New York: Raven Press. Neurological Sciences 37:107–12.
13. Konishi, T., H. Nishitani, F. Matsubara, and M. Ohta. 29. Ukachoke, C., P. Ashby, A. Basinski, and J. A. Sharpe.
1981. Myasthenia gravis: Relation between jitter in 1994. Usefulness of single fiber EMG for distinguish-
single-fiber EMG and antibody to acetylcholine recep- ing neuromuscular from other causes of ocular mus-
tor. Neurology 31:386–92. cle weakness. The Canadian Journal of Neurological
14. Kelly, J. J. Jr., J. R. Daube, V. A. Lennon, F. M. Howard Sciences 21:125–8.
Jr., and B. R. Younge. 1982. The laboratory diagno- 30. Bertorini, T. E., E. Stålberg, C. P. Yuson, and W. K.
sis of mild myasthenia gravis. Annals of Neurology Engel. 1994. Single-fiber electromyography in neu-
12:238–42. romuscular disorders: Correlation of muscle histo-
15. Padua, L., E. Stålberg, M. LoMonaco, A. Evoli, chemistry, single-fiber electromyography, and clinical
A. Batocchi, and P. Tonali. 2000. SFEMG in ocular findings. Muscle & Nerve 17:345–53.
myasthenia gravis diagnosis. Clinical Neurophysiology 31. Foote, R. A., W. M. O’Fallon, and J. R. Daube. 1978.
111:1203–7. A comparison of single fiber and routine EMG in nor-
16. Valls-Canals, J., J. Montero, and J. Pradas. 2000. Stim- mal subjects and patients with inflammatory myopa-
ulated single fiber EMG of the frontalis muscle in thy. Bulletin of the Los Angeles Neurological Societies
the diagnosis of ocular myasthenia. Muscle & Nerve 43:95–103.
23:779–83. 32. Thiele, B., and E. Stålberg. 1975. Single fibre EMG
17. Weinberg, D. H., J. F. Rizzo III, M. T. Hayes, M. D. findings in polyneuropathies of different aetiology.
Kneeland, and J. J. Kelly Jr. 1999. Ocular myasthenia Journal of Neurology, Neurosurgery, and Psychiatry
gravis: Predictive value of single-fiber electromyogra- 38:881–7.
phy. Muscle & Nerve 22:1222–7. 33. Schwarz, J., J. Planck, J. Briegel, and A. Straube.
18. Sonoo, M., H. Uesugi, A. Mochizuki, Y. Hatanaka, 1997. Single-fiber electromyography, nerve conduc-
and T. Shimizu. 2001. Single fiber EMG and repet- tion studies, and conventional electromyography in
itive nerve stimulation of the same extensor digito- patients with critical-illness polyneuropathy: Evidence
rum communis muscle in myasthenia gravis. Clinical for a lesion of terminal motor axons. Muscle & Nerve
Neurophysiology 112:300–3. 20:696–701.
19. Srivastava, A., J. Kalita, and U. K. Misra. 2007. A com- 34. Bril, V., M. R. Werb, D. A. Greene, and A. A.
parative study of single fiber electromyography and Sima. 1996. Single-fiber electromyography in diabetic
repetitive nerve stimulation in consecutive patients peripheral polyneuropathy. Muscle & Nerve 19:2–9.
35. Papua, L., I. Aprile, P. D’Amico, et al. 2001. A use- unstable postpolio subjects, stable postpolio subjects,
ful electrophysiological test for diagnosis of mini- and control subjects. Archives of Physical Medicine
mal conduction block. Clinical Neurophysiology 112: and Rehabilitation 78:986–91.
1041–8. 38. Schwartz, M. S., E. Stålberg, and M. Swash. 1980. Pat-
36. Stålberg, E., M. S. Schwartz, and J. V. Trontelj. 1975. tern of segmental motor involvement in syringomyelia:
Single fibre electromyography in various processes A single fibre EMG study. Journal of Neurology,
affecting the anterior horn cell. Journal of Neurolog- Neurosurgery, and Psychiatry 43:150–5.
ical Sciences 24:403–15. 39. Daube, J. R., and D. W. Mulder. 1982. Prognostic fac-
37. Rodriquez, A. A., J. C. Agre, and T. M. Franke. 1997. tors in amyotrophic lateral sclerosis (abstract). Muscle
Electromyographic and neuromuscular variables in & Nerve 5(Suppl):S107.
Chapter 29
Quantitative Motor Unit Number

Estimates
Jasper R. Daube
INTRODUCTION Surface Averaging of Needle EMG MUPs

MUNE BY STANDARD EMG MUNE from All-or-None
Interference Pattern Increments in the CMAP
Recruitment Analysis Statistical (STAT) MUNE
MUNE BY STANDARD MOTOR NCS CLINICAL APPLICATIONS
QUANTITATIVE MUNE SUMMARY
Underlying Assumptions of
Quantitative MUNE
INTRODUCTION • The hallmark of neurogenic diseases is

damage or destruction to motor neurons
Peripheral neuromuscular diseases are divided and/or peripheral axons.
broadly into neurogenic and myopathic dis- • None of the standard assessments are
eases, with diseases of the neuromuscular able to quantitatively and reproducibly
junction often being included with myopathic directly measure the loss of motor units or
diseases. Neurogenic and myopathic diseases axons.
both result in the clinical problem of muscle • Motor unit number estimates (MUNEs)
weakness. The task of clinicians is, first, to dis- provide quantitative measures of the num-
tinguish between these two broad groups of ber of axons in peripheral neuromuscular
diseases and, second, to identify the specific diseases.
type. • MUNE can thereby provide direct confir-
mation of the severity of disease and its
rate of progression.
Purpose and Role of Motor Unit Number
Estimates
The critical difference between neurogenic
• The major advantage of electrophysiologic and myopathic diseases is that motor neurons
testing of neuromuscular disease is the or their axons are the primary site of damage
ability to measure changes caused by dis- in neurogenic diseases whereas muscle is the
ease quantitatively. primary site of damage in myopathic diseases.
493
Damage to either the anterior horn cell or its Each of these MUNE methods is in active use
peripheral axon is the essence of a neurogenic in different EMG laboratories.
process, even though the associated change in The important terms used in this chapter are
muscle may be more apparent clinically. Neu- defined as follows:
rogenic diseases may produce either histologic
changes in nerve or muscle, or physiologic 1. Motor unit—a single anterior horn cell or
changes without histopathologic correlates. An brain stem motor neuron, its peripheral
example of a physiologic abnormality is fas- axon (which travels in a cranial or periph-
ciculation arising from axonal irritability, with eral nerve), and each of the muscle fibers
spontaneous firing in the motor nerve. Phys- the axon innervates.
iologic disorders produce no loss of function 2. Motor unit potential (MUP)—the electri-
(as in slowing of conduction) or loss of func- cal potential recorded from a motor unit
tion (as in conduction block). Degenerative or by needle EMG electrode in the muscle.
destructive processes result in the loss of motor 3. Surface-recorded motor unit potential
neurons or peripheral axons. Conduction block (SMUP)—the electrical potential recor-
or loss of either motor axons or neurons is the ded from a motor unit by a skin surface
basis of the weakness found in most patients electrode over the muscle.
with a neurogenic disease. The severity of the 4. Number of motor units—the number of
clinical deficit is related directly to the number functioning motor neurons or function-
of motor neurons or axons (or both) that are ing motor axons innervating a muscle or
blocked or lost. Therefore, an important part group of muscles.
of the assessment of neuromuscular disease is 5. Motor unit number estimate (MUNE)—a
to determine the number of functioning motor physiologic determination of the number
units.1, 2 It would be ideal to have an actual of motor units in a muscle or group of
measure of the number of motor units, but cur- muscles.
rent methods—clinical, physiologic, histologic,
and histopathologic—are not able to provide Physiologic estimates of the number of motor
such a measure. Electrophysiologic methods units have been hampered by the lack of a stan-
are required to estimate the number of motor dard determination of the actual number of
units in a muscle.3, 4 motor units in a muscle. At best, histopatho-
This chapter describes the older, standard logic and anatomical determinations are rough
methods for estimating the number of motor measures. Attempts have been made to mea-
units and their loss in neurogenic disease. Stan- sure the number of motor units in human fetal
dard needle electromyography (EMG) and and newborn tissue.5 In both types of studies,
nerve conduction studies (NCS) provide non-
the individual muscles and the motor nerves
quantitative information about the loss of
innervating them were dissected and counted.
motor units and motor axons. These are dis-
The number of large myelinated axons in the
cussed in detail in other chapters, but will be
motor nerve was counted and divided by two
reviewed briefly in this chapter to put them in
to estimate the proportion of these fibers that
perspective with the newer quantitative meth-
were motor rather than sensory. This propor-
ods for determining the number of motor units
tion was based on animal studies in which
in a muscle. One of the new quantitative meth-
degeneration of the peripheral sensory axons
ods uses needle electrode recordings in the
after section of the dorsal root indicated that
muscle, while the other three new methods
approximately half of the axons in a motor
rely on stimulation of the peripheral nerve.
nerve are sensory.6 This proportion of large
These are as follows:
myelinated axons then served as the estimate
for the number of motor units innervating the
Surface averaging of needle EMG motor muscle. These studies also counted the num-
unit potentials ber of muscle fibers in a muscle to determine
All-or-none increments in the compound the innervation ratio of muscle fibers to axons.
muscle action potential Although the results of the two studies were
F-wave measurements similar, the values were sufficiently different
Statistical MUNE to preclude the designation of a true standard
Quantitative Motor Unit Number Estimates 495
measurement of the number of motor units whether from subjective or automated meth-
innervating individual human muscles. How- ods, have been used to make judgments about
ever, these studies do serve as baseline com- the loss of motor units. When the loss is mod-
parisons for the physiologic methods that have erate to severe, these methods can identify a
been developed for MUNE. The absence of a loss of motor units, for example, as in a reduced
standard makes direct comparison of the val- interference pattern or a single motor unit
ues obtained by different methods an equally firing pattern.
important part of the assessment of the validity
of individual methods of MUNE. Key Points
• MUNE is reflected in the number of
Key Points motor units activated during voluntary
• Weakness in neurogenic diseases is due to contraction.
loss of axons or loss of motor neurons. • Assessment of voluntary EMG activity can
• MUNE defines the extent of loss of axons provide a rough judgment of the number
or motor neurons. of motor units lost, but not a reproducible,
• Standard EMG and NCS provide non- quantitative value.
quantitative information about the num-
ber of axons innervating a muscle.
• There are four categories of other elec- Interference Pattern
trophysiologic methods that estimate an
actual number of motor units in a muscle The methods of interference pattern analy-
or group of muscles. sis can be classified into two approaches (see
Chapter 27). The more common one is the
measurement of the number of turns (i.e.,
MUNE BY STANDARD EMG reversals of potential per unit time) and the
amplitude of the spikes in the EMG pattern.
Standard diagnostic clinical EMG has always The other method analyzes the frequency com-
included a subjective estimate of the num- ponents of the EMG pattern. Both methods
ber of motor units in a muscle (see Chap- provide some measure of density and, indi-
ter 26). Electromyographers have used recruit- rectly, of the number of motor units. Nei-
ment analysis or interference pattern analy- ther method is reliable enough to provide a
sis (or both) with voluntarily activated EMG quantitative measure of the number of motor
recordings to judge the number of motor units. Alterations in MUPs with disease fur-
units in a muscle.7, 8 The EMG recorded with ther reduce the reliability of both methods.
either surface or intramuscular electrodes dur- For example, in some disease processes, the
ing strong voluntary contractions summates the potential generated by individual motor units
activity of the muscle fibers in the activated becomes more complex with multiple phases.
motor units to produce an interference pat- These additional phases contribute to both the
tern. The greater the number of motor units high-frequency components and the number
activated, the greater the density of the EMG of turns, but they do not reflect a change in
pattern. The increased density with increased the number of motor units. Thus, interference
effort is the result of an increase in the num- pattern analysis is unsatisfactory for MUNE.
ber of motor units activated and an increase in
the firing rate of individual motor units. The
combination of the firing rate and the number Recruitment Analysis
of motor units reduces the reliability of den-
sity measures in determining the number of The second method used by clinical elec-
motor units. Density measures of the number tromyographers to judge the number of motor
of motor units are further beset by the problem units in a muscle is recruitment analysis (see
of having to rely on the effort of the patient Chapter 27). Recruitment refers to the initia-
for obtaining activation. Clearly, the effort of tion of firing of additional motor units as the
the patient alters the number of motor units effort of voluntary contraction increases. The
activated. Nonetheless, measures of density, intrinsic anatomical and physiologic properties
of motor neurons result in a fixed pattern of • The ratio of the rate of firing of any single
activation in response to voluntary effort. Dur- motor unit to the total number of active
ing voluntary activation, low-threshold motor units provides a reproducible measure of
neurons begin firing at rates of 5–7 Hz and the axons lost in a neurogenic process.
increase their firing frequency with increasing • Subjective assessment of recruitment
effort. As the effort increases, additional motor requires years of EMG experience, which
units are activated, and they, in turn, increase is not quantitative, and thus remains diffi-
their frequency of firing with further increases cult to compare between examiners.
in effort. The recruitment of motor units dur-
ing activation is a fixed relationship between
the number of activated motor units and their MUNE BY STANDARD
firing rates. In recruitment analysis, the num- MOTOR NCS
ber of MUPs activated at any given level of
effort is compared with the rate of firing of Motor NCS are an important part of the
individual motor units. This ratio provides an electrophysiologic analysis of peripheral neuro-
indirect measure of the number of motor units muscular disease (see Chapter 23). The ampli-
in the muscle. tudes of compound muscle action potentials
Clinical EMG judgments about the number (CMAPs) are related directly to the number
of motor units compare the rate of firing of sin- and size of muscle fibers in a muscle group and
gle units with the total number of motor units. indirectly to the number of motor units in the
However, determining the rate of firing is one muscle group.9 If a disease is known to be neu-
of the more difficult steps in standard EMG. rogenic and acute, the amplitude of a CMAP is
The ratio of the number of units firing to the a rough estimate of the number of motor units.
rate of firing can provide a rough gauge of the Conduction block provides a quantitative mea-
number of motor units. This semiquantitative sure of the proportion of motor units that are
method of determining reduced recruitment not functioning distal to the block, but not the
provides a more accurate and reproducible actual number.
estimate of the number of motor units than The value of CMAP amplitude is further
does interference pattern analysis. Although limited by three other factors. First, the ampli-
recruitment analysis is reasonably reproducible tude is decreased in myopathies with loss of
and clinically reliable, it is usually a subjec- muscle fiber tissue. Second, estimates of the
tive judgment made by electromyographers on number of motor units made on the basis of
the basis of experience. It requires taking into the amplitude of the CMAP are hampered
account differences in recruitment in differ- further by the wide range of normal ampli-
ent areas of individual muscles and the even tudes. For example, even in the case of acute
greater differences among different muscles. traumatic section of a peripheral nerve that
Automated methods for formally quantifying disrupts half of the motor axons, the ampli-
the recruitment pattern have been developed tude of the ulnar or hypothenar CMAP may
(see Chapter 27). These formal quantitative decrease from 12 to 6 μV and still be within
measures can provide evidence of the relia- normal limits. Third, the CMAP reduction that
bility of the clinical methods; however, they occurs with the destruction of axons can be
are time-consuming and complex and have not compensated for partially or fully by reinnerva-
found clinical application. None of these meth- tion from collateral sprouting of intact axons. In
ods provides an actual estimate of the number the clinical setting in which the baseline ampli-
of motor units in a muscle. tude is not known, the CMAP is only a rough
guide to the number of motor units. Therefore,
the amplitude of the CMAP cannot be used to
Key Points obtain a reliable MUNE.
• Interference pattern analysis of voluntary
Key Points
EMG activity cannot provide quantitative
or even partially quantitative estimates of • CMAP amplitude measures can quanti-
the number of axons lost in a neurogenic tate the proportion of axons lost in disease
process. in a disorder with conduction block, but
cannot define the specific number of axons Underlying Assumptions of

remaining. Quantitative MUNE
• CMAP amplitude measurement is an
unreliable measure of the remaining axons All MUNE techniques rely on several under-
because of lying assumptions that must be understood to
◦ Reduction in CMAP amplitude by appreciate fully the advantages and drawbacks
myopathies where there is no loss of of quantitative MUNE:
axons.
◦ Reinnervation of denervated muscle
fibers by collateral sprouting from intact 1. MUNEs are assumed to be measure-
axons. ments from a single muscle, but maxi-
◦ The wide range of normal amplitudes mal stimulation of any peripheral motor
for all CMAPs. nerve activates all the muscles inner-
vated by that nerve distal to the point of
stimulation. Therefore, measurements of
the CMAP are the summation of activity
QUANTITATIVE MUNE from multiple muscles. For example, the
median or thenar CMAP is the summa-
Needle EMG and nerve stimulation methods tion of the activity of the opponens polli-
have been developed in the past twenty years cis, abductor pollicis brevis, flexor pollicis
to provide quantitative MUNE.10–15 Both the brevis, and, to a lesser extent, the lum-
needle EMG and the motor nerve stimula- brical muscles. The ulnar or hypothenar
tion approaches to MUNE make basic assump- CMAP is the summation of all the other
tions about the electric characteristics of MUPs intrinsic muscles of the hand. Thus,
that are important to understand to make MUNE is more accurately an estimate
use of quantitative MUNE. Six major meth- of the number of motor units in a group
ods of quantitative MUNE have been devel- of muscles rather than in a single mus-
oped. Four of them record CMAPs in response cle. Moreover, the SMUP in more distant
to nerve stimulation, and two rely on nee- muscles contribute less to the CMAP than
dle EMG. Variations of each of these four do SMUP in nearby muscles.
basic techniques continue to evolve to improve 2. Each of the MUNE methods assumes
on the accuracy and reliability of MUNE. that each SMUP has the same size each
Each method makes two basic measurements: time it is activated. Thus, defects of neu-
(1) the average size of the potential gener- romuscular transmission that result in
ated by a single motor unit as recorded on the varying sizes of the SMUP will cause inac-
surface of the skin—surface-recorded motor curacy. This is significant only if there is a
unit potential (SMUP)—and (2) the size of the large variation in SMUP size.
CMAP obtained with maximal stimulation of a 3. Although the assumption that all motor
motor nerve. MUNE is obtained by dividing axons are activated by maximal stimula-
the maximal CMAP by the average size of the tion is generally true, it may not be true
SMUP. The techniques differ in the way that of disease involving high-threshold axons
the average size of the SMUP is obtained. (e.g., severe demyelination and regener-
ated axons). In these situations, the max-
imal CMAP may be difficult to obtain,
Key Points
thereby underestimating the number of
• A number of quantitative MUP meth- intact axons.
ods have been devised—four using nerve 4. Peripheral nerve stimulation methods
stimulation and two using voluntary EMG assume a random activation of axons
are of particular interest. with differing thresholds and conduction
• All methods measure the size of the aver- velocities. Collision stimulation method
age single motor unit potential (SMUP) studies of peripheral axons have shown
and the CMAP. that the first activated axons are not rep-
• MUNE is derived by dividing the CMAP resentative of the entire pool, thereby
by the SMUP. posing limitations on the reliability of
MUNE with threshold peripheral nerve Key Points

stimulation.16 • Each of the methods of quantitative
5. The most critical issue in determining the
MUNE makes the same underlying
size of individual SMUPs is the adequacy
assumptions that can reduce the reliability
of sampling of the many SMUPs in the
of the estimate. The most critical is
muscle.17 In patients with severe neuro- ◦ Reliable direct measurement of the
genic disease, it is possible to identify
number of SMUP can be made if there
each motor unit and to obtain a reliable,
are fewer than 20.
reproducible, direct count of up to 10 or ◦ Increasing number of samples are
20 motor units. This is the practical max-
needed as the total number of SMUP
imum of such direct counts. With more
increases in the muscle group.
motor units, none of the methods allows ◦ Variation in the size of the sampled
reliable measures of each motor unit, so
SMUP increases the number of samples
it is uncertain whether an adequate sam-
needed.
ple has been obtained. If the number • Stimulation must be high enough to acti-
of motor units is greater than 20 or so,
vate all motor axons when recording the
it is necessary to measure the size of a
CMAP.
subset of the total population of motor • Large defects of neuromuscular transmis-
units and then calculate an average size
sion resulting in different sizes of SMUP
for that subgroup. If all the motor units
recorded with a sequence of stimuli to the
are of nearly identical size, sampling a
same single axon are unreliable.
subset of them to obtain an average size • MUNE is made from a group of mus-
of the SMUP is a valid approach. With
cles innervated by a single nerve rather
greater variation in the size of SMUP,
than just the muscle under the recording
particularly if the size range is not a nor-
electrode.
mal distribution, estimates of the true
size become less reliable. Each of the
methods for measuring the size of SMUP
must address the adequacy of the sam- Surface Averaging of Needle
pling based on the variation of mea- EMG MUPs
sured SMUP sizes to determine whether
a reasonable representation of the total Both of the two methods for determining the
population of motor unit sizes has been average SMUP size rely on two channel record-
obtained. ings. One channel records the MUP recorded
with a needle electrode in the muscle; the
other records and averages the surface poten-
It had been assumed that there should be tials generated by the MUP recorded with the
a lower limit of the size of SMUP used to needle electrode. The CMAP size is divided
calculate MUNE.18 However, a more recent by the size of these surface-recorded SMUP
study19 showed that there are significant num- to obtain the MUNE. Two methods have been
bers of small SMUP in normal subjects and described for needle EMG-based identifica-
those with a neurogenic disorder. Calculations tion. Both can be used for MUNE in large
that excluded small SMUP reduced the differ- proximal muscles such as the biceps.
ence in MUNE between normal subjects and
patients with a specific disease. These find- SPIKE-TRIGGERED AVERAGING
ings again illustrate the importance of correct MUNE
sampling procedures.
Each of these assumptions is a possible The first of these methods selects a sin-
source of error in each MUNE method. gle MUP with an amplitude trigger, and is
Nonetheless, the ease of measuring the small referred to as spike-triggered averaging (STA).
number of potentials in neurogenic disorders STA relies on the ability to isolate MUP by
with moderate to marked loss of motor units voluntary activation on needle EMG with a
minimizes these sources of error and makes two-channel EMG machine.20, 21 Intramuscular
MUNE highly reliable in those situations. MUPs are recorded on one channel with any
one of several electrodes, including single fiber peak amplitudes (Fig. 29–1). The triggered,
EMG, bipolar concentric, standard concentric, surface-recorded MUPs are recorded simul-
or fine-wire electrodes, but most often stan- taneously on a second channel triggered by
dard concentric. Individual MUPs are isolated the needle-recorded MUP on the first channel.
on the first channel, by an amplitude trigger The activity is averaged on the second chan-
window that selects potentials on the basis of nel from the same surface electrodes used to
Figure 29–1. Spike-triggered averaging for motor unit number estimate. A, Compound muscle action potential. B and C,
Each line (MU1 –MU9 ) shows the triggered motor unit (MU) potential recorded with a needle (solid lines) and the averaged
surface CMAP (dotted line). (From Brown, W. F., M. J. Strong, and R. Snow. 1988. Methods for estimating numbers of
motor units in biceps-brachialis muscles and losses of motor units with aging. Muscle & Nerve 11:423–32. By permission of
John Wiley & Sons.)
record the CMAP. The technique requires the DECOMPOSITION-BASED

isolation of at least 10, and preferably 20, MUP QUANTITATIVE EMG MUNE
whose spike-triggered average can be recorded
A more recently developed, simplified method
on the surface. The amplitude or area of the
of surface-averaged MUP MUNE selects
surface-recorded potentials is used to calculate
MUP with a sophisticated, automated MUP
the average size of the MUP in the muscle.
identification—decomposition-based quantita-
MUNE is then calculated by dividing the size
tive EMG (DQEMG). A 20-second train of
(area or amplitude) of the maximal CMAP by
needle-recorded MUP is recorded at the same
the average size of the MUP.
time as a surface electrode over the mus-
Several of the assumptions made in using
cle. The program isolates up to eight dis-
this technique are possible sources of error.
tinct MUPs and averages the simultaneous
First, the method assumes that all MUPs can
occurring SMUP. Studies have demonstrated
be recorded at the surface. Studies by Brown
the reliability of the programs and defined
and coworkers22 suggest that this is true only
the differences in distal and proximal mus-
in superficial muscles. Second, the technique
cles, the effect of age, the effect of force
assumes that voluntary activation recruits the
of contraction, and the abnormalities in dis-
full range of sizes of motor units. This likely
ease.27–30 Figure 29–2 shows six averaged
is not true. Larger motor units probably are
not activated with standard voluntary contrac- MUPs recorded from a single train of MUP
tion. Despite these issues, the values obtained in a normal anal sphincter with correspond-
with the method are comparable to those ing recordings of the simultaneous SMUP from
expected on the basis of animal studies and that train.
those obtained with other methods of record-
ing MUNE. Key Points
De Koning et al.23 have modified the tech- • SMUP can be identified from MUP
nique by using macro EMG needles to record recorded from a needle electrode in the
the CMAP. It is assumed that a macro nee- muscle.
dle provides a better representation of the full • Two methods are based on recording
range of motor units, particularly those deeper MUP on needle EMG.
in the muscle. Milner-Brown and Brown24 ◦ STA selects different MUPs by their
used the technique of microstimulation of amplitudes and then averages the
nerve terminals in the end plate region to surface-recorded potential occurring at
activate motor units recorded with a nee- the same time as the MUP.
dle electrode. This reduces the bias in the ◦ DQEMG simultaneously isolates differ-
selection of motor unit sizes that occurs with ent MUPs with a specialized computer
voluntary activation. Each of these methods algorithm. The averaged surface poten-
gives comparable MUNE. Animal studies com- tials occurring with each of these is
paring the actual number of anterior horn averaged to obtain SMUP.
cells with the MUNE using microstimulation
of individual axons in the motor nerve25, 26
showed this method to be more accurate than
standard STA MUNE from All-or-None
The STA methods require specialized record- Increments in the CMAP
ing devices and programs not available on
most commercial EMG machines. They are Axons have the special property of activation
also generally more time-consuming and more at a specific level of depolarization, called the
complex to perform because of the need for threshold. Below that they are inactive, while
two channels of recording: a MUP triggering above that they are fully activated and send
channel and an SMUP averaging channel. The an action potential down the axon. This prop-
accuracy of the methods depends on the ability erty allows individual axons to be stimulated
of the patient and electromyographer to acti- with very low stimulus intensities. An axonal
vate, identify, and trigger individual MUPs for action potential will travel down all branches
a period long enough to allow the size of the of the axon to activate all the muscle fibers that
SMUP to be measured. it innervates. Activation will thus activate an
Figure 29–2. Example of DQEMG-MUNE recordings. DQEMG program identified six MUP from 15 seconds of ongo-
ing EMG in a normal anal sphincter. Column 1, Superimposition of each of the MUPs identified as the same MUP confirms
the quality selection process; column 2, Average of the MUP superimposed; column 3, Average surface electrode record-
ing of the synchronous SMUP recorded with each MUP; column 4, Interdischarge interval between MUPs in that row to
confirm the firing pattern; column 5, Vertical lines to depict the time of occurrence of each MUP selected in that row (gaps
occur when superimposition of MUP prevents their isolation).
SMUP that can be recorded on the surface as the stimulus current is finely controlled in very
an SMUP. small steps designed to allow isolated stimula-
tion of individual axons in a stepwise fashion
(Fig. 29–3). For example, if a muscle contains
MUNE FROM MULTIPLE
only two motor units, the CMAP consists of the
ALL-OR-NONE INCREMENTS AT
summed SMUP of these two potentials only.
ONE STIMULATION POINT
Incremental testing with slowly and gradually
All-or-none increment measurement, intro- increasing current will show no response to a
duced by McComas,31 was the first method stimulus below the threshold of either axon of
used for quantitative MUNE. The method the two motor units. When the threshold of one
is deceptively simple and provides the easi- of the axons is reached, that axon is fully acti-
est and most direct and reliable method of vated and the CMAP suddenly changes from
obtaining MUNE if the number of axons or no response to the response of that SMUP.
motor neurons is reduced.32, 33 It was based When the threshold of the second axon is
on the all-or-none characteristic of the motor reached, this axon also fires and the maximal
axons activation. In the incremental method, CMAP is obtained (Fig. 29–4). Changing the
Figure 29–3. Incremental method of MUNE. A, Ten
response increments in 500 μV give a MUNE of 100. B,
Ten response increments in 2000 μV give a MUNE of 25.
(From Brown, W. F., and T. E. Feasby. 1974. Estimates
of functional motor axon loss in diabetics. Journal of Neu-
rological Sciences 23:275–93. By permission of Elsevier
Science.)
Start: 20.0 mA 20 μV 20–5 kHz 2 ms

% Scan areas: End: 34.5 mA
100
R
e
90 c
o
80 m
m
n
70 d
e
60 d Max area: 8746 uVms. Amp 2.093 mV
T
e
50 s
t 500 uV 20–5 kHz 2 ms
40 i
n
g
30
R
a
20 n
g
e
10 s
Last area: 8561 uVms Amp: 2.031 mV
0 Curve Max area: 8746 uVms Amp: 2.093 mV
0 10 20 30
Figure 29–4. Left, Stimulation of the median nerve from threshold to maximal (20–34.5 mA) in 0.3 mA increments. Note
that two SMUPs are elicited: (1) Top right at 20 μV sensitivity shows 11 superimposed, 30-μV SMUP with baseline noise.
(2) Bottom right at 500 μV sensitivity shows 16 superimposed 2000-μV SMUP.
502
stimulus current above and below these thresh- any axon varies within a small range so that for
olds produces stepwise activation of two steps; any given stimulus, an SMUP has a percent-
that is, the first SMUP and then the full CMAP. age likelihood of firing. Therefore, any one of
If there are three motor units in the muscle, the three axons with nearly identical thresh-
three steps would be recorded, and similarly olds might be activated for each stimulus. An
for a larger number of motor units. axon that is activated first in one trial may be
With this technique, the size of the SMUP activated second or third in subsequent tri-
is estimated from the incremental change als. Thus, the sizes of potentials that could be
in the CMAP that occurs with control of obtained when there are three motor units of
the stimulus current to demonstrate the pro- different sizes—A, B, and C—are those gener-
gressively increasing number of motor units. ated by A alone, B alone, C alone; by A and C,
The more of these distinct steps of the total B and C, A and B; and by A, B, and C together.
CMAP that can be measured, the more reliable Thus the three SMUPs might be recorded as
MUNE becomes with incremental measure- three to seven steps.
ments. Normal nerves have so many motor A third problem with the incremental tech-
axons that incremental steps becomes more nique is the inability to separate very small
difficult to separate. Thus, the incremental potential, as occurs in severe myopathies, facial
method is truly reliable only with markedly muscles, or with nascent MUPs. The inability
reduced numbers of axons. to record the smallest steps results in underes-
Incremental MUNE using multiple current timation of MUNE in myopathies.
steps to selectively stimulate different axons Several modifications have been developed
has two major potential sources of error. First, to minimize some of these errors: (1) use
there may be a selection bias for larger or of automated computer identification of the
smaller motor units. Second, and of greater templates of different SMUPs allows recog-
concern, is that the occurrence of alternation nition of different SMUPs and decreases the
that occurs when different axons have simi- likelihood of measuring alternation, (2) use
lar thresholds for activation. Alternation is best of recording electrodes of different sizes and
illustrated by the example of a muscle contain- shapes to selectively stimulate different axons,
ing three axons of nearly the same threshold (3) automation of the incrementing stimulus
that result in seven rather than three appar- size to allow finer control,34 and (4) micros-
ent increments (Fig. 29–5). The threshold of timulation of single nerve terminals at the end
plate region.25, 26
Each of the single point incremental CMAP
techniques uses the average values of the size
(amplitude or area) of all the SMUPs iden-
tified and compares them with the maximal
size of the CMAP. Normal values determined
by several authors have shown that the mean
normal MUNE for median or thenar mus-
cles is approximately 350 (range, 100–500) and
for the peroneal or extensor digitorum bre-
vis, the other well-studied muscle, 200 (range,
50–300).
Despite the problems, incremental CMAP
is simple and direct enough that it should
be learned by every electromyographer. Many
patients with a neurogenic disease have low-
amplitude CMAP providing a measure of the
severity of axon loss. However, in some slowly
Figure 29–5. Alternation of three SMUPs during F- progressing processes or with residuals of old
wave recording to give seven F waves (A–G). (From processes, CMAP amplitude may be normal
Feasby, T. E., and W. F. Brown. 1974. Variation of motor
unit size in the human extensor digitorum brevis and
through collateral sprouting of intact axons
thenar muscles. Journal of Neurology, Neurosurgery, and despite the loss of axons. In such patients,
Psychiatry 37:916–26. By permission of the publisher.) recording the CMAP at a high sensitivity
Figure 29–6. Examples of maintenance of CMAP amplitude with marked loss of axons. Comparison of ulnar CMAP
amplitude with MUNE in patients with ALS. The lower limit of normal MUNE is 100 and the lower limit of normal ulnar
CMAP is 6.0 μV. Note that 13 patients have low MUNE with normal CMAP, and some as low as eight motor units.
(100–200 μV) with fine stimulus control near • The incremental method is less commonly
threshold will demonstrate the presence of used than other methods because of tech-
individual, high-amplitude SMUP. Some EMG nical problems.
machines provide automated stimulus control
to make this even easier to do. This method
can demonstrate as few as 3–10 remaining MULTIPOINT STIMULATION (MPS)
SMUPs with normal CMAP amplitude when
there should be over 100 (Fig. 29–6). The second method of isolating SMUP by
peripheral nerve stimulates axons at very low
intensity to activate a single axon and record
its SMUP. Stimulation applied at different
Key Points
points along the nerve (MPS) can isolate sin-
• Single point, incremental stimulation near gle SMUP at each point,14, 35–37 thereby elim-
threshold can sequentially isolate the inating the problem of alternation. MPS has
SMUP from different nerves. become one of the two most commonly used
• Normal values for standardized incremen- MUNE methods. The method is conceptually
tal MUNE methods have been deter- simple, but requires the manual skills to main-
mined. tain the position of the stimulator well enough
• Alternation of activation of axons with to repeatedly activate the same SMUP with
similar thresholds can erroneously sig- small stimulus variation at the axon threshold.
nificantly increase the MUNE with this Unique axons are identified by the record-
method particularly if the number of ing of distinct, reproducible SMUP. Stimu-
axons is near normal. lation is applied at as many sites as pos-
• Alternation occurs from the summation of sible at short intervals along the nerve to
similar size SMUPs in what appear to be selectively activate different axons (Fig. 29–7).
more SMUPs than are actually present. MUNE is calculated by dividing the maximal
• Averaging 10 or more such SMUPs pro- CMAP by the average size of the individual
vides a reasonable average size to divide SMUP.
into the maximal CMAP for MUNE. While MPS effectively eliminates the prob-
• Other common technical problems of lem of alternation, the other problems remain.
sample size, variation in SMUP size, A nerve with normal numbers of axons requires
and SMUP decrement can also decrease fine control of the stimulus to reliably isolate
MUNE value and reproducibility. individual axons. This can be enhanced by a
Figure 29–7. Example of data recorded for MPS. A, Traces 1–10 are single SMUP recorded over thenar muscles with
threshold stimulation at ten different points along the median nerve. Their mean area is 167 μVms. B, Thenar CMAP
recorded from the same thenar electrodes has an area of 46,793 μVms. Dividing CMAP area by average SMUP area gives a
MUNE of 280 (normal). (From Doherty, T. J., and W. F. Brown. 1993. The estimated numbers and relative sizes of thenar
motor units as selected by multiple point stimulation in young and older adults. Muscle & Nerve 16:355–66. By permission
of the publisher.)
nonstandard stimulus control that allows stim- to isolate an SMUP, (3) a myopathy with
ulus gradation in 0.1 mA steps. Even so, it SMUP too small to distinguish, (4) SMUP
is often difficult to isolate more than 15–20 selection bias by selective stimulation of low-
SMUPs. That number would be sufficient threshold, large motor axons, and (5) depend-
if the range of SMUP size was generally ing on the other possible technical factors MPS
the same. However, in nerves with reduced MUNE for one nerve can take 5–10 min-
numbers of axons and variation in collateral utes.
sprouting, the SMUP size can vary greatly. Despite these problems, MPS provides suf-
Sample size could then be insufficient to ficient reliability for clinical use (Fig. 29–8).
make a reproducible MUNE. Other possi- This figure also shows the marked improve-
ble problems include (1) defects of neuro- ment in reproducibility as the total number
muscular transmission that cause sufficient of axons in a nerve decreases.38 A variation of
variation in SMUP size to make the recog- MPS isolating only three SMUPs at each point
nition of individual SMUP less certain, (2) makes the recording more efficient with similar
patient tremor or other movement when trying values and reproducibility.39
Figure 29–8. Reproducibility of MUNE with MPS method. A, Subjects with decreased MUNE because of amyotrophic
lateral sclerosis. B, Normal subjects. (From Felice, K. J. 1995. Thenar motor unit number estimates using the multiple
point stimulation technique: Reproducibility studies in ALS patients and normal subjects. Muscle & Nerve 18:1412–16. By
permission of John Wiley & Sons.)
Key Points submaximal stimulation and template match-

ing may make the method useful clinically.
• MPS records individual SMUP by selec-
Simmons et al.42 have described an automated
tive stimulation of different low-threshold method of measuring only recurrent, identi-
axons at different points along a nerve.
• Awareness of a limited number of technical F waves that are more likely to represent
SMUP. This method has not been used widely
cal problems can make this method clini- because of the difficulty with implementing it.
cally useful.
• MUNE values for different nerves in nor- Nonetheless, F waves can be useful in judg-
ing the extent of collateral sprouting that
mal subjects are in the same range as those results from loss of axons. The remaining indi-
for single point incremental stimulation.
vidual motor units gain muscle fibers with an
increase in size of the SMUP recorded as F
waves. F waves from single SMUP can be rec-
F-WAVE MEASUREMENTS ognized as large, identical F waves (Fig. 29–9).
F waves have been suggested as a method for Their presence is confirmatory evidence of sig-
determining the size of the SMUP.40, 41 Max- nificant loss of axons, even if CMAP amplitude
imal activation of all the motor axons in a remains normal.
peripheral nerve is associated with antidromic
activation of some of the anterior horn cells. Key Points
The small proportion of anterior horn cells • F waves represent one or a combination of
activated antidromically produces small late a few SMUPs.
potentials, F waves. Repeated maximal stim- • As axons are lost, F waves are more likely
uli activate different anterior horn cells and to represent single SMUP.
produce different F waves. Recording a range • Published reports of MUNE from F-wave
of sizes of F waves can be used to estimate sizes have not been sufficiently accurate
the average size of the SMUP. This average or reproducible to provide clinically useful
size can be divided into the maximal CMAP to MUNE.
obtain MUNE. Simmons et al.42 have shown • Large reproducible F waves provide non-
that the drawbacks of this method are simi- quantitative evidence of loss of axons.
lar to those described in the preceding section
for the incremental method and that multi-
ple motor units are activated more commonly Statistical (STAT) MUNE
than single motor units. These drawbacks
result in overestimating the average size of the The fourth method of estimating the size of
SMUP and in underestimating the MUNE. the SMUP uses direct stimulation of the motor
Automated correction of these drawbacks by nerve, similar to the all-or-none incremental
Figure 29–9. Peroneal motor conduction study. Left, Maximal and threshold recordings at 1000 μV per division. Right,
F-wave recordings at 200 μV per division. Note that a 600-μV SMUP was recorded twice as identical F waves. This large
SMUP from collateral sprouting in a peripheral neuropathy makes up 40% of the 2.4-μV M wave.
method, but it is conceptually different.11, 43, 44 the following assumptions: each motor unit
With the STAT MUNE, no attempt is made has a similar size; it is the same size each
to identify the potentials associated with time it is activated; the samples tested are
individual motor units. The method relies on unbiased; and all motor axons are activated.
the known relationship between the variance Two other issues must be considered for STAT
of multiple measures of step functions and the MUNE. First, with a larger number of SMUP
size of the individual steps when the steps making up the CMAP, the distributions shift
have a Poisson distribution. Poisson statistics gradually from a Poisson to a normal distri-
is used to calculate the number of quanta bution. This may produce an error of up to
released from a nerve terminal at the neuro- 10% in the STAT MUNE. Second, because
muscular junction when the individual quanta all measurements are statistical, the results
are too small to be distinguished, as in myas- vary with each sample. Consequently, the num-
thenia gravis. In Poisson statistics, the sizes of ber of samples must be increased to provide
a series of measurements are multiples of the reproducibility comparable to that of the other
size of a single component. Therefore, a Pois- methods.
son distribution has discrete values at which In the STAT MUNE, recording electrodes
responses are found. A Poisson distribution has are applied as they are for standard NCS, with
decreasing numbers of responses with higher the stimulating electrode taped firmly in place
values. In a Poisson distribution, the variance over the appropriate nerve. An initial “scan”
of series of measurements is equal to the size of the responses of the nerve to 30 stimuli of
of the individual components that make up equal increments between threshold and max-
each measurement. The STAT MUNE method imal CMAP identifies large increments that
measures the variance of the CMAP and does may result from a large SMUP (Fig. 29–10).
not require identification of individual compo- Such large SMUPs do not need further statis-
nents; it can be used when the sizes of SMUP tical testing. A sequence of 30 or more sub-
are too small to be isolated, which is often the maximal stimuli is given at a fixed stimulus
case in normal muscles and myopathies. Also, it intensity in selected regions of small incre-
can be used with high-amplitude CMAPs that ments on the scan. The threshold of individual
require gains at which the SMUP cannot be motor axons fluctuates so that at any given
isolated. intensity the likelihood of firing ranges from
Although the STAT MUNE has advantages, 0% to 100%, with a finite range where the
it has the potential sources of error described axon fires only some percentage of the time
above for the other methods. These include (Fig. 29–11). This inherent variability of the
Figure 29–10. MUNE scan from, A, normal subject and, B, patient with amyotrophic lateral sclerosis. Between the
threshold and the maximal CMAP, 30 equal increments in stimulus intensity were applied to the nerve. The elicited CMAPs
are superimposed above the histograms. The histograms depict the area of each of the 30 responses. In A, note the smooth
curve with small increments. In B, the increments are larger, with a particularly large increment just before the maximal
CMAP. The latter is caused by activation of a single large motor unit.
threshold of individual axons causes intermit- a muscle with unstable motor units such as
tent firing of axons and continuous variations in amyotrophic lateral sclerosis (ALS);49 but a
the size of the CMAP. Therefore, the problem number of technical issues that will not be
of alternation with activation of different motor reviewed here were not addressed in that pub-
units described for the incremental and F-wave lication. Another recent study reported further
methods is not an issue with the STAT MUNE. improvements in STAT MUNE.50
Because the method is a statistical measure- Key Points
ment, a somewhat different result is obtained
each time, and multiple trials are needed to • STAT MUNE uses a distinctly different
obtain the most accurate measurement. Exper- method of CMAP recording to determine
imental testing with trials of more than 300 the number of axons innervating a muscle
stimuli has shown that repeated measurement group that provides a number of distinct
of groups of 30 until the standard deviation of advantages over other methods.
the repeated trials is less than 10% provides • The variance of the changes in the sizes
a close estimate of the number obtained with of the CMAP due to intermittent firing of
many more stimuli. Ongoing studies by differ- motor unit axons at threshold provides an
ent investigators have shown improvement in estimate of the sizes of the SMUP that are
the reproducibility and reliability of the STAT firing.
MUNE.42, 45–48 • The CMAP variance is measured at four
One recent study has raised questions about different levels of stimulation to sample
the validity of STAT MUNE when testing the entire range of axons in the nerve.
Figure 29–11. Statistical method of MUNE in a normal subject. A, Normal variation in CMAPs with 30 equal stimuli. B,
Calculated SMUPs and MUNE for four groups of 30 stimuli (table on right) and two runs at different stimulus intensities
(table on left). Superimposed CMAPs from the 30 stimuli are shown on top left in A and B.
509
CLINICAL APPLICATIONS spinal muscular atrophy, and ALS. Serial mea-

surements of MUNE are helpful in following
Applications of the methods of MUNE to dis- the course of a disease, especially in treat-
eases have shown the expected decrease in the ment trials or studies of the evolution of
number of motor units in peripheral nerve dis- disease.59–62
ease (carpal tunnel syndrome, lumbar radicu- MUNE has also been applied to assess the
lopathies, and peripheral neuropathies) and in timing of axonal loss during the course of inher-
motor neuron disease51–58 (Fig. 27–12). MUNE ited (SOD1) motor neuron disease. Remark-
is particularly helpful in chronic disorders in ably, in that disorder, axonal loss does not
which collateral sprouting results in a CMAP begin until only a few months before weakness
of normal amplitude despite a loss of motor appears, despite the lifelong presence of the
units such as residuals of old poliomyelitis, genetic abnormality (Fig. 29–13).63
Figure 29–12. Decrease in thenar MUNE (STAT MUNE) in 16 patients with amyotrophic lateral sclerosis. Note the
rapid decrease in MUNE over a few weeks in patients whose initial values were normal, but the decrease was slower after
the initial reduction in MUNE. F, female; M, male. Numerals following F and M indicate the age of patients.
Figure 29–13. Statistical MUNE from thenar muscles in normal subjects (“controls”), patients with sporadic ALS, known
carriers of the SOD1 ALS genetic defect, and SOD1 negative subjects in families with the genetic defect. Note that none of
the carriers had manifested the disease at the time of testing, but did subsequently with a course similar to the ALS patients
shown in Figure 29–12. (From Aggarwal, A., and G. Nicholson. 2009. Motor unit number estimates in patients with SOD1
positive and negative amyotrophic lateral sclerosis in motor unit number estimation. In Motor Unit Number Estimation and
Quantitative EMG, 60, ed. M. B. Bromberg. Proceedings of the Second International Symposium on MUNE and QEMG,
Snowbird, Utah, USA, August 19–20, 2006. Philadelphia: Elsevier. By permission of the publisher.)
SUMMARY 8. Schulte-Mattler, W. J., D. Georgiadis, K. Tietze, and

S. Zierz. 2000. Relation between maximum discharge
The all-or-none increment, the STA, DQEMG, rates on electromyography and motor unit number
estimates. Muscle & Nerve 23:231–8.
MPS, and the STAT MUNE methods of 9. Lateva, Z. C., K. C. McGill, and C. G. Burgar. 1996.
obtaining MUNE give similar values in each Anatomical and electrophysiological determinants of
of the muscles compared in normal subjects. the human thenar compound muscle action potential.
Alternative MUNE methods continue to be Muscle & Nerve 19:1457–68.
10. Armon, C., M. E. Brandstater, and G. W. Peterson.
developed, but none have been used suffi- 1997. Motor unit number estimates and quantitative
ciently in a clinical setting to assess them.64 muscle strength measurements of distal muscles in
In addition, Felice has shown that MUNE patients with amyotrophic lateral sclerosis. Muscle &
is more reliable than other measurements in Nerve 20:499–501.
documenting the course of ALS.38 Thus each 11. Daube, J. R. 1995. Estimating the number of motor
units in a muscle. Journal of Clinical Neurophysiology
of these methods can be used whenever and 12:585–94.
wherever it is most feasible and that different 12. Doherty, T. J., and W. F. Brown. 1993. The estimated
methods may be appropriate in different set- numbers and relative sizes of thenar motor units as
tings. For example, when each of the motor selected by multiple point stimulation in young and
older adults. Muscle & Nerve 16:355–66.
units can be identified with small increments of 13. Doherty, T., Z. Simmons, B. O’Connell, et al. 1995.
stimuli in a severe neurogenic process, MUNE Methods for estimating the numbers of motor units in
is defined most rapidly and accurately by actu- human muscles. Journal of Clinical Neurophysiology
ally counting the total number of increments. 12:565–84.
When the number of motor units is too large to 14. Felice, K. J. 1995. Thenar motor unit number esti-
mates using the multiple point stimulation technique:
do this or their size is too small for them to be Reproducibility studies in ALS patients and normal
identified accurately in the CMAP, the STAT subjects. Muscle & Nerve 18:1412–16.
MUNE or multipoint methods are appropri- 15. Roos, M. R., C. L. Rice, and A. A. Vandervoort. 1997.
ate. In muscles in which CMAPs cannot be Age-related changes in motor unit function. Muscle &
Nerve 20:679–90.
obtained reliably, as in proximal muscles that 16. Hennings, K., E. N. Kamavuako, and D. Farina. 2007.
are difficult to immobilize during stimulation The recruitment order of electrically activated motor
of the motor nerve, the STA and DQEMG neurons investigated with a novel collision technique.
methods are most appropriate. The value of Clinical Neurophysiology 118(2):283–91.
other methods of MUNE remains to be deter- 17. Slawnych, M., C. Laszlo, and C. Hershler. 1997. Motor
unit number estimation: Sample size considerations.
mined.65–69 Muscle & Nerve 20:22–8.
18. Bromberg, M. B. 2007. Updating motor unit num-
ber estimation (MUNE). Clinical Neurophysiology
118:1–8.
REFERENCES 19. van Dijk, J. P., J. H. Blok, B. G. Lapatki, I. N, van
Schaik, M. J. Zwarts, and D. F. Stegeman. 2008.
1. Armon, C., and M. E. Brandstater. 1999. Motor unit Motor unit number estimation using high-density
number estimate-based rates of progression of ALS surface electromyography. Clinical Neurophysiology
predict patient survival. Muscle & Nerve 22:1571–5. 119(1):33–42.
2. Johnston, S. C. 2000. Prognostication matters. Muscle 20. Bromberg, M. B. 1993. Motor unit estimation: Repro-
& Nerve 23:839–42. ducibility of the spike-triggered averaging technique in
3. Chan, K. M., and W. F. Brown. 1998. Quantitation normal and ALS subjects. Muscle & Nerve 16:466–71.
in EMG. Canadian Journal of Neurological Sciences 21. Bromberg, M. B., and J. L. Abrams. 1995. Sources of
25(1):S27–31. error in the spike-triggered averaging method of motor
4. Daube, J. R. 2006. Motor unit number estimates— unit number estimation (MUNE). Muscle & Nerve
From A to Z. Journal of Neurological Sciences 242 18:1139–46.
(1–2):23–35. 22. Brown, W. F., T. J. Doherty, M. Chan, A. Andres, and
5. Feinstein, B., B. Lindegård, E. Nyman, and G. Wohl- S. M. Provost. 2000. Human motor units in health and
fart. 1955. Morphologic studies of motor units in disease. Muscle & Nerve 9(Suppl):S7–18.
normal human muscles. Acta Anatomica Scandinavica 23. de Koning, P., G. H. Wieneke, D. van der Most van
23:127–42. Spijk, A. C. van Huffelen, W. H. Gispen, and F. G.
6. Azzouz, M., N. Leclerc, M. Gurney, J. M. Warter, P. Jennekens. 1988. Estimation of the number of motor
Poindron, and J. Borg. 1997. Progressive motor neu- units based on macro-EMG. Journal of Neurology,
ron impairment in an animal model of familial amy- Neurosurgery, and Psychiatry 51:403–11.
otrophic lateral sclerosis. Muscle & Nerve 20:45–51. 24. Milner-Brown, H. S., and W. F. Brown. 1976. New
7. Erim, Z., C. J. De Luca, K. Mineo, and T. Aoki. 1996. methods of estimating the number of motor units in
Rank-ordered regulation of motor units. Muscle & a muscle. Journal of Neurology, Neurosurgery, and
Nerve 19:563–73. Psychiatry 39:258–65.
25. Arasaki, K., M. Tamaki, Y. Hosoya, and N. Kudo. 1997. of motor unit number estimation in individual sub-
Validity of electromyograms and tension as a means jects. Muscle & Nerve 24:467–73.
of motor unit number estimation. Muscle & Nerve 43. Olney, R. K., and C. Lomen-Hoerth. 2000. Motor unit
20:552–60. number estimation (MUNE): How may it con-
26. Arasaki, K., and M. Tamaki. 1998. A loss of functional tribute to the diagnosis of ALS? Amyotrophic Lat-
spinal alpha motor neurons in amyotrophic lateral eral Sclerosis & Other Motor Neuron Disorders
sclerosis. Neurology 51:603–5. 1(Suppl 2):S41–4.
27. McNeil, C. J., T. J. Doherty, D. Stashuk, and C. L. 44. Yuen, E. C., and R. K. Olney. 1997. Longitudinal study
Rice. 2005. Motor unit number estimates in the tib- of fiber density and motor unit number estimate in
ialis anterior muscles of young, old and very old men. patients with amyotrophic lateral sclerosis. Neurology
Muscle & Nerve 31(4):461–7. 49:573–8.
28. Boe, S. G., D. W. Stashuk, W. F. Brown, and 45. Aggarwal, A., and G. Nicholson. 1998. Motor unit
T. J. Doherty. 2005. Decomposition-based quantita- number estimation in preclinical and amyotrophic
tive electromyography: Effect of force on motor unit lateral sclerosis subjects (abstract). Muscle & Nerve
potentials and motor unit number estimates. Muscle 7(Suppl):S203.
& Nerve 31(3):365–73. 46. Lomen-Hoerth, C., and R. Olney. 1999. A compar-
29. Boe, S. G., D. W. Stashuk, and T. J. Doherty. 2006. ison of multipoint and statistical motor unit number
Within-subject reliability of motor unit number esti- estimation (abstract). Muscle & Nerve 22:1304.
mates and quantitative motor unit analysis in distal and 47. Olney, R. K., E. C. Yuen, and J. W. Engstrom.
proximal upper limb muscles. Clinical Neurophysiol- 2000. Statistical motor unit number estimation: Repro-
ogy 117(3):596–603. ducibility and sources of error in patients with
30. Boe, S. G., D. W. Stashuk, and T. J. Doherty. 2007. amyotrophic lateral sclerosis. Muscle & Nerve 23:
Motor unit number estimates and quantitative motor 193–7.
unit analysis in healthy subjects and patients with amy- 48. Shefner, J. M., D. Jillapalli, and D. Y. Bradshaw.
otrophic lateral sclerosis. Muscle & Nerve 36(1):62–70. 1999. Reducing intersubject variability in motor unit
31. McComas, A. J. 1995. Motor unit estimation: Anxieties number estimation. Muscle & Nerve 22:1457–60.
and achievements. Muscle & Nerve 18:369–79. 49. Shefner, J. M., M. E. Cudkowicz, H. Zhang,
32. Galea, V., H. de Bruin, R. Cavasin, and A. J. McComas. D. Schoenfeld, and D. Jillapalli. 2007. Northeast ALS
1991. The numbers and relative sizes of motor units Consortium. Revised statistical motor unit number
estimated by computer. Muscle & Nerve 14:1123–30. estimation in the Celecoxib/ALS trial. Muscle & Nerve
33. Stein, R. B., and J. F. Yang. 1990. Methods for esti- 35(2):228–34.
mating the number of motor units in human muscles. 50. Hong, Y. H., J. J. Sung, K. S. Park, O. Kwon, J. H.
Annals of Neurology 28:487–95. Min, and K. W. Lee. 2007. Statistical MUNE: A
34. Aoyagi, Y., F. J. Strohschein, and K. Ming Chan. comparison of two methods of setting recording win-
2000. Use of the collision technique to improve the dows in healthy subjects and ALS patients. Clinical
accuracy of motor unit number estimation. Clinical Neurophysiology 118(12):2605–11.
Neurophysiology 111:1315–19. 51. Bromberg, M. B., D. A. Forshew, K. L. Nau,
35. Doherty, T. J., D. W. Stashuk, and W. F. Brown. J. Bromberg, Z. Simmons, and T. J. Fries. 1993.
1993. Determinants of mean motor unit size: Impact Motor unit number estimation, isometric strength, and
on estimates of motor unit number. Muscle & Nerve electromyographic measures in amyotrophic lateral
16:1326–31. sclerosis. Muscle & Nerve 16:1213–19.
36. Doherty, T. J., and W. F. Brown. 1994. A method for 52. Bromberg, M. B., and W. L. Larson. 1996. Rela-
the longitudinal study of human thenar motor units. tionships between motor-unit number estimates and
Muscle & Nerve 17:1029–36. isometric strength in distal muscles in ALS/MND.
37. Wang, F. C., and P. J. Delwaide. 1995. Number and Journal of Neurological Sciences 139(Suppl):38–42.
relative size of thenar motor units estimated by an 53. Brown, W. F., and T. E. Feasby. 1974. Estimates of
adapted multiple point stimulation method. Muscle & functional motor axon loss in diabetics. Journal of
Nerve 18:969–79. Neurological Sciences 23:275–93.
38. Felice, K. J. 1997. A longitudinal study comparing 54. Dantes, M., and A. McComas. 1991. The extent
thenar motor unit number estimates to other quantita- and time course of motoneuron involvement in amy-
tive tests in patients with amyotrophic lateral sclerosis. otrophic lateral sclerosis. Muscle & Nerve 14:416–21.
Muscle & Nerve 20:179–85. 55. Kuwabara, S., K. Mizobuchi, K. Ogawara, and
39. Guiheneuc, P. 1987. Automatic detection and pat- T. Hattori. 1999. Dissociated small hand muscle
tern recognition of motor unit potentials. Elec- involvement in amyotrophic lateral sclerosis detected
troencephalography and Clinical Neurophysiology by motor unit number estimates. Muscle & Nerve
39(Suppl):21–6. 22:870–3.
40. Felice, K. J. 1998. Nerve conduction velocities of 56. McComas, A. J., C. Quartly, and R. C. Griggs. 1997.
single thenar motor axons based on the automated Early and late losses of motor units after poliomyelitis.
analysis of F waves in amyotrophic lateral sclerosis. Brain 120:1415–21.
Muscle & Nerve 21:756–61. 57. Smith, B. E., J. C. Stevens, W. J. Litchy, et al. 1995.
41. Stashuk, D. W., T. J. Doherty, A. Kassam, and W. F. Longitudinal electrodiagnostic studies in amyotrophic
Brown. 1994. Motor unit number estimates based lateral sclerosis patients treated with recombinant
on the automated analysis of F-responses. Muscle & human ciliary neurotrophic factor (abstract). Neurol-
Nerve 17:881–90. ogy 45(Suppl 4):A448.
42. Simmons, Z., D. K. Epstein, B. Borg, D. T Mauger, 58. Krajewski, K. M., R. A. Lewis, D. R. Fuerst,
M. J. Kothari, and J. M. Shefner. 2001. Reproducibility et al. 2000. Neurological dysfunction and axonal
degeneration in Charcot-Marie-Tooth disease type 1A. on MUNE and QEMG, Snowbird, Utah, USA, August
Brain 123:1516–27. 19–20, 2006. Philadelphia: Elsevier.
59. Chan, K. M., D. W. Stashuk, and W. F. Brown. 64. Nandedkar, S. D., D. S. Nandedkar, P. E. Barkhaus,
1998. A longitudinal study of the pathophysiologi- and E. V. Stalberg. 2004. Motor unit number index
cal changes in single human thenar motor units in (MUNIX). IEEE Transactions on Biomedical Engi-
amyotrophic lateral sclerosis. Muscle & Nerve 21: neering 51(12):2209–11.
1714–23. 65. Fang, J., B. T. Shahani, and D. Graupe. 1997. Motor
60. Gooch, C. L., M. Pleitez, and Y. Harati. 1999. MUNE unit number estimation by spatial-temporal summa-
and single motor unit tracking: Experience in a clin- tion of single motor unit potentials. Muscle & Nerve
ical ALS trial. EEG and Clinical Neurophysiology 20:461–8.
107(Suppl 110):1001. 66. Roeleveld, K., A. Sandberg, E. V. Stalberg, and
61. Gooch, C. L. 1998. Repetitive axonal stimulation of the D. F. Stegeman. 1998. Motor unit size estimation of
same single motor unit: A longitudinal tracking study. enlarged motor units with surface electromyography.
Muscle & Nerve 21:1537–9. Muscle & Nerve 21:878–86.
62. Kuwabara, S., K. Ogawara, K. Mizobuchi, M. Mori, 67. Shahani, B. T., J. Fang, and U. K. Dhand. 1995.
and T. Hattori. 2001. Mechanisms of early and late A new approach to motor unit estimation with surface
recovery in acute motor axonal neuropathy. Muscle & EMG triggered averaging technique. Muscle & Nerve
Nerve 24:288–91. 18:1088–92.
63. Aggarwal, A., and G. Nicholson. 2009. Motor unit 68. Slawnych, M., C. Laszlo, and C. Herschler. 1996.
number estimates in patients with SOD1 positive and Motor unit estimates obtained using the new
negative amyotrophic lateral sclerosis in motor unit “MUESA” method. Muscle & Nerve 19: 626–36.
number estimation. In Motor Unit Number Estima- 69. Sun, T. Y., T. S. Lin, and J. J. Chen. 1999. Multi-
tion and Quantitative EMG, 60, ed. M. B. Bromberg. electrode surface EMG for noninvasive estimation of
Proceedings of the Second International Symposium motor unit size. Muscle & Nerve 22:1063–70.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART E
Reflexes and Central Motor
Control
Motor function is controlled by a complex com- useful primarily in elucidating central disorders
bination of central nervous system circuits that of motor function or neuronal excitability.
involve all levels of the neuraxis. Local reflexes They have helped elucidate the central disor-
at the level of the spinal cord or brain stem der in disorders with upper and lower motor
mediate and integrate local sensory input and neuron involvement, like amyotrophic lateral
input from descending motor pathways to the sclerosis.
motor unit. Descending motor pathways from Multichannel surface electromyographic
the cerebral hemisphere to the spinal cord recordings from agonist and antagonist mus-
include the rapidly conducting direct corti- cles in the limbs and trunk can be used to char-
cospinal pathways and several indirect path- acterize several motor disorders on the basis
ways that arise in the spinal cord and brain of the patterns of activation and the timing of
stem. Descending motor activity in these path- activity in different muscles, either in one limb
ways is directed and controlled by motor areas or longitudinally in the body (Chapter 33).
in the cerebral cortex, basal ganglia, and cere- New knowledge has allowed improvement in
bellum. The cerebellum and basal ganglia form the analysis and classification of tremor.
feedback loops that extend through the thala- Surface electromyographic recordings in
mus to the cerebral cortex and control motor posturography and electronystagmography are
activities. also used in measuring the motor control of
Many of these pathways and functions posture and vestibular function. These mea-
can be monitored electrically, as described surements (Chapter 34) assess the long path-
in the chapters of this section. H reflexes ways that control motor function and their
(Chapter 30) and cranial nerve reflexes integration in the neuronal pools. Posturogra-
(Chapter 31) are localized responses of the phy and electronystagmography are useful in
motor neurons in the spinal cord and brain evaluating many disorders of both the vestibu-
stem to localized sensory input. Both groups of lar pathways and the motor control pathways.
reflexes can be used to assess peripheral sen- Their applications have been expanded with
sory and motor function as well as their central new approaches to Dix-Hallpike, dynamic
connections in the spinal cord and brain stem. walking, and optokinetic rotary chair testing.
In contrast, long-loop reflexes and the silent These new approaches have helped increase
period depend more on the descending motor the application of the tests in Parkinson’s
pathways from the brain to the spinal cord and Alzheimer’s diseases, and in vestibular
(Chapter 32). Therefore, these reflexes are rehabilitation.
517
Chapter 30
H Reflexes
Ruple S. Laughlin
INTRODUCTION PEDIATRIC H REFLEXES

PHYSIOLOGIC BASIS CLINICAL APPLICATIONS
TECHNIQUE Proximal Conduction
Soleus Technique Central Nervous System Excitability
Gastrocnemius Technique
SUMMARY
Flexor Carpi Radialis Technique
INTRODUCTION for the H-reflex test. In the research litera-

ture, however, the H reflex is often discussed
In 1918, Hoffmann1 recorded a late response as a test of motor neuron excitability and used
from the soleus muscle that occurred with sub- to assess conditions that may cause spasticity
maximal electric stimulation of the tibial nerve, or rigidity, such as myelopathy, motor neu-
similar to the response recorded as a result of ron disease, Parkinson’s disease, and cerebellar
a tendon tap. Magladery and McDougal2 later disorders.
termed this response the Hoffmann reflex, or
H reflex. This reflex has been studied widely Purpose and Role of H Reflex
and has been found useful in assessing exper-
imental and clinical aspects of disorders of the • To serve as the electrophysiologic corre-
central and peripheral nervous systems. The late for the ankle jerk.
H reflex may be obtained from other muscles, • To assess proximal nerve segments that
including the gastrocnemius and flexor carpi are often inaccessible by routine nerve
radialis. conduction studies.
Lachman et al.3 later showed that the H • To serve as a tool in assessment of motor
reflex can be used to evaluate proximal nerve neuron excitability.
segments that are generally inaccessible by
routine nerve conduction studies in cases of
suspected radiculopathy, plexopathy, or sen- PHYSIOLOGIC BASIS
sory neuropathy, especially if routine nerve
conduction studies do not detect an abnormal- Stimulation of the tibial nerve in the popliteal
ity that is suspected clinically. These are the fossa at intensities less than needed to generate
most practical and clinically applicable reasons an early compound muscle action potential
519
(M response), from the soleus muscle elicits The latency of the H reflex depends on sev-
the H reflex. Magladery and McDougal2 first eral factors. These include the time to activate
demonstrated that the H reflex is a monosy- the primary Ia afferents, conduction veloc-
naptic reflex response produced by activation ity of the primary afferents, central conduc-
of a small proportion of group Ia afferents trav- tion delay, conduction velocity of motor axons
eling orthodromically to alpha motor neurons and terminal conduction delay, neuromuscu-
in the spinal cord (Fig. 30–1). Although similar transmission delay, distance from the site of
lar to a tendon reflex, the H reflex is evoked stimulation to the spinal cord, and the time to
by direct activation of the afferents bypassing detect a compound muscle action potential by
the muscle spindle and the influence of sen- the recording electrode.5, 6
sory endings and gamma motor neuron activ- The amplitude of the H reflex, and hence
ity on spindle sensitivity. In the spinal cord the ability to record the reflex, is extremely
the Ia afferents make monosynaptic connec- variable. The value of measuring the ampli-
tions to the alpha motor neuron and initiates tude of the H reflex in clinical settings is
a volley of activation in the motor nerve trav- debated. Normal amplitude values of the tib-
eling orthodromically from the cell body to ial H reflex vary from 0.1 to 7 mV and
the muscle. With gradually increasing stimu- also vary with recording electrode position,
lus intensities, the amplitude of the H reflex stimulus intensity, stimulus duration, posture,
increases as more spindle afferents are acti- age, and temperature.7 Additionally, H-reflex
vated. However, as stimulus intensity increases amplitude is sensitive to the degree of motor
further causing motor axon activation, the H- neuron excitability. Repeated recordings of
reflex amplitude begins to decrease as more the H reflex may demonstrate moment-to-
and more of the reflex volley is blocked in moment amplitude variability by as much as
the motor axons by antidromically conducted 1.5 mV.8, 9 Activation of one group of motor
motor impulses toward the spinal cord. A stim- units can inhibit motor units of other muscle
ulus of long duration allows for more selective groups, presumably through recurrent inhibi-
activation of afferent axons, whereas a stimulus tion. Thus, agonist contraction can increase
of short duration increases the likelihood of the amplitude of the H reflex and antagonist
motor axon activation.4 contraction can decrease it.10 Furthermore,
H-reflex F wave
Stimulus Stimulus
Figure 30–1. Physiology of the H reflex: Selective activation of muscle spindle afferents and monosynaptic reflex response
of soleus motor axons. The F wave represents recurrent discharge of motor neurons.
H Reflexes 521
excitability can vary at different spinal cord seg- Key Points

ments. Studies of recovery of the H reflex have • The H reflex is a monosynaptic reflex usu-
demonstrated that it is slower for the soleus H
ally evoked with tibial nerve stimulation.
reflex than for the flexor carpi radialis H reflex, • The H reflex is evoked by direct activa-
presumably because the motor neurons in the
tion of the afferents bypassing the muscle
sacral segments have different characteristics
spindle and the influence of sensory end-
from those in the cervical segments.11 Thus,
ings and gamma motor neuron activity on
H-reflex amplitude alone may not be a reliable
spindle sensitivity.
measure for clinical studies.8, 9 • A stimulus of long duration is used when
Factors that affect the H reflex and that
performing H reflexes, to allow for more
should be considered in performing this study
selective activation of afferent axons.
are listed in Table 30–1. It is important to dis- • The amplitude of the H reflex, and hence
tinguish the H reflex from the F wave, which
the ability to record the reflex, is extremely
has a similar latency. The F wave is not a reflex
variable and a less reliable measure than
but represents recurrent discharge of several
latency.
motor neurons, with alpha motor neurons and
axons serving as both the afferent and effer-
ent pathways. This produces a smaller ampli-
tude wave of varying latency and morphology TECHNIQUE
obtained with supramaximal stimulation. The
F wave can be recorded from many skeletal The H reflex can be recorded from many mus-
muscles (Table 30–2), and it can be distin- cles in normal neonates as well as in adults
guished from an H reflex by paired stimulation, with upper motor neuron lesions. In normal
which blocks the recurrent F wave but not the adults, it is most reliably obtained from the
H reflex. soleus and flexor carpi radialis muscles and has
Table 30–1 Factors That Affect the Presence or Amplitude of the H Reflex1, 3, 7, 23
Suppression Facilitation
Contraction of antagonist muscles Mild contraction of agonists

Strong contraction of agonists Passive stretch
Passive shortening Labyrinthine vestibular stimulation
Strong electric stimulus Bite, grasp
Sleep
Vibration
Drugs (nicotine, pentobarbital, diazepam)
Stimulation rate <1/second
Tendon tap
Strong flexion / extension of neck muscles
Proximal ischemia
Spinal anesthesia
Table 30–2 Comparison of the H Reflex and F Wave in Normal Subjects8, 31, 32
H Reflex F Wave
Suppressed by supramaximal stimulation Maximal with supramaximal stimulation

Optimal with submaximal stimulation Infrequent with submaximal stimulation
Amplitude 50%–100% of Mmax Amplitude 10% of Mmax
Latency relatively constant Latency variable
Morphology constant Morphology variable
Recorded from forearm and leg Recorded from hand and foot muscles
been reported in the gastrocnemius, palmaris M response is obtained (Fig. 30–2). With fur-
longus, flexor carpi ulnaris, anterior tibialis, ther increase in voltage, the H-reflex amplitude
vastus medialis, masseter, extensor digitorum declines and eventually disappears as the M
communis, and ulnar-innervated intrinsic hand wave reaches maximal amplitude at supramax-
muscles.9 The H reflex is elicited most read- imal stimulus intensity.4, 7, 13
ily in the soleus muscle by stimulation of the When recorded from the soleus muscle, the
tibial nerve. Although electric stimulation is H reflex appears as a biphasic wave with a
used most commonly, an Achilles tendon tap or large negative deflection, which is normally
quick Achilles tendon stretch may also be used 50%–100% of the maximal amplitude of the
to elicit the reflex.9, 12 M wave (Fig. 30–2). Stimulus intensity should
be increased and decreased as needed until a
reproducible maximal H reflex is obtained.9, 14
Soleus Technique This technique is important in differentiating
the H reflex from the F wave (Table 30–2).
The H reflex can be easily elicited recording Latency is measured to the initial deflection
over the soleus muscle. Because the soleus from baseline. In adults, the H reflex usually
muscle is the primary source of the reflex com- occurs at 28–35 ms and varies with leg length
pound muscle action potential, recording over and age. Braddom and Johnson15 developed a
this muscle results in an initial negative deflec- nomogram and formula that allow for age and
tion and a larger waveform as compared to leg-length differences. However, the side-to-
recording over the gastrocnemius muscle. It side comparison of the latency of the H reflex
has also been shown that there is a large separa- is probably more widely used than the abso-
tion between the stimulation thresholds of the lute value and should vary by no more than
H and M waves in the soleus muscle, making it 3 ms9, 14–17 (Table 30–3). As previously noted,
easier to perform and interpret as compared to the reliability of amplitude measurements is
other muscles. not sufficient for clinical use.
To perform the soleus technique, if neces-
sary, the patient, may be placed supine to avoid
excessive lengthening or shortening of the Gastrocnemius Technique
soleus muscle and to allow adequate patient
relaxation. The active recording electrode is The H reflex can be elicited over the gastrocne-
placed medially over the soleus muscle approx- mius muscle. Because the soleus muscle is the
imately midway between the site of stimu- primary source of the reflex compound muscle
lation and the medial malleolus. The tibial action potential, recording over the gastrocne-
nerve is stimulated immediately lateral to the mius muscle results in a smaller waveform as
popliteal artery at the level of the popliteal compared to recording over the soleus mus-
crease. To avoid stimulation of the peroneal cle. For this reason, it should be noted that
nerve by current spread, the resultant plantar when the H reflex is recorded over the gas-
flexion muscle twitch of the gastrocnemius– trocnemius muscle, an initial positivity may
soleus contraction without associated dorsiflex- be present. To perform the study, the patient
ion of the anterior tibialis or peroneal muscles is placed supine as in the soleus technique
should be observed. The cathode is placed described above. The recording electrodes are
proximal to the anode. Square-wave pulses placed posteriorly between the two heads of
of 0.5–1.0-ms duration, with an intensity of the gastrocnemius muscle. Stimulation of the
10–80 mA, are delivered at a rate of 0.5 Hz. tibial nerve is applied in the same fashion as
Intensity should be increased by small incre- for soleus recordings. Latencies are similar for
ments from an initial low level until a response soleus and gastrocnemius recordings.
larger than the preceding M wave is elicited.
With gradually increasing stimulus intensity,
the H reflex usually appears before the M Flexor Carpi Radialis Technique
response or shortly thereafter. As the stimu-
lus voltage is increased, the amplitude of the The H reflex also can be recorded from
H reflex reaches a maximum before a maximal the flexor carpi radialis muscle. According to
H Reflexes 523
Figure 30–2. Top, H reflex recorded from soleus muscle with increasing stimulus intensity (from top to bottom). Note
maximal amplitude of the H reflex with submaximal stimulation and initial negativity of the waveform, allowing reliable
latency measurement. Bottom, Graphical representation of stimulus intensity vs. amplitude depicting increasing M-wave
amplitude with decreasing H-wave amplitude as stimulation intensity increases.
the method described by Jabre,18 the active Key Points

recording electrode is placed over the mid-
portion of the flexor carpi radialis, approxi- • The H reflex can be reliably elicited over
mately 10 cm distal to the elbow crease, with several muscles in adults, most commonly
the reference electrode over the brachioradi- the soleus, gastrocnemius, and flexor carpi
alis (Fig. 30–3). A flexor carpi radialis study radialis.
usually is performed with the patient seated
rather than supine to increase the likelihood • The soleus technique is the preferred
of recording a response.19 Normal values are method due to patient comfort and repro-
listed in Table 30–3. ducibility.
Table 30–3 Normal Values for H-Reflex Latency

Latency, ms
Reference Muscle recorded Mean Range Side-to-side

comparison
Braddom and Johnson15 Gastrocnemius — 28–35 <1.2

Kimura8 Soleus 29.5 27–35 <1.4
Mayo Clinic Soleus — 25–35 <3
Shahani and Young33 Soleus — ≤34 <2
Ongerboer de Visser, Flexor carpi radialis 16.8 15–20 <0.85
Schimsheimer, and
Hart16
Jabre18 Flexor carpi radialis 15.9 11–21 <0.7
Buschbacher5 Gastrocnemius 30.3 30–38 <2
Figure 30–3. Technique of H reflex recording from flexor carpi radialis with median nerve stimulation.
PEDIATRIC H REFLEXES Key Points

• Conduction velocities of the H reflex do
Cai and Zhang20 demonstrated that the con-
not reach adult values until 6 years of age.
duction velocities of the H reflex do not reach • In infants, H reflexes are easily elicitable
adult values until 6 years of age. Normal abso-
from calf and hand muscles.
lute latencies of the H reflex are correlated
with body length until 1 year of age.21 In
infants, H reflexes are easily elicitable from calf
and hand muscles. Mitsudome et al.22 showed CLINICAL APPLICATIONS
that although the evolution of motor velocities
is the same for both premature and full-term
infants, the evolution of the velocity of the Proximal Conduction
H reflex may be slower in premature than in
full-term infants. This factor should be consid- The H reflex is used best as a measure
ered when evaluating infants. of proximal conduction. For instance, it is
H Reflexes 525
most commonly used in assessing S1 radicu- the F wave in cases of peripheral neu-
lopathies. Braddom and Johnson15 proposed ropathy.3 Prolonged latency of the H reflex
that a side-to-side latency difference in the has been demonstrated in diabetic, alcoholic,
H reflex greater than 1.2 ms indicates a dis- nutritional, paraneoplastic, cisplatin, and vas-
crete lesion of the S1 nerve root, and may culitic neuropathies. In uremic neuropathy,
be the initial finding in acute radiculopathy this prolongation decreases after successful
or in cases of radiculopathy in which the only renal transplantation.24 Prolonged latency also
needle electromyographic findings are fibril- has been demonstrated in Guillain–Barré syn-
lation potentials in paraspinal muscles.14 The drome, chronic inflammatory polyradiculoneu-
prolongation or absence of the H reflex cor- ropathies, and hereditary mixed motor and sen-
relates with clinically reduced or absent ankle sory neuropathy type I.25 Because the H reflex
jerk. Any lesion along the pathway of the H may be absent in persons older than 60 years, a
reflex may prolong its latency. Moreover, if large number of whom also have lumbar steno-
the largest diameter axons are not affected by sis, the usefulness of the H reflex in studying
a root lesion, the H reflex may remain nor- neuropathy in this age group is limited.23, 26 The
mal. Therefore, additional corroborating elec- latency of the H reflex in the flexor carpi radi-
trophysiologic evidence is needed to support alis muscle has proved useful in assessing slow-
the diagnosis of S1 radiculopathy.19, 23 ing of conduction through the proximal median
The latency of the H reflex also may be pro- nerve fibers in radiation-induced plexopa-
longed in some cases of peripheral neuropathy thy.16, 25 This may be used to assess C7 radicu-
due to slowing of peripheral or proximal con- lopathy, comparable to assessing S1 radiculopa-
duction. The H reflex may disappear before thy with recordings from the soleus muscle.
Figure 30–4. H reflex recorded from the abductor digiti minimi during ulnar nerve stimulation in a patient with
cerebral palsy.
Central Nervous System It should be noted, however, that many

Excitability recent studies have questioned the stability of
the M wave in various circumstances includ-
If the H reflex is recorded from muscles other ing degree of muscle contraction and time of
than the soleus during standard nerve conduc- day.29 Therefore, it remains to be seen whether
tion studies in adults, the possibility of a central the Hmax/Mmax ratio is any more reliable
nervous system disorder must be considered from trial to trial as compared to the H-reflex
(Fig. 30–4). As the H reflex directly activates amplitude as a clinical tool.
the Ia afferents, it can be used as a measure The second way the H reflex has been uti-
of the excitability of the neuronal components lized as a measure of motor neuron excitability
of the arc, without concerns for muscle spin- is to produce H-reflex recovery curves, first
dle sense organ sensitivity.6, 8, 10, 27, 28 However, described by Magladery and McDougal.2 The
both the latency and amplitude of the H reflex H reflex is elicited before, during, and after
alone has been shown to have trial-to-trial vari- conditioning stimuli, usually electric stimuli, at
ability as previously mentioned. This can not a range of interstimulus intervals. It is impor-
only occur between subjects but also within the tant to keep the latency and amplitude of the
same subject from day to day. In efforts to more M-response constant during these trials. These
reliably study the excitability of the motor neu- curves also have been used to study spasticity.
ron pool using the H reflex, two methods have However, the changes in the curves are rarely
been proposed. striking. The curve is influenced greatly by the
The first way compares the ratio of maximal position and comfort of the patient, the angle
H-reflex amplitude with maximal M-response of lower extremity joints, relaxation, and the
amplitude (Hmax/Mmax). By expressing the H positioning of the head. Construction of these
reflex as a ratio relative to the M response, it curves is time-consuming, and reproducibil-
is thought that the influence of many variables ity is poor, making them impractical.10, 12 The
such as impedance and technical factors of size pathophysiologic basis of these curves is poorly
and electrode location is reduced.29 This ratio understood and as a result, they are seldom
may be 1 in normal subjects and increased in used in clinical practice.
disorders causing spasticity, although it corre-
lates poorly with the degree of spasticity noted
clinically. The Hmax/Mmax ratio is normal in Key Points
patients with rigidity.27 • Clinically, the H reflex is most utilized in
Many recent studies have extrapolated the
assessing early S1 radiculopathies.
Hmax/Mmax ratio not only as an indicator • The H reflex can be a useful adjunct
of an upper motor neuron lesion, but also
to routine nerve conduction studies in
as a means to assess clinical responsiveness
assessing proximal peripheral nerve disor-
in cases of spasticity. For instance, Stokie
ders.
and Yablon recently employed this ratio as • The H reflex and the Hmax/Mmax ratio
an adjunct for evaluating responsiveness to
may prove to be useful in assessment
intrathecal baclofen in cases of spasticity.30 In
of upper motor neuron excitability and
this study, the authors found that a decreas-
response to clinical therapy of upper
ing Hmax/Mmax ratio is seen with increasing
motor neuron lesions.
doses of baclofen. The most useful decline
in the ratio was seen early at lower doses of
baclofen, often before any clinical response
was noted. They also found that the ratio would
often nadir before any clinical benefit was seen
with high doses of baclofen. In this study the SUMMARY
Hmax/Mmax ratio was employed as a measure
of spinal cord responsiveness to baclofen ther- Technically, the H-reflex test is relatively sim-
apy rather than a direct reflection of spinal cord ple to perform in a clinical neurophysiology
spasticity. They employed this ratio, therefore, laboratory. A thorough understanding of the
as a pharmacological marker that would permit physiologic basis, sources of error, and clinical
more aggressive dose escalation of medication applications and limitations enhances the use-
in the absence of clinical responsiveness. fulness of the H reflex.
H Reflexes 527
REFERENCES 17. Falco, F. J., W. J. Hennessey, G. Goldberg, and R. L.

Braddom. 1994. H reflex latency in the healthy elderly.
1. Hoffmann, P. 1918. Ueber die Beziehungen der Muscle & Nerve 17:161–7.
Schnenreflexe zur willkürlichen Bewegung und zum 18. Jabre, J. F. 1981. Surface recording of the H-reflex of
Tonus. Zeitschrift für Biologie 68:351–70. the flexor carpi radialis. Muscle & Nerve 4:435–8.
2. Magladery, J. W., and D. B. McDougal Jr. 1950. Elec- 19. Sabbahi, M. A., and M. Khalil. 1990. Segmental H-
trophysiological studies of nerve and reflex activity in reflex studies in upper and lower limbs of healthy
normal man. I. Identification of certain reflexes in subjects. Archives of Physical Medicine and Rehabil-
the electromyogram and the conduction velocity of itation 71:216–22.
peripheral nerve fibres. Bulletin of the Johns Hopkins 20. Cai, F., and J. Zhang. 1997. Study of nerve conduction
Hospital 86:265–90. and late responses in normal Chinese infants, children,
3. Lachman, T., B. T. Shahani, and R. R. Young. 1980. and adults. Journal of Child Neurology 12:13–18.
Late responses as aids to diagnosis in peripheral neu- 21. Jones, H. R. Jr., R. L. Harmon, C. M. Harper Jr., and
ropathy. Journal of Neurology, Neurosurgery, and C. F. Bolton. 1996. An approach to pediatric elec-
Psychiatry 43:156–62. tromyography. In Pediatric clinical electromyography,
4. Panizza, M., J. Nilsson, and M. Hallett. 1989. Opti- ed. H. R. Jones Jr., C. F. Bolton, and C. M. Harper Jr.,
mal stimulus duration for the H reflex. Muscle & Nerve 1–36. Philadelphia: Lippincott-Raven Publishers.
12:576–9. 22. Mitsudome, A., S. Yasumoto, and H. Ogata. 1996.
Late responses in full-term newborn infants. In Recent
5. Buschbacher, R. M. 1999. Normal range for H-reflex
Advances in Clinical Neurophysiology: Proceedings of
recording from the calf muscles. American Jour-
the Xth International Congress of EMG and Clini-
nal of Physical Medicine & Rehabilitation 78(Suppl
cal Neurophysiology, ed. J. Kimura, and H. Shibasaki,
6):S75–9.
761–5. Kyoto, Japan, October 15–19, 1995. New York:
6. Fisher, M. A. 1992. AAEM minimonograph #13: H
Elsevier.
reflexes and F waves: Physiology and clinical indica-
23. Wilbourne, A. J., and M. J. Aminoff. 1993. Radicu-
tions. Muscle & Nerve 15:1223–33.
lopathies. In Clinical electromyography, ed. W. F.
7. Nishida, T., A. Kompoliti, I. Janssen, and K. F. Levin. Brown, and C. F. Bolton, 2nd ed., 177–209. Boston:
1996. H reflex in S-1 radiculopathy: Latency ver- Butterworth-Heinemann.
sus amplitude controversy revisited. Muscle & Nerve 24. Bolton, C. F. 1993. Metabolic neuropathy. In Clinical
19:915–17. electromyography, ed. W. F. Brown, and C. F. Bolton,
8. Funase, K., and T. S. Miles. 1999. Observations on the 2nd ed., 561–98. Boston: Butterworth-Heinemann.
variability of the H reflex in human soleus. Muscle & 25. Schimsheimer, R. J., B. W. Ongerboer de Visser,
Nerve 22:341–6. B. Kemp, and L. J. Bour. 1987. The flexor carpi
9. Kimura, J. 1989. Electrodiagnosis in diseases of nerve radialis H-reflex in polyneuropathy: Relations to con-
and muscle: Principles and practice, 2nd ed., 359–61. duction velocities of the median nerve and the soleus
Philadelphia: FA Davis Company. H-reflex latency. Journal of Neurology, Neurosurgery,
10. Iles, J. F., and J. Pardoe. 1999. Changes in transmission and Psychiatry 50:447–52.
in the pathway of heteronymous spinal recurrent inhi- 26. Wilbourn, A. J. 1993. Diabetic Neuropathies. In
bition from soleus to quadriceps motor neurons during Clinical electromyography, ed. W. F. Brown, and
movement in man. Brain 122:1757–64. C. F. Bolton, 2nd ed., 477–515. Boston: Butterworth-
11. Rossi-Durand, C., K. E. Jones, S. Adams, and P. Bawa. Heinemann.
1999. Comparison of the depression of H-reflexes 27. Leonard, C. T., and T. Moritani. 1992. H-reflex testing
following previous activation in upper and lower to determine the neural basis of movement disorders
limb muscles in human subjects. Experimental Brain of neurologically impaired individuals. Electromyogra-
Research 126:117–27. phy and Clinical Neurophysiology 32:341–9.
12. Rico, R. E., and E. J. Jonkman. 1982. Measure- 28. Misiaszek, J. E. 2003. The H-reflex as a tool in neu-
ment of the Achilles tendon reflex for the diagnosis of rophysiology: Its limitations and uses in understanding
lumbosacral root compression syndromes. Journal of nervous system function. Muscle & Nerve 28: 144–60.
Neurology, Neurosurgery, and Psychiatry 45: 791–5. 29. Tucker, K. J., M. Tuncer, and K. S. Turker. 2005.
13. Nishida, T., C. E. Levy, E. J. Lewit, and I. Janssen. A review of the H-reflex and M-wave in the human
1999. Comparison of three methods for recording triceps surae. Human Movement Science 24: 667–88.
tibial H reflex: A clinical note. American Journal of 30. Stokie, D. S., and S. A. Yablon. 2007. Neurophysio-
Physical Medicine & Rehabilitation 78:474–6. logical basis and clinical application of the H-reflex
14. DeLisa, J. A., H. J. Lee, E. M. Baran, K-S. Lai, and as an adjunct for evaluation response to intrathecal
N. Spielholz. 1994. Manual of nerve conduction veloc- baclofen for spasticity. Acta Neurochirurgica. Supple-
ity and clinical neurophysiology, 3rd ed., 168–75. New ment 97(Pt 1):231–41.
York: Raven Press. 31. Liveson, J. A., and D. M. Ma. 1992. Laboratory ref-
15. Braddom, R. I., and E. W. Johnson. 1974. Standard- erence for clinical neurophysiology. 238–46. Philadel-
ization of H reflex and diagnostic use in S1 radiculopa- phia: FA Davis.
thy. Archives of Physical Medicine and Rehabilitation 32. Maryniak, O., and R. Yaworski. 1987. H-reflex: Opti-
55:161–6. mum location of recording electrodes. Archives of
16. Ongerboer de Visser, B. W., R. J. Schimsheimer, and Physical Medicine and Rehabilitation 68:798–802.
A. A. Hart. 1984. The H-reflex of the flexor carpi 33. Shahani, B. T., and R. R. Young. 1980. Studies of reflex
radialis muscle; a study in controls and radiation- activity from a clinical viewpoint. In Electrodiagnosis
induced brachial plexus lesions. Journal of Neurology, in clinical neurology, ed. M. J. Aminoff, 209–304. New
Neurosurgery, and Psychiatry 47:1098–101. York: Churchill Livingstone.
Chapter 31
Cranial Reflexes and Related

Techniques
Benn E. Smith
INTRODUCTION MASSETER INHIBITORY REFLEX

BLINK REFLEX Methods
Neuroanatomy Applications
Methods
GREAT AURICULAR SENSORY
Applications
NERVE CONDUCTION STUDIES
LATERAL SPREAD OF THE FACIAL Methods
NERVE RESPONSE: Applications
ASSESSMENT OF FACIAL
TRIGEMINAL CONTACT HEAT
SYNKINESIS AND HEMIFACIAL SPASM
EVOKED POTENTIAL
JAW JERK (MASSETER REFLEX)
STIMULATOR STUDIES
Neuroanatomy
SUMMARY
Methods
Applications
INTRODUCTION in the head which is not a cranial nerve of

branchial arch origin, the great auricular sen-
The integrity of the trigeminal and facial sory nerve, and the other to interrogate trigem-
nerves and their central connections can be inal sensory pathways from the sensory recep-
evaluated electrophysiologically by studying tor level to the parietal cortex, contact heat
the reflex activity mediated by these nerves. evoked potential stimulator studies—will also
For this reason, these reflexes are of great- be discussed.
est value in assessing cranial neuropathies.
They can also provide useful information in
Purpose and Role of Cranial Reflexes
some cases of polyradiculoneuropathy, periph-
eral neuropathy, and brain stem lesions. The • To assess for cranial neuropathies.
reflexes discussed in this chapter are the elec- • Used in the evaluation of polyradiculo-
trically evoked blink reflex, the jaw jerk (or pathies or severe peripheral neuropathies.
masseter reflex), and the masseter inhibitory • Assist in assessing brain stem lesions.
reflex (MIR). In addition, two additional • To evaluate for facial movement disorders,
techniques—one to assess a sensory nerve such as hemifacial spasm or synkinesis.
529
BLINK REFLEX early response is relayed centrally through an

oligosynaptic pathway involving the principal
Overend1 observed that a blink response sensory nucleus of the trigeminal nerve in
occurs after a light tap on the forehead. Kugel- the pons.5 The afferent fibers relaying the R2
berg2 elicited the reflex with an electric stim- response descend into the medulla and synapse
ulus and demonstrated two distinct responses: in the spinal nucleus of the trigeminal nerve.
an early well-synchronized response occurring Through polysynaptic pathways that pass ipsi-
ipsilateral to the stimulus (R1) and poorly syn- laterally and contralaterally, the afferent limb is
chronized bilateral responses with a longer connected with the nucleus of the facial nerve,
latency (R2). Rushworth3 demonstrated that the efferent limb of the reflex.
the afferent limb of the reflex is carried by the
first division of the trigeminal nerve and the
efferent component is transmitted by the facial Methods
nerve. Shahani4 showed that the reflex was
mediated by stimulation of cutaneous recep- The blink reflex is elicited by mechanical
tors rather than proprioceptive receptors, as or electric stimulation over the face with a
previously thought. graded threshold, with the lowest threshold
being around the eye. The reflex is usually
elicited by stimulating the supraorbital branch
Neuroanatomy of the trigeminal nerve over the supraorbital
notch, while recording simultaneously from
The afferent limb of the reflex arc is transmit- both the left and the right orbicularis oculi
ted in the ophthalmic division of the trigem- muscles (Fig. 31–2). The stimuli are applied
inal nerve, with the afferent nerve cell body irregularly at least 5 seconds apart so as to
in the gasserian ganglion (Fig. 31–1). The minimize habituation. The stimulus current
efferent limb travels in the facial nerve. The required to evoke the reflex is small and not
generally considered painful. In normal sub-
jects, an early (R1) response is obtained ipsi-
lateral to the side of stimulus, and the late (R2)
responses are obtained bilaterally (Fig. 31–3).
Figure 31–1. Neural pathways associated with the blink

reflex. For the R1 response, afferent fibers from the
trigeminal nerve synapse with the principal sensory
nucleus in the pons (Vp). Connections are then made
with the facial nucleus (VII) via one or two interneurons.
For the R2 response, the trigeminal nerve fibers synapse Figure 31–2. Electrode placement for blink reflex stud-
with the spinal nucleus of the trigeminal nerve via a mul- ies. The supraorbital nerve is stimulated with the cathode
tisynaptic pathway. Connections are then relayed to the at the supraorbital notch. Responses are recorded ipsilater-
facial nuclei bilaterally. (From Ongerboer de Visser, V. W., ally and contralaterally.G1, active electrode;G2, reference
and H. G. J. M. Kuypers. 1978. Late blink reflex changes electrode. (From Auger, R. G. 1987. Brain stem disor-
in lateral medullary lesions: An electrophysiological and ders and cranial neuropathies. In Clinical electromyogra-
neuro-anatomical study of Wallenberg’s syndrome. Brain phy, ed. W. F. Brown, and C. F. Bolton, 417–29. Boston:
101:285–94. By permission of Oxford University Press.) Butterworths. By permission of the publisher.)
Cranial Reflexes and Related Techniques 531
response may be obtained by stimulating the

opposite trigeminal nerve. Amplitudes of the
responses are not usually measured because
differences in amplitude of up to 40% occur in
normal subjects.6
Normal values, in milliseconds, are as
follows:
Ipsilateral R1 < 13
Ipsilateral R2 < 41
Contralateral R2 < 44
Ipsilateral vs. contralateral differences, in mil-

liseconds, are as follows:
R1 < 1.2
R2 < 8
The R1 latency in young children reaches adult

Figure 31–3. Responses obtained with stimulation of the values by the age of 2 years. However, the
supraorbital nerve and simultaneously recording from both R2 responses are sometimes absent in children
left and right orbicularis oculi muscles. R1, first response; younger than 2 years. The R2 waveforms attain
R2, second response.
adult values at age 5–6 years. Because of the
variability of R2 responses, they are less useful
Later bilateral (R3) responses are also seen in in children younger than 6 years.7
normal subjects. With slowly increasing stim- The R2 response correlates with contrac-
ulation intensities, the R2 responses are usu- tion of the orbicularis oculi muscle and has the
ally elicited before R1. If an early response same latency as the corneal reflex. The physi-
cannot be obtained, a paired stimulus can be ologic significance of the R1 response remains
used, with an interstimulus interval of 5 ms to unknown.
take advantage of a period of facilitation that In patients with suspected lesions of the
may last 1–9 ms after the initial conditioning maxillary or mandibular division of the trigem-
stimulus (Fig. 31–4). Confusing results may inal nerve, the blink reflex obtained with
be obtained if the stimulating electrode is too supraorbital nerve stimulation may be nor-
close to the midline, because a contralateral R1 mal. In this situation, the infraorbital branch
(recording from the superior portion of the
orbicularis oculi) can be stimulated to assess
the function of the maxillary division. Stimula-
tion of the infraorbital nerve consistently elicits
R2 responses in normal subjects, but the R1
response is frequently absent. Stimulation of
the mental nerve (the mandibular division) also
elicits R2 responses, but the R1 response is
never observed.8
Applications
TRIGEMINAL NERVE LESIONS
In lesions of the trigeminal nerve, blink reflex
Figure 31–4. In this patient, the R1 response could not
be elicited with a single stimulus. A normal response was
responses are usually delayed or absent—when
obtained with paired stimuli, with an interstimulus interval recorded from either side—with stimulation
of 5 ms. of the involved side, but they are normal
Figure 31–6. Blink reflex studies in a patient with a

lesion of the right facial nerve. R1 max, upper limit of nor-
mal for the onset of the R1 response; R2 max, upper limit
of normal for the onset of the R2 response.
Figure 31–5. Blink reflex studies in a patient with a

lesion of the left trigeminal nerve. R1 max, upper limit of facial-innervated muscles are delayed or absent
normal for the onset of the R1 response; R2 max, upper with stimulation of the trigeminal nerve on
limit of normal for the onset of the R2 response. either side. However, the R1 and R2 responses
recorded from the unaffected facial muscles
bilaterally with stimulation of the unaffected are normal if trigeminal nerve function is pre-
side (Fig. 31–5). In severe lesions of the served (Fig. 31–6). As an illustrative example,
trigeminal nerve, stimulation of the involved in a patient with a right facial neuropathy, stim-
side may not elicit a response. In patients with ulation of the right supraorbital (trigeminal)
trigeminal sensory neuropathy associated with nerve will produce abnormally prolonged or
connective tissue diseases, there may be abnor- absent R1 and R2 responses from the right
malities of R1 only or bilateral abnormalities orbicularis oculi, but a normal contralateral R2
of R1 and R2; it is uncommon for only R2 to response is recorded from the left orbicularis
be affected in isolation.9 The blink reflex is also oculi, since the right trigeminal and left facial
helpful in identifying lesions of the sensory root nerves are intact. In the same patient, stimula-
and gasserian ganglion, including tumors, vas- tion of the left supraorbital nerve will produce
cular malformations, and postherpetic lesions. normal ipsilateral (left) R1 and R2 responses,
The blink reflex is usually normal in idio- but a prolonged or absent contralateral (right)
pathic trigeminal neuralgia and atypical facial R2 response. In severe lesions of the facial
pain.10 Therefore, an abnormal blink response nerve, the R1 and R2 responses may be absent
in a patient with suspected trigeminal neuralgia on the involved side.
would suggest a structural lesion involving the In Bell’s palsy, routine conduction studies
trigeminal nerve or brain stem. Trichloroethy- of the facial nerve evaluate the segment of
lene, long known to produce trigeminal neu- the nerve distal to the stylomastoid foramen,
ropathy, is associated with a prolonged mean which is clinically helpful only when degen-
latency of R1 in persons with occupational eration begins distally. The blink reflex, how-
exposure or a contaminated water supply.11 ever, assesses the entire nerve, including the
intraosseous portion. All the patients with Bell’s
palsy studied by Kimura, Giron, and Young12
FACIAL NERVE LESIONS
had a delayed or absent R1 response on the
In lesions of the facial nerve, the R1 and affected side of the face. On the weak side,
R2 responses recorded from the affected the R2 response was also abnormal in all
patients. Of 127 patients tested serially, 100

had return of the previously absent R1 and
R2 responses, whereas responses with stim-
ulation of the facial nerve at the mastoid
process remained relatively normal. The R1
latency was increased by more than 2 ms ini-
tially, suggesting demyelination of facial nerve
fibers. The latency of R1 decreased during
the second month and returned to normal by
the fourth month after onset. These patients
generally had good return of function within
a few months. The other 27 patients had
smaller amplitudes with direct stimulation of
the facial nerve, and the blink reflex responses
did not return. These patients had degener-
Figure 31–7. Assessment of facial synkinesis in a patient
ation of the facial nerve and poorer recover- with left hemifacial spasm. A synkinetic response is present
ies. More recent studies have reached similar in the left orbicularis oris with stimulation of the left
conclusions—when the blink reflex is normal supraorbital nerve. On the right (normal) side, synkine-
or R1 is only delayed, the prognosis is excellent. sis is not present. (From Auger, R. G. 1979. Hemifacial
spasm: Clinical and electrophysiologic observations. Neu-
Absence of the blink reflex has been associated rology 29:1261–72. By permission of Lippincott Williams
with a poor prognosis in 56% of cases.13, 14 & Wilkins.)
Delay or absence of only the R1 response
indicates dysfunction of either the trigeminal present in other movement disorders involv-
or facial nerve or the corresponding central ing the face, such as facial myokymia, ble-
connections in the brain stem. Clinical find- pharospasm, or Meige syndrome.
ings, for example numbness of the face, may Rubin et al.16 described a patient in whom
indicate which nerve is responsible. If both an unusual movement disorder affecting the
nerves are affected, as they may be in acous- muscles of mastication developed after surgi-
tic neuroma, it may be impossible to provide a cal removal of a tumor involving the trigeminal
more precise interpretation. If the R2 compo- nerve. They used the blink reflex to demon-
nent is also affected, the examiner usually can strate a synkinetic response in the masseter
clarify which of the nerves is involved. Abnor- muscle with stimulation of the supraorbital
mality of the R2 responses with a normal R1 nerve, thereby confirming that the masseter
response suggests a central lesion of the spinal in their patient was innervated by aberrantly
tract and nucleus of the trigeminal nerve. regenerated branches of the facial nerve.
ASSESSMENT OF FACIAL PERIPHERAL NEUROPATHY

SYNKINESIS
Delayed responses to supraorbital nerve stim-
In some instances, the location of facial record- ulation are found in hereditary motor and
ing electrodes during blink reflex studies can sensory neuropathy type I (Charcot–Marie–
be modified to allow for objective assessment Tooth disease).17 The delay is most marked
of facial synkinesis. This can be accomplished in the distal segment of the facial nerve
by recording simultaneously over the orbicu- even though no facial weakness may be
laris oculi and another muscle on the same demonstrated. In contrast, the delayed laten-
side of the face which is also supplied by the cies in Guillain–Barré syndrome are associ-
facial nerve, such as the mentalis or orbicu- ated with facial muscle weakness. The blink
laris oris. In normal subjects, stimulation of reflexes also may be abnormal in acquired
the supraorbital nerve produces reflex activa- demyelinating neuropathies, including chronic
tion of only the orbicularis oculi muscle. In inflammatory demyelinating polyradiculoneu-
hemifacial spasm and aberrant regeneration of ropathy (CIDP) (Fig. 31–8). In most axonal
the facial nerve after injury, the second muscle sensorimotor peripheral neuropathies, includ-
often demonstrates a synkinetic response15 ing diabetic neuropathy,18 the blink reflex is
(Fig. 31–7). This synkinetic response is not unaffected .
Figure 31–9. Blink reflex studies in a patient with mul-

tiple sclerosis and trigeminal neuralgia. The R1 and R2
Figure 31–8. Blink reflex studies in a patient with responses are markedly delayed. R1 max, upper limit of
chronic inflammatory demyelinating polyradiculoneuropa- normal for the onset of the R1 response; R2 max, upper
thy. The R1 and R2 responses are significantly delayed. limit of normal for the onset of the R2 response.
R1 max, upper limit of normal for the onset of the R1
response; R2 max, upper limit of normal for the onset of
the R2 response. was used in some laboratories to search for
evidence of clinically silent lesions that would
The blink reflex may be helpful in the eval-
provide further support for the diagnosis of
uation of patients who present clinically with
multiple sclerosis (Fig. 31–9). Currently, it is
subacute sensory neuronopathy. In a study of
rarely used for this purpose, but it may be
patients who were evaluated in our laboratory,
helpful in documenting objective abnormali-
the blink reflex was normal in patients with
ties in patients with vague facial paresthesias of
paraneoplastic sensory neuronopathy (malig-
uncertain significance.
nant inflammatory sensory polyganglionopa-
In 260 patients with multiple sclerosis
thy), but frequently abnormal in those with
observed over a 7-year period, Kimura20 found
Sjögren syndrome/sicca complex or idiopathic
that the R1 response was delayed on one or
sensory neuronopathy (nonmalignant inflam-
both sides in 66% of those with definite mul-
matory sensory polyganglionopathy).19
tiple sclerosis, in 56% of those with probable
multiple sclerosis, and in 29% of those with
POSTERIOR FOSSA LESIONS possible multiple sclerosis. R1 was abnormal
In patients with cerebellopontine angle lesions, in 78% of the patients with neurologic signs
the blink reflex response latencies may be that suggested pontine involvement in 57%
delayed. Previously, the blink reflex was used of those with signs of disease of the medulla
as a diagnostic test to screen for suspected or midbrain, and in 40% of those who had
tumors in the cerebellopontine angle. How- no brain stem signs. R2 was less diagnostic
ever, with the advent of sophisticated imaging than R1 in detecting brain stem lesions and
techniques, the blink reflex has no place in the was most often abnormal in those with pon-
diagnostic evaluation when these lesions are tine signs. When R1 was normal and R2 was
suspected. delayed, the patients had symptoms suggesting
medullary involvement. As might be expected
in patients with multiple sclerosis, the inci-
MULTIPLE SCLEROSIS
dence of delayed R1 responses increases with
Before the advent of modern signal averag- time, although there may be improvement
ing and imaging techniques, the blink reflex when the disease is in remission.
EXTRAPYRAMIDAL DISEASE • The blink reflex is typically normal in

trigeminal neuralgia and in cases of atypi-
In normal subjects, habituation of the R2 com- cal face pain.
ponent of the reflex occurs with regularly • The blink reflex and facial nerve conduc-
applied electric stimuli, resulting in delayed tion studies (NCS) can be used as predic-
or absent R2 responses. This is the basis for tive tools in determining the outcome of
the glabellar tap sign elicited during neurologic acute Bell’s palsy.
examination. In patients with Parkinson dis- • Facial synkinesis can be demonstrated by
ease or other disorders affecting the extrapyra- recording over the orbicularis oculi mus-
midal system, habituation may be impaired. cle and another facial nerve innervated
muscle such as the mentalis muscle while
OTHER FACTORS AFFECTING stimulating the supraorbital nerve.
BLINK REFLEX RESPONSES • In patients with cerebral infarction asso-
ciated with hemiparesis, the R1 response
Although the blink reflex pathways are con- may be delayed for up to 1 week, and
fined to the brain stem, lesions rostral to both direct and consensual R2 responses
the brain stem may affect latencies and thus may be absent or diminished for several
diminish the localizing value of the reflex. weeks.
In almost half of the patients with cerebral
infarction associated with hemiparesis, the R1
response may be delayed for up to 1 week, and
both direct and consensual R2 responses may
be absent or diminished for several weeks.21 LATERAL SPREAD OF THE
Therefore, in the acute phase, whether facial FACIAL NERVE RESPONSE:
paresis is caused by a central or a peripheral ASSESSMENT OF FACIAL
process cannot be determined solely on the SYNKINESIS AND HEMIFACIAL
basis of prolonged R1 latency. These effects
may be the result of removal of crossed cortical
SPASM
facilitation in the brain stem.22
The R2 responses are attenuated with sleep Hemifacial spasm is characterized by unilat-
eral, involuntary twitching of periorbital mus-
and with the pharmacologic influence of seda-
tive drugs. cles and muscles of the lower portion of the
face. In most patients, the disorder is proba-
bly associated with compression of the nerve
Key Points
by a vascular loop near the brain stem. Rarely,
• The trigeminal blink reflex is mediated by a tumor, aneurysm, or vascular malformation
the first division of the trigeminal nerve may compress the facial nerve. Focal demyeli-
(afferent limb) and the facial nerve (effer- nation ensues, producing ectopic generation
ent limb). of action potentials or ephaptic transmission
• In normal individuals, two responses (R1 to other axons adjacent to the site of com-
and R2) are obtained when recording pression.23 Recent evidence suggests that in
from the muscles ipsilateral to the stimu- hemifacial spasm the facial motor nucleus
lated nerve, and a single (R2) response is may become hyperactive, with spasms result-
obtained from the contralateral muscles. ing from neuronal “backfiring” that is similar to
• Abnormalities in trigeminal blink reflex an exaggerated F response.24, 25 These patients
latencies can often be localized to specific also have synkinesis, which appears similar to
segments of the peripheral and/or central that seen in aberrant regeneration, but in dis-
neural pathways traversed by trigeminal tinction the synkinetic response is variable in
and facial axons. early cases and may not be present between
• For disorders affecting the ophthalmic spasms.15
division, the blink reflex is useful for Abnormal communication between axons of
identifying lesions of the sensory nerve, the facial nerve, termed lateral spread, can be
root, or gasserian ganglion. demonstrated by stimulating the mandibular
• The blink reflex is less useful for disorders branch of the facial nerve while recording over
affecting the second or third divisions of the orbicularis oculi muscle or by stimulating
the trigeminal nerve. the zygomatic branch of the facial nerve
Figure 31–10. Electrode placement to assess the lateral

spread response in a patient with hemifacial spasm. G1,
active electrode; G2, reference electrode; Stim., stimulate.
(From Harper, C. M. Jr. 1991. AAEM case report 21: Hemi-
facial spasm: Preoperative diagnosis and intra-operative
management. Muscle & Nerve 14:213–18. By permission of
Mayo Foundation for Medical Education and Research.)
Figure 31–11. Lateral spread response in a patient with hemifacial spasm. With stimulation of the mandibular branch of
the facial nerve (top), a delayed response is recorded from the ipsilateral orbicularis oculi. With stimulation of the zygomatic
branch of the facial nerve (bottom), a delayed response is recorded from the ipsilateral mentalis.
while recording over the mentalis muscle Key Points

(Figs. 31–10 and 31–11). Surgical microvas-
• The lateral spread response can provide
cular decompression of the facial nerve gen-
electrophysiological evidence of synki-
erally results in loss of lateral spread and
synkinesis and relief of clinical hemifacial netic activity in patients with hemifacial
spasm.26, 27 spasm.
JAW JERK (MASSETER REFLEX)
Neuroanatomy
The jaw jerk, or masseter reflex, is a monosy-
naptic muscle stretch reflex elicited by a tap on
the jaw. Afferent impulses from muscle spin-
dles in the masseter muscle are conveyed via
the motor root of the trigeminal nerve to the
mesencephalic nucleus in the midbrain. Axons
from this nucleus synapse in the motor nucleus
of the trigeminal nerve to activate the effer-
ent limb of the reflex arc. This reflex is unique
among stretch reflexes in that the cell bodies
of the afferent limb (i.e., the mesencephalic
nucleus) lie intra-axially in the brain stem Figure 31–13. Superimposed responses recorded from
rather than in the gasserian ganglion, which is the masseter muscles following four successive taps on
the brain stem counterpart of a spinal dorsal the chin with a reflex hammer in a normal subject. The
reflex hammer contains a microswitch that, upon contact,
root ganglion. The afferent nerve cell bod- initiates a sweep across the monitor.
ies subserving all other stretch reflexes reside
extra-axially in dorsal root ganglia.
Wide variation in amplitude among normal
subjects precludes the use of amplitude mea-
Methods surements in clinical studies. In normal sub-
jects, particularly elderly and obese ones, the
Recording electrodes are taped over the belly reflex is sometimes difficult to record using
of the masseter muscle bilaterally, and the ref- surface electrodes.
erence electrodes are placed over the zygoma
(Fig. 31–12). The reflex hammer contains a
microswitch that triggers a sweep across the Applications
monitor upon contact with the examiner’s fin-
ger, which is held on the patient’s chin. The The main indication for using the jaw jerk
latency is measured to the initial reproducible is to assess the function of the mandibular
deflection from baseline (Fig. 31–13). The nor- division of the trigeminal nerve. If a patient’s
mal range of latencies is 6–10.5 ms. The maxi- symptoms are in the distribution of this divi-
mal side-to-side difference in latency is 1.5 ms. sion, the blink reflex may well be normal. In
this situation, the jaw jerk may provide objec-
tive evidence of involvement of the mandibu-
lar division (Fig. 31–14). The most common
abnormality is the absence of the jaw jerk
rather than prolongation of its latency.
In patients with inflammatory polygan-
glionopathies presenting with pure sensory
neuronopathy, the jaw jerk may be normal even
though the patient is otherwise areflexic, has
no sensory responses in the limbs, and may not
have a blink reflex.28 This likely occurs because
the afferent nerve cell body involved with the
jaw jerk is in the mesencephalic nucleus of the
Figure 31–12. Electrode placement for study of the jaw trigeminal nerve, which is in the brain stem and
jerk (masseter reflex). (From Auger, A. G. 1987. Brain stem
disorders and cranial neuropathies. In Clinical electromyo-
protected by the blood–brain barrier. Although
graphy, ed. W. F. Brown, and C. F. Bolton, 417–29. Boston: neurons in the dorsal root ganglia and gasse-
Butterworths. By permission of the publisher.) rian ganglion are protected by the blood–nerve
MASSETER INHIBITORY REFLEX

The masseter inhibitory reflex is important in
the reflex control of biting and chewing. Its
primary purpose is to protect intraoral struc-
tures against the powerful jaw-closing muscles.
It is elicited by applying a mechanical or elec-
tric stimulus to the skin and mucous membrane
supplied by the maxillary or mandibular divi-
sion of the trigeminal nerve during voluntary
contraction of the masseter muscle.
Methods
The recording electrodes are placed on the
Figure 31–14. Superimposed responses recorded from masseter muscle bilaterally, in the same man-
the masseter muscles following four successive taps on the
chin with a reflex hammer in a patient with a left acoustic
ner as in the jaw jerk (Fig. 31–12). A mechani-
neuroma. The response is normal on the right and absent cal tap usually produces only one silent period
on the left. (SP), which begins between 11 and 15 ms after
the tap and lasts for 14–30 ms (Fig. 31–15).
With an electric stimulus, two silent periods
barrier, this is not as protective as the blood– typically occur: the first one (SP1) corresponds
brain barrier. to the SP after a mechanical tap and the second
Ongerboer De Visser and Goor29 studied
jaw jerk and masseter electromyograms in
patients with vascular or neoplastic disease of
the midbrain and pons. Mesencephalic lesions
were associated with abnormal jaw reflexes
and normal masseter electromyograms. In the
group with pontine lesions, both the mas-
seter electromyograms and the jaw jerk were
often abnormal. An abnormal jaw jerk sug-
gests midbrain disease, whereas an abnormal
R1 response, with or without an associated
change in the jaw jerk, suggests a rostral pon-
tine lesion.30, 31 The latency of the jaw jerk is
not influenced by supratentorial or primary
cerebellar disease.32
Key Points
• The jaw jerk (masseter reflex) is medi-
ated by the third division of the trigeminal
nerve.
• The masseter reflex is unique among
stretch reflexes in that the cell bodies of
the afferent limb (in the mesencephalic
nucleus) lie intra-axially in the brain stem
rather than in the gasserian ganglion.
• Even though the jaw jerk can help to Figure 31–15. Superimposed responses recorded from
the masseter muscles following four successive taps on the
identify disease in the pons or midbrain, chin (top) and four successive electric stimuli to the mental
it has largely been supplanted by high- nerve (bottom) while the subject clenched this teeth. SP1,
resolution neuroimaging techniques. first silent period; SP2, second silent period.
one (SP2) begins 30–60 ms after the stimulus The MIR may be abolished in some forms
(Fig. 31–15). of sensory neuropathy involving predominantly
intraoral sensory nerves, giving rise to severe
impairment in chewing and swallowing.35
Applications In some laboratories, the MIR is used to
assess function of the maxillary and mandibu-
The MIR is sometimes useful in the eval- lar divisions of the trigeminal nerve by apply-
uation of peripheral neuropathies.33, 34 In ing an electric stimulus to the infraorbital
severe demyelinating neuropathies in which and mental nerves, respectively, during volun-
no response occurs to stimuli in the limbs, tary contraction of the masseter. Abnormalities
the MIR can be used to assess conduction of conduction can be detected by comparing
delay, because it can still be measured in neu- the latency of the onset of SP1 on the two
ropathies that are severe enough to abolish sides.36
even the blink reflex (Fig. 31–16). In these sit- The MIR also has been used to assess cen-
uations, the latency of the onset of the MIR tral inhibition in some disease states. In trismus
may be severely prolonged, thereby providing associated with tetanus, the MIR may be abol-
evidence for a demyelinating component. The ished. In the rare condition of hemimasticatory
reflex is normal in axonal neuropathies. spasm, the MIR is attenuated on the side of the
Figure 31–16. MIR following a tap on the chin in a patient with CIDP compared with that in normal subject. The onset
of the first silent period (SP1) is delayed in the patient with CIDP.
spasm, implying impaired inhibition during the

period of spasm.37, 38
Key Points
• The MIR can be useful in demyelinating
neuropathies when limb responses cannot
be obtained.
• In situations which produce impaired cen-
tral inhibition, such as tetanus, the MIR
can be abolished.
GREAT AURICULAR SENSORY

NERVE CONDUCTION STUDIES
Nerve conduction studies (NCS) abound in
clinical neurophysiology practice for assess-
ing predominantly distal limb sensory nerves.
In the lower extremity, the sural, plantar,
and superficial fibular (peroneal) nerves may
be studied, while in the upper extremity,
the median, ulnar, radial, and antebrachial Figure 31–17. Greater auricular sensory nerve conduc-
cutaneous nerves may be interrogated. Sen- tion study technique. (“Reprinted with the permission of
sory NCS in proximal segments are limited to The Cleveland Clinic Center for Medical Art & Photogra-
phy © 2009. All Rights Reserved”.)
trigeminal “blink” reflexes and a few difficult-
to-perform and less-than-reliable techniques
to study such nerves as the saphenous, lat-
eral femoral cutaneous, and posterior femoral Key Points
cutaneous sensory nerves.
• The great auricular sensory response is
useful in sensory neuropathies when limb
Methods sensory responses cannot be obtained.
• For focal disorders affecting sensation in
The recording electrode is placed directly the C2,3 dermatomes, the great auricu-
behind the earlobe, while the reference elec- lar sensory response can provide evidence
trode is placed 2 cm away posterior to the of the integrity of postganglionic sensory
helix. The stimulating electrode is placed on pathways.
the posterior border of the sternocleidomas-
toid muscle 8 cm inferior to the recording elec-
trode (Fig. 31–17). Amplitude values in normal TRIGEMINAL CONTACT HEAT
subjects range from 8–48 μV, while the distal EVOKED POTENTIAL
latency falls between 1.4 and 2.4 ms.39, 40 STIMULATOR STUDIES
The Contact Heat Evoked Potential Stimula-
Applications tor (CHEPS) method utilizes rapidly delivered
pulses of heat to stimulate the differen-
The great auricular sensory nerve (C2,3) is tial warm or heat thresholds of receptors
a readily investigated proximal spinal sensory innervated by cutaneous Aδ and C fibers. The
nerve that can be learned quickly, performed resulting evoked potentials can be recorded
reliably, and studied in individuals suspected of and measured over the scalp with standard sur-
having peripheral processes affecting proximal face EEG electrodes. CHEPS has been used to
segments, ventral rami of the cervical plexus, excite preferentially cutaneous Aδ and C fibers
or local disorders affecting the great auricular in human volunteers and correlate with sub-
nerve in isolation.41 jectively reported scores of pain intensity.42–44
A marked similarity between CHEPS and laser The great auricular sensory NCS is a use-
evoked potentials has been noted;45 CHEPS ful method to assess proximal somatic sensory
offers the advantages of being easier to admin- function in the upper cervical dermatomes.
ister, not requiring eye protection, and reduc- The CHEPS technology provides a method
ing risk of causing burns to the skin. The to study somatic small fiber sensory pathways
skin of the face can be stimulated in the from the trigeminal dermatomes to the sensory
first, second, or third division of the trigemi- cortex. Although not discussed in this chap-
nal nerve, eliciting responses over the midline ter, needle electrode examination of muscles
scalp region between 5 and 30 μV in amplitude innervated by the trigeminal and facial cranial
and between 300 and 500 ms in latency. nerves are usually performed in combination
In summary, CHEPS method offers an addi- with cranial nerve reflex studies.
tional clinical tool for the assessment of small
diameter somatic sensory nerve fiber func-
tion from the trigeminal territory (and other
regions), complementing the assessment of REFERENCES
autonomic small diameter fiber pathways by
autonomic testing methods.46 1. Overend, W. 1896. Preliminary note on a new cranial
reflex. Lancet 1:619.
2. Kugelberg, E. 1952. Facial reflexes. Brain 75:385–96.
3. Rushworth, G. 1962. Observations on blink reflexes.
Key Points Journal of Neurology, Neurosurgery, and Psychiatry
25:93–108.
• CHEPS studies assess somatic spinothala- 4. Shahani, B. 1970. The human blink reflex. Journal of
mic sensory function mediated by Aδ and Neurology, Neurosurgery, and Psychiatry 33:792–800.
C fibers. 5. Trontelj, M. A., and J. V. Trontelj. 1978. Reflex arc
• Each of the three divisions of the trigem- of the first component of the human blink reflex:
A single motoneurone study. Journal of Neurology,
inal sensory territory can be investigated Neurosurgery, and Psychiatry 41:538–47.
using CHEPS, stimulating the skin on the 6. Stoehr, M., and F. Petruch. 1978. The orbicularis
face and recording over the parietal scalp oculi reflex: Diagnostic significance of the reflex ampli-
using standard EEG electrodes. tude. Electromyography and Clinical Neurophysiol-
ogy 18:217–24.
7. Anday, E. K., M. E. Cohen, and H. S. Hoffman.
1990. The blink reflex: Maturation and modification
in the neonate. Developmental Medicine and Child
SUMMARY Neurology 32:142–50.
8. Jaaskelainen, S. K. 1995. Blink reflex with stimu-
Although a certain level of expertise is neces- lation of the mental nerve. Methodology, reference
values, and some clinical vignettes. Acta Neurologica
sary, electrophysiologic study of cranial reflexes Scandinavica 91:477–82.
is not technically demanding, time-consuming, 9. Hagen, N. A., J. C. Stevens, and C. J. Michet Jr.
or associated with substantial patient discom- 1990. Trigeminal sensory neuropathy associated with
fort. The information obtained may document connective tissue diseases. Neurology 40:891–6.
objective abnormality and assist with localiza- 10. Kimura, J., R. L. Rodnitzky, and M. W. Van Allen.
1970. Electrodiagnostic study of trigeminal nerve.
tion. The blink reflexes are useful for studying Orbicularis oculi reflex and masseter reflex in trigem-
the function of the trigeminal and facial nerves inal neuralgia, paratrigeminal syndrome, and other
and their central connections in the brain stem. lesions of the trigeminal nerve. Neurology 20:574–83.
When NCS in the limbs suggest a demyeli- 11. Feldman, R. G., C. Niles, S. P. Proctor, and
J. Jabre. 1992. Blink reflex measurement of effects
nating peripheral neuropathy, the blink reflex of trichloroethylene exposure on the trigeminal nerve.
can provide information about involvement of Muscle & Nerve 15:490–5.
proximal nerve segments. Patterns of involve- 12. Kimura, J., L. T. Giron, and S. M. Young. 1976.
ment of the facial and trigeminal nerves are Electrophysiological study of Bell palsy: Electri-
often helpful in suggesting the type of neu- cally elicited blink reflex in assessment of prognosis.
Archives of Otolaryngology 102:140–3.
ropathy under investigation. The jaw jerk is 13. Ghonim, M. R., and C. Gavilan. 1990. Blink reflex:
useful in assessing the mandibular division of Prognostic value in acute peripheral facial palsy. ORL;
the trigeminal nerve, and it can aid in eval- Journal for Oto-Rhino-Laryngology and its Related
uating patients with suspected sensory gan- Specialities 52:75–9.
14. Heath, J. P., R. E. Cull, I. M. Smith, and J. A. Murray.
glionopathies. The MIR is sometimes helpful 1988. The neurophysiological investigation of Bell’s
in evaluating patients with demyelinating neu- palsy and the predictive value of the blink reflex.
ropathies and in assessing central inhibition. Clinical Otolaryngology 13:85–92.
15. Auger, R. G. 1979. Hemifacial spasm: Clini- 32. Hopf, H. C., C. Hinrichs, P. Stoeter, P. P. Urban,
cal and electrophysiologic observations. Neurology J. Marx, and F. Thomke. 2000. Masseter reflex laten-
29:1261–72. cies and amplitudes are not influenced by supraten-
16. Rubin, D. I., J. Y. Matsumoto, G. A. Suarez, and R. G. torial and cerebellar lesions. Muscle & Nerve 23:
Auger. 1999. Facial trigeminal synkinesis associated 86–9.
with a trigeminal schwannoma. Neurology 53:635–7. 33. Auger, R. G. 1996. Latency of onset of the masseter
17. Kimura, J. 1971. An evaluation of the facial and inhibitory reflex in peripheral neuropathies. Muscle &
trigeminal nerves in polyneuropathy: Electrodiagnos- Nerve 19:910–11.
tic study in Charcot–Marie–Tooth disease, Guillain– 34. Cruccu, G., R. Agostino, M. Inghilleri, P. Innocenti,
Barré syndrome, and diabetic neuropathy. Neurology A. Romaniello, and M. Manfredi. 1998. Mandibu-
21:745–52. lar nerve involvement in diabetic polyneuropathy and
18. Kirk, V. H. Jr., W. J. Litchy, J. L. Karnes, and P. J. chronic inflammatory demyelinating polyneuropathy.
Dyck. 1991. Measurement of blink reflexes not useful Muscle & Nerve 21:1673–9.
in detection or characterization of diabetic polyneu- 35. Auger, R. G., and P. G. McManis. 1990. Trigemi-
ropathy (abstract). Muscle & Nerve 14:910–11. nal sensory neuropathy associated with decreased oral
19. Auger, R. G., A. J. Windebank, C. F. Lucchinetti, sensation and impairment of the masseter inhibitory
and C. H. Chalk. 1999. Role of the blink reflex in reflex. Neurology 40:759–63.
the evaluation of sensory neuronopathy. Neurology 36. Ongerboer de Visser, B. W., and G. Cruccu. 1993.
53:407–8. Neurophysiologic examination of the trigeminal, facial,
20. Kimura, J. 1975. Electrically elicited blink reflex in hypoglossal, and spinal accessory nerves in cranial neu-
diagnosis of multiple sclerosis. Review of 260 patients ropathies and brain stem disorders. In Clinical elec-
over a seven-year period. Brain 98:413–26. tromyography, ed. W. F. Brown and C. F. Bolton, 2nd
21. Berardelli, A., N. Accornero, G. Cruccu, F. Fabiano, ed., 61–92. Boston: Butterworth-Heinemann.
V. Guerrisi, and M. Manfredi. 1983. The orbicularis 37. Auger, R. G., W. J. Litchy, T. L. Cascino, and J. E.
oculi response after hemispheral damage. Journal of Ahlskog. 1992. Hemimasticatory spasm: Clinical and
Neurology, Neurosurgery, and Psychiatry 46:837–43. electrophysiologic observations. Neurology 42:2263–6.
22. Kimura, J., J. T. Wilkinson, H. Damasio, H. R. Adams 38. Kim, H. J., B. S. Jeon, and K. W. Lee. 2000. Hemi-
Jr., E. Shivapour, and T. Yamada. 1985. Blink reflex masticatory spasm associated with localized sclero-
in patients with hemispheric cerebrovascular accident derma and facial hemiatrophy. Archives of Neurology
(CVA). Blink reflex in CVA. Journal of the Neurological 57:576–80.
Sciences 67:15–28. 39. Smith, B. E., and P. J. B. Dyck. 2004. Proximal
23. Nielsen, V. K. 1984. Pathophysiology of hemofacial spinal neurophysiological assessments: Greater auric-
spasm: I. Ephaptic transmission and ectopic excitation. ular nerve conduction study (abstract 284). Annals of
Neurology 34:418–26. Neurology 56 (Suppl 8):S62.
24. Moller, A. R. 1991. Interaction between the blink 40. Peuker, E. T., and T. J. Filler. 2002. The nerve supply
reflex and the abnormal muscle response in patients of the human auricle. Clinical Anatomy 15:35–7.
with hemifacial spasm: Results of intraoperative 41. Kimura, I., H. Seki, S. Sasao, and D. R. Ayyar.
recordings. Journal of the Neurological Sciences 1987. The great auricular nerve conduction study:
101:114–23. A technique, normative data and clinical useful-
25. Valls-Sole, J., and E. S. Tolosa. 1989. Blink reflex ness. Electromyography and Clinical Neurophysiology
excitability cycle in hemifacial spasm. Neurology 27:39–43.
39:1061–6. 42. Granovsky, Y., D. Matre, A. Sokolik, J. Lorenz, and
26. Auger, R. G., D. G. Piepgras, E. R. Laws Jr., and K. L. Casey. 2005.Thermoreceptive innervation of
R. H. Miller. 1981. Microvascular decompression human glabrous and hairy skin: A contact heat evoked
of the facial nerve for hemifacial spasm: Clinical potential analysis. Pain 115(3):238–47.
and electrophysiologic observations. Neurology 31: 43. Chen, A. C., D. M. Niddam, and L. Arendt-Nielsen.
346–50. 2001. Contact heat evoked potentials as a valid means
27. Moller, A. R., and P. J. Jannetta. 1987. Monitoring to study nociceptive pathways in human subjects. Neu-
facial EMG responses during microvascular decom- roscience Letters 316(2):79–82.
pression operations for hemifacial spasm. Journal of 44. Le Pera, D., M. Valeriani, D. Niddam, A. C. Chen, and
Neurosurgery 66:681–5. L. Arendt-Nielsen. 2002. Contact heat evoked poten-
28. Auger, R. G. 1998. Role of the masseter reflex in the tials to painful and non-painful stimuli: Effect of atten-
assessment of subacute sensory neuropathy. Muscle & tion towards stimulus properties. Brain Topography
Nerve 21:800–1. 15(2):115–23.
29. Ongerboer De Visser, B. W., and C. Goor. 1976. 45. Valeriani, M., D. Le Pera, D. Niddam, A. C. Chen,
Jaw reflexes and masseter electromyograms in mes- and L. Arendt-Nielsen. 2002. Dipolar modelling of
encephalic and pontine lesions: An electrodiagnostic the scalp evoked potentials to painful contact heat
study. Journal of Neurology, Neurosurgery, and Psy- stimulation of the human skin. Neuroscience Letters
chiatry 39:90–2. 318(1):44–8.
30. Hopf, H. C., F. Thomke, and L. Gutmann. 1991. 46. Atherton, D. D., P. Facer, K. M. Roberts, et al. 2007.
Midbrain vs. pontine medial longitudinal fasciculus Use of the novel contact heat evoked potential stim-
lesions: The utilization of masseter and blink reflexes. ulator (CHEPS) for the assessment of small fibre
Muscle & Nerve 14:326–30. neuropathy: Correlations with skin flare responses and
31. Thomke, F. 1999. Isolated cranial nerve palsies due to intra-epidermal nerve fibre counts. BMC Neurology
brain stem lesions. Muscle & Nerve 22:1168–76. 7:21–30.
Chapter 32
Long Latency Reflexes and the

Silent Period
John N. Caviness
INTRODUCTION The Flexor Reflex

LONG LATENCY REFLEXES
THE SILENT PERIOD
LLRs to Stretch
Cutaneous Silent Period
LLRs to Mixed Nerve Stimulation
Cutaneous Reflexes SUMMARY
INTRODUCTION LONG LATENCY REFLEXES

Long latency reflexes (LLRs) and the silent There is no exact moment when the reflex
period are normal phenomena recorded from response to a stimulus ends and the voluntary
muscles that are highly dependent on the cen- reaction begins. Rather, from the onset of a
tral nervous system connections. The clinical monosynaptic stretch reflex to the time of the
application of these phenomena is limited, but first conscious voluntary reaction, the corti-
they can be helpful in selected situations. Some cal influence over the spinal and brain stem
LLRs are abnormal in patients with myoclonus, reflex activity gradually increases. Unique elec-
and the silent period shows abnormalities with tromyographic (EMG) phenomena called long
hyperexcitable motor neurons or peripheral latency reflexes arise during this transition
nerves. period. Like spinal reflexes, LLRs have pre-
dictable latencies, but their amplitudes are
modulated profoundly by context and volition.
Purpose and Role of LLRs and the Silent
These characteristics make them important
Period
tools in the study of normal and abnormal
• LLRs are useful in certain central nervous motor control.1
system disorders. Hammond2 discovered the LLRs while
• The silent period is abnormal in disorders studying the EMG responses evoked in the
with hyperexcitable motor neurons. biceps muscle by sudden stretch. EMG activity
543
appeared at 70 ms, later than the monosynaptic it is unlikely that all LLRs have an identical
stretch reflex and earlier than the volun- physiologic basis.11
tary reaction time of 113 ms. A command LLRs are altered in disorders anywhere
to “resist” the stretch augmented these long along their pathway from the periphery to
latency responses and the instruction to “let go” the cortex. They may be absent, delayed, or
resulted in their virtual disappearance. Mars- enhanced.12
den et al.3 studied similar stretch reflexes in
the long flexor of the thumb during movement • Absent LLR II—lesions of the lemnis-
and theorized that LLRs served to reinforce cal pathways, cortex, corticospinal tracts,13
volition or intent against unexpected perturba- and Huntington’s disease.
tions. The reflexes were thought to emanate • Delayed LLR II—demyelinating lesions
from a transcortical loop, with one arm of in multiple sclerosis and Friedreich’s dis-
the loop ascending to the sensorimotor cortex ease.14
in the dorsal column-medial lemniscus system • Enhanced LLR I—cortical and subcor-
and the other arm descending in the corti- tical myoclonus,15 corticobasal degenera-
cospinal tract. tion, Parkinson’s disease, essential tremor,
Several lines of evidence support the con- and dystonia.16
cept of a transcortical reflex loop. They are as
follows: Key Points
• LLRs are delayed or absent in patients • LLRs arise during this transition period
with lesions of the dorsal columns or sen- from the onset of a monosynaptic stretch
sorimotor cortex. reflex to the time of the first conscious
• Cortical potentials precede LLRs by voluntary reaction.
30–50 ms and the two events correlate in • The reflexes are thought to emanate from
amplitude.4 a transcortical loop, with one arm of the
• LLRs occur bilaterally and nearly simul- loop ascending to the sensorimotor cor-
taneously in response to a unilateral stim- tex in the dorsal column-medial lemniscus
ulus in patients with congenital mirror system and the other arm descending in
movements.5 the corticospinal tract.
• Patients with cortical reflex myoclonus
have hyperexcitable LLRs, which clearly
are cortically mediated.6 LLRs to Stretch
• Modulation was also reflected in the
amplitude of the sensorimotor cortex All protocols for stretch reflex testing involve
potentials just preceding the LLR.7 a computer-controlled torque motor that can
be programmed to maintain a steady load or
However, the persistence of LLRs in spinal to introduce rapid perturbations. Generally,
animals forces one to consider other possi- the torque is delivered through a manipu-
ble explanations. For example, repetitive firing landum that the subject holds. The subject
of muscle spindles or transmission of sensory receives visual feedback about the position of
influences by slowly conducting fibers could the manipulandum and attempts to hold it
explain the appearance of reflex activity at long stationary against a low constant torque. The
latencies. It is likely that the neural circuits that computer delivers random torque pulses, and
generate LLRs depend on the type of stimu- the surface EMG signals are recorded over
lus.8, 9 Although the physiologic basis of LLRs the agonist and antagonist of the joint that
is a matter of controversy, the weight of evi- is stretched. The EMG signal is rectified and
dence strongly suggests that “loops” involving averaged.
the motor cortex are involved in their genera- In a normal response, an M1 component
tion.10 occurs at 30 ms and corresponds to the
LLRs have been recorded by various tech- monosynaptic stretch reflex. The M2 compo-
niques using stretch or different forms of elec- nent, the most frequent long latency compo-
tric stimulation. The precise character of the nent, appears at 55–65 ms in the wrist. Occa-
reflexes depends on the testing protocol, and sionally, a later M3 component may be seen
Long Latency Reflexes and the Silent Period 545
individually or it may merge with the M2 com-

ponent.
In Parkinson’s disease, the M2 component is
enlarged in the wrist flexors. This abnormality
corresponds to the degree of the patient’s rigid-
ity. In contrast, a decrease in or absence of the
M2 component distinguishes Huntington’s dis-
ease.17 M2 may be prolonged in dystonia. As
noted above, lesions of the dorsal columns ipsi-
lateral to the tested limbs or the contralateral
sensorimotor cortex may delay or abolish M2.
Key Points Figure 32–1. Abnormal presence of LLR II in a patient

with cortical reflex myoclonus.
• In a normal LLR response to stretch, an
M1 component occurs at 30 ms and corre-
sponds to the monosynaptic stretch reflex.
• The M2 component, the most fre- In myoclonic disorders, a late response
that corresponds to a reflex myoclonic jerk
quent long latency component, appears at
is recorded at 40–60 ms in the hand at rest
55–65 ms in the wrist.
• A later M3 component may be seen indi- (Fig. 32–1). Sutton and Mayer19 named this
the C reflex, a distinct abnormality that may
vidually or it may merge with the M2
be seen in cortical reflex myoclonus, reticular
component.
reflex myoclonus, hyperekplexia, Alzheimer’s
disease, and other symptomatic myoclonic dis-
orders. The pattern of an increased LLR I
LLRs to Mixed Nerve Stimulation and LLR III with a normal LLR II may typ-
ify Parkinson’s disease. The LLR II is absent in
Upton et al.18 discovered a late response, Huntington’s disease but is unaffected in other
termed V2, to the electric stimulation of a choreatic movement disorders.17, 20 Patients
mixed nerve. Since that time, a series of LLRs with focal dystonia may display increased LLR
in the hand muscles has been distinguished in I or reduced LLR II.21 Possibly in a sub-
response to median nerve stimulation. group of patients with essential tremor, the
The median nerve is stimulated at the wrist, LLR I is increased.22 A delayed LLR II reflects
with the cathode proximal. Surface EMG elec- slowing of central conduction in patients with
trodes are placed over the abductor pollicis multiple sclerosis.23 LLR II disappears after
brevis. The stimulus duration is set to 1 ms, and thalamic infarction, corresponding to the clini-
the intensity is increased to the level that pro- cal deficit and abnormalities on somatosensory
duces the first small twitch. Initially, recordings evoked potential testing.24 Deuschl provides
are made with the muscle at rest and single useful normal values for LLRs of the hand
shocks are given. Next, the patient maintains a muscles to median nerve stimulation.12
moderate contraction, and stimuli are given at
a rate of 1–3 Hz. The signal must be rectified
Key Points
and then averaged. A total of 100–500 stimuli
provide reproducible results. • LLRs in the hand muscles can occur in
In normal subjects with their hands at rest, response to median nerve stimulation.
no late response is seen after the F wave and • In normal subjects with their hands at rest,
before the voluntary reaction time. With con- no late response is seen after the F wave
traction and averaging, a short latency reflex and before the voluntary reaction time.
develops at approximately 28 ms and is identi- With contraction and averaging, a short
cal with the H reflex. A late response named latency reflex develops at approximately
the LLR II appears in all normal subjects at 28 ms and is identical with the H reflex.
50 ms, and in one-third of them an additional • A late response named the LLR II appears
LLR I may be recorded at 40 ms or an LLR III in all normal subjects at 50 ms, and in
at 75 ms. one-third of them an additional LLR I may
be recorded at 40 ms or an LLR III at The Flexor Reflex

75 ms.
• In myoclonic disorders, a late response, Certain stimuli may trigger reflex withdrawal,
called the C reflex, which corresponds flexor reflexes. Generally, the adequate stim-
to a reflex myoclonic jerk is recorded at ulus must be cutaneous and noxious. Such
40–60 ms in the hand at rest. stimuli affect a polysynaptic network of spinal
neurons, termed flexor reflex afferents, that
program patterned withdrawal behavior.
Meinck et al.30 standardized a technique
Cutaneous Reflexes for eliciting flexor reflexes. The medial plan-
tar nerve is stimulated in the ball of the foot
LLRs evoked by purely cutaneous nerve stim- with a train of five shocks by using a stimulus
ulation demonstrate well-defined inhibitory as duration at 0.1 ms and an interstimulus inter-
well as excitatory periods.25 Stimuli are deliv- val of 3–5 ms. The stimulus intensity is adjusted
ered to the index finger by ring electrodes. The to the motor threshold of the flexor hallucis
surface EMG is recorded with an electrode brevis muscle, a level that will be perceived
over the belly of the first dorsal interosseous as mildly painful. EMG is recorded from the
and a reference electrode over the radial sty- anterior tibial muscle, rectified, and then aver-
loid process. The patient maintains a steady aged. A total of eight trains are averaged, with
force at 20% of maximum. Stimulus inten- the stimuli repeated every 1–3 seconds. With
sity is adjusted to four times the sensory this protocol, the normal activity is triphasic,
threshold, with a stimulus duration of 0.2 ms. with a large F1 response at 70 ms, a period of
With a stimulus rate of 3 Hz, 100–500 sam- silence, and then a small F2 burst at approxi-
ples are collected, full-wave rectified, and then mately 150 ms.
averaged. With spinal cord lesions, exaggerated with-
A first excitatory period, E1, is present at drawal corresponds to a large response at or
approximately 40 ms, followed by inhibition, beyond the latency of F2. Very high-intensity
I1, at approximately 51 ms (Fig. 32–2). A final stimulation may shorten the latency of this
excitatory wave, E2, appears at approximately response to a value consistent with F1. In spas-
66 ms.26, 27 I1 and E2 depend on the motor ticity caused by a hemispheric lesion, stimula-
cortex and are absent with contralateral cor- tion may trigger alternating clonic bursts in the
tical lesions.26 E2 may be delayed in multi- anterior tibial and gastrocnemius muscles.
ple sclerosis. The depth of I1 is decreased
in patients with Parkinson’s disease as com- Key Points
pared with control subjects.28 E2 is increased
• Withdrawal or flexor reflexes may be
in myoclonus associated with akinetic rigid
disorders, and the latency of this compo- recorded following electrical stimulation.
• The normal flexor reflex activity is tripha-
nent may separate Parkinson’s disease and
multiple systems atrophy from corticobasal sic, with a large F1 response at 70 ms,
degeneration.29 a period of silence, and then a small F2
burst at approximately 150 ms.
• With spinal cord lesions, exaggerated
withdrawal corresponds to a large response
at or beyond the latency of F2.
THE SILENT PERIOD

There are two general categories of silent
period. The most familiar and most thoroughly
Figure 32–2. The cutaneous reflex. The index finger was studied is elicited by a strong shock to a nerve
stimulated electrically, and EMG activity was recorded
from the flexor digitorum indicis. Two phases of excitation,
innervating a contracting muscle. A second,
E1 and E2, and an intervening phase of inhibition, I1, are less well-known form occurs in response to a
seen in the record. cutaneous nerve not innervating the region of
the contracting muscles and is considered a 60 ms, may reflect activation of recurrent col-
protective reflex mediated by spinal inhibitory laterals of Renshaw cells. These two intervals
circuits.31 combined are referred to as the peripheral
If a strong shock is delivered to the nerve of a silent period. The final period of silence should
muscle that is tonically contracting, a period of be viewed as a long latency inhibitory reflex
relative or absolute silence begins immediately often referred to as the cortical silent period.
and persists for about 100 ms (Fig. 32–3). The Recent evidence, including the study of the
depth of the silent period depends entirely on silent periods after cortical magnetic stimula-
the intensity of the shock. With supramaximal tion, raises the possibility of spinal inhibition
shocks, which are commonly used, the silence of corticospinal inputs or of cortically mediated
is generally complete except for an intervening inhibitory reflexes.34, 35
F wave. With lower stimulation intensities, the Few normative data exist about the depth
LLRs I–III described above appear. and duration of the silent period; thus, this
Initially, Merton32 thought the silent period period is interpreted in an “all-or-nothing”
resulted from the muscle twitch and the fashion. In states of hyperexcitability of the
unloading of muscle spindles induced by the distal nerve or muscle, the silent period may
shock. This hypothesis became untenable with be absent because ectopic impulses arise dis-
the demonstration that the silent period per- tal to the stimulus. In tetanus, the silent period
sists with the stimulation of a cutaneous nerve may be abbreviated or absent.36 A shortened
or a nonhomologous nerve or with stimula- silent period or its absence has been reported
tion proximal to a nerve block—all condi- in the case of a cervical cord tumor that pro-
tions in which twitch is absent.33 The silent duced arm rigidity.37 Prolonged duration of the
period should be viewed as a multifactorial silent period has been reported in dystonia and
phenomenon. With supramaximal stimulation, Parkinson’s disease.38 The silent period per-
approximately the first 30 ms of silence results sists in patients with pure sensory neuropathy
from the collision of impulses in the nerve and absence of sensory nerve action poten-
trunk. The next period, up to approximately tials, raising the possibility that it provides an
Figure 32–3. The silent period. After a supramaximal shock, the electromyographic activity is inhibited. A, Peripheral
silent period. B, Cortical silent period. Recorded signals above and averaged below. The silence is interrupted by an M
wave and H reflex.
electrophysiologic assessment of the integrity SUMMARY

of smaller, slower conducting sensory fibers.39
Recent studies have suggested that the corti- LLRs and the silent period are EMG phenom-
cal silent period may be useful in identifying ena that reflect the complex interplay of spinal,
cortical involvement in suspected amyotrophic brain stem, and cortical influences in motor
lateral sclerosis.40–42 A number of other recent control. These techniques have been applied to
intriguing studies suggest possible clinical the study of disorders of motor control such as
value of the cortical silent period in other Parkinson’s disease, Huntington’s disease, and
upper motor neuron disorders, including con- dystonia. Abnormalities of these reflexes may
cussion,43 hereditary spastic paraplegia44 and help to detect lesions of the central nervous
focal epilepsy,45 and restless leg syndrome.46 system.
Cutaneous Silent Period

REFERENCES
Strong shocks to cutaneous nerves in the arm
or leg and to the trigeminal nerve inhibit ongo- 1. Johnson, M. T., A. N. Kipnis, M. C. Lee, and T. J.
ing muscle contraction in nearby muscles.31 Ebner. 1993. Independent control of reflex and voli-
Latencies vary with the location of the stimu- tional EMG modulation during sinusoidal pursuit
tracking in humans. Experimental Brain Research
lus, but are generally in the range of 50–80 ms, 96:347–62.
and last approximately 50 ms depending on 2. Hammond, P. H. 1956. The influence of prior instruc-
the strength of contraction. They appear to tion to the subject on an apparently involuntary neuro-
be mediated by Aδ, small myelinated fibers. muscular response (abstract). Journal of Physiology
Cutaneous silent period (CSP) does not habit- (London) 132:17P–18P.
3. Marsden, C. D., P. A. Merton, and H. B. Morton.
uate and may be followed by post-inhibitory 1976. Servo action in the human thumb. Journal of
facilitation for 50–100 ms. CSP clinical utility Physiology (London) 257:1–44.
has not been reproducibly reported. They have 4. Goodin, D. S., M. J. Aminoff, and P. Y. Shih. 1990.
not been consistently abnormal in peripheral Evidence that the long-latency stretch responses of the
human wrist extensor muscle involve a transcerebral
neuropathies, but may be useful in identifying pathway. Brain 113:1075–91.
intact sensory roots after plexus trauma. They 5. Capaday, C., R. Forget, R. Fraser, and Y. Lamarre.
are lost in syringomyelia. There has been one 1991. Evidence for a contribution of the motor cortex
report of CSP abnormality in the restless leg to the long-latency stretch reflex of the human thumb.
syndrome.46 Journal of Physiology 440:243–55.
6. Hallett, M., D. Chadwick, and C. D. Marsden. 1979.
Cortical reflex myoclonus. Neurology 29:1107–25.
Key Points 7. Kourtis, D., H. F. Kwok, N. Roach, and A. M. Wing.
2008. Praamstra P. Maintaining grip: Anticipatory and
• Following a strong shock delivered to the reactive EEG responses to load perturbations. Journal
nerve of a muscle that is tonically contract- of Neurophysiology 99(2):545–53.
ing, a period of relative or absolute silence 8. Palmer, E., and P. Ashby. 1992. Evidence that a
long latency stretch reflex in humans is transcortical.
called the silent period begins immedi- Journal of Physiology (London) 449:429–40.
ately and persists for about 100 ms. 9. Palmer, E., and P. Ashby. 1992. The transcortical
• The silent period results from a combina- nature of the late reflex responses in human small
tion of peripheral, spinal cord, and cortical hand muscle to digital nerve stimulation. Experimental
Brain Research 91:320–6.
phenomena. 10. Matthews, P. B. 1991. The human stretch reflex and
• In states of hyperexcitability of the dis- the motor cortex. Trends in Neurosciences 14:87–91.
tal nerve or muscle, the silent period may 11. Hallett, M. 1989. Long-latency reflexes. In Disorders
be absent because ectopic impulses arise of movement: Clinical, pharmacological and physio-
distal to the stimulus. logical aspects, ed. N. P. Quinn, and P. G. Jenner,
529–41. London: Academic Press.
• The silent period may prove to be of value 12. Deuschl, G. 2003. Long-latency reflexes following
in a number of upper motor neuron clini- stretch and nerve stimulation. In Movement disorders.
cal disorders. Handbook of clinical neurophysiology, ed. M. Hallett,
• The testing of the silent period should Vol. 1, 285–94. Amsterdam: Elsevier.
13. Black, I., D. Nichols, M. Pelliccio, and J. Hidler. 2007.
be considered in instances where hyper- Quantification of reflex activity in stroke survivors dur-
excitable motor neurons are suspected but ing an imposed multi-joint leg extension movement.
the reason is unclear. Experimental Brain Research 183(2): 271–81.
14. Lee, Y. C., J. T. Chen, K. K. Liao, Z. A. Wu, and parameters on the reflex response. Electroencephalog-
B. W. Soong. 2003. Prolonged cortical relay time of raphy and Clinical Neurophysiology 61:287–98.
long latency reflex and central motor conduction in 31. Floeter, M. K. 2003. Cutaneous silent periods. Muscle
patients with spinocerebellar ataxia type 6. Clinical & Nerve 28:391–401.
Neurophysiology 114(3):458–62. 32. Merton, P. A. 1951. The silent period in a muscle
15. Defebvre, L. 2006. Myoclonus and extrapyramidal of the human hand. Journal of Physiology (London)
diseases. Neurophysiologie Clinique 36(5–6):319–25. 114:183–98.
16. Kanovsky, P. 2002. Dystonia: A disorder of motor pro- 33. Leis, A. A., M. A. Ross, T. Emori, Y. Matsue, and
gramming or motor execution? Movement Disorders T. Saito. 1991. The silent period produced by electri-
17(6):1143–7. cal stimulation of mixed peripheral nerves. Muscle &
17. Noth, J., K. Podoll, and H. H. Friedemann. 1985. Nerve 14:1202–8.
Long-loop reflexes in small hand muscles studied in 34. Fuhr, P., R. Agostino, and M. Hallett. 1991. Spinal
normal subjects and in patients with Huntington’s motor neuron excitability during the silent period
disease. Brain 108:65–80. after cortical stimulation. Electroencephalography and
18. Upton, A. R., A. J. McComas, and R. E. Sica. 1971. Clinical Neurophysiology 81:257–62.
Potentiation of “late” responses evoked in muscles dur- 35. Leis, A. A., I. Stetkarova, A. Beric, and D. S. Sto-
ing effort. Journal of Neurology, Neurosurgery, and kic. 1995. Spinal motor neuron excitability during the
Psychiatry 34:699–711. cutaneous silent period. Muscle & Nerve 18:1464–70.
19. Sutton, G. G., and R. F. Mayer. 1974. Focal reflex 36. Risk, W. S., E. P. Bosch, J. Kimura, P. A. Cancilla, K. H.
myoclonus. Journal of Neurology, Neurosurgery, and Fischbeck, and R. B. Layzer. 1981. Chronic tetanus:
Psychiatry 37:207–17. Clinical report and histochemistry of muscle. Muscle
20. Deuschl, G., K. Strahl, E. Schenck, and & Nerve 4:363–6.
C. H. Lucking. 1988. The diagnostic significance 37. Weinberg, D. H., E. L. Logigian, and J. J. Kelly Jr.
of long-latency reflexes in multiple sclerosis. Electro- 1988. Cervical astrocytoma with arm rigidity: Clinical
encephalography and Clinical Neurophysiology and electrophysiologic features. Neurology 38:1635–7.
70:56–61. 38. Pullman, S. L., B. Ford, B. Elibol, A. Uncini, P. C.
21. Naumann, M., and K. Reiners. 1997. Long-latency Su, and S. Fahn. 1996. Cutaneous electromyographic
reflexes of hand muscles in idiopathic focal dysto- silent period findings in brachial dystonia. Neurology
nia and their modification by botulinum toxin. Brain 46:503–8.
120:409–16. 39. Leis, A. A., M. Kofler, and M. A. Ross. 1992. The silent
22. Deuschl, G., C. H. Lucking, and E. Schenck. 1987. period in pure sensory neuronopathy. Muscle & Nerve
Essential tremor: Electrophysiological and pharmaco- 15:1345–8.
logical evidence for a subdivision. Journal of Neurol- 40. Schelhaas, H. J., I. M. Arts, S. Overeem, et al. 2007.
ogy, Neurosurgery, and Psychiatry 50:1435–41. Measuring the cortical silent period can increase diag-
23. Deuschl, G., C. H. Lucking, and E. Schenck. 1989. nostic confidence for amyotrophic lateral sclerosis.
Hand muscle reflexes following electrical stimulation Amyotrophic Lateral Sclerosis 8(1):16–9.
in choreatic movement disorders. Journal of Neurol- 41. Karandreas, N., M. Papadopoulou, P. Kokotis,
ogy, Neurosurgery, and Psychiatry 52:755–62. A. Papapostolou, G, Tsivgoulis, and T. Zambelis. 2007.
24. Chen, C. C., J. T. Chen, Z. A. Wu, K. P. Kao, and K. K. Impaired interhemispheric inhibition in amyotrophic
Liao. 1998. Long latency responses in pure sensory lateral sclerosis. Amyotrophic Lateral Sclerosis
stroke due to thalamic infarction. Acta Neurologica 8(2):112–18.
Scandinavica 98:41–8. 42. Attarian, S., A. Verschueren, and J. Pouget. 2007.
25. Caccia, M. R., A. J. McComas, A. R. Upton, and Magnetic stimulation including the triple-stimulation
T. Blogg. 1973. Cutaneous reflexes in small muscles technique in amyotrophic lateral sclerosis. Muscle &
of the hand. Journal of Neurology, Neurosurgery, and Nerve 36(1):55–61.
Psychiatry 36:960–77. 43. De Beaumont, L., M. Lassonde, S. Leclerc, and
26. Fuhr, P., and W. G. Friedli. 1987. Electrocutaneous H. Theoret. 2007. Long-term and cumulative effects
reflexes in upper limbs—reliability and normal values of sports concussion on motor cortex inhibition. Neu-
in adults. European Neurology 27:231–8. rosurgery 61(2):329–36; discussion 336–7.
27. Jenner, J. R., and J. A. Stephens. 1982. Cutaneous 44. Sartucci, F., S. Tovani, L. Murri, and L. Sagliocco.
reflex responses and their central nervous pathways 2007. Motor and somatosensory evoked potentials in
studied in man. Journal of Physiology (London) Autosomal Dominant Hereditary Spastic Paraparesis
333:405–19. (ADHSP) linked to chromosome 2p, SPG4. Brain
28. Fuhr, P., T. Zeffiro, and M. Hallett. 1992. Cuta- Research Bulletin 74(4):243–9.
neous reflexes in Parkinson’s disease. Muscle & Nerve 45. Badawy, R. A., J. M. Curatolo, M. Newton, S. F.
15:733–9. Berkovic, and R. A. Macdonell. 2007. Changes in
29. Chen, R., P. Ashby, and A. E. Lang. 1992. Stimulus- cortical excitability differentiate generalized and focal
sensitive myoclonus in akinetic-rigid syndromes. Brain epilepsy. Annals of Neurology 61(4):324–31.
115:1875–88. 46. Han, J. K., K. Oh, B. J. Kim, et al. 2007. Cutaneous
30. Meinck, H. M., S. Kuster, R. Benecke, and B. Conrad. silent period in patients with restless leg syndrome.
1985. The flexor reflex—influence of stimulus Clinical Neurophysiology 118(8):1705–10.
Chapter 33
Movement Disorders
John N. Caviness
INTRODUCTION PSYCHOGENIC JERKS

TECHNIQUES STARTLE DISORDERS
Surface EMG PERIODIC LIMB MOVEMENTS
EEG OF SLEEP
EEG–EMG Polygraphy with DYSTONIA
Back-Averaging Recording Techniques
Elicited Responses Abnormal Patterns
SURFACE EMG: NORMAL PATTERNS TICS, CHOREA, AND ATHETOSIS
TREMOR VOLUNTARY MOVEMENT
Recording Techniques ABNORMALITIES
Abnormal Patterns SUMMARY
MYOCLONUS
Recording Techniques
Abnormal Patterns
INTRODUCTION trained human eye that, together with the clin-

ical history, provides an accurate diagnosis in
The study of movement disorders encompasses most cases.
abnormalities of motor control that result in Clearly, reliance on diagnosis by visual
either too little or too much movement. At inspection is not always sufficient. Although
one pole are akinetic-rigid syndromes, such as observation is excellent for perceiving the
Parkinson’s disease, and at the other are invol- overall pattern of movement, it is less profi-
untary movements, such as tremor, myoclonus, cient in discerning the fine details of move-
dystonia, chorea, and tics. Movement disor- ment, such as timing (Which body part moved
ders stem from complex and poorly understood first?) and regularity (Is the movement tremu-
pathophysiologic processes that occur in the lous or irregular?). At times, details such as
central nervous system. The most valuable tool these are critical in diagnosis. Also, experimen-
in evaluating clinical movement disorders is the tal studies in motor control demonstrate clearly
551
that the brain, spinal cord, and musculoskeletal disposable adhesive Ag/AgCl ECG electrodes
system are able to produce a specific move- convenient because they can be applied rapidly
ment with a large number of different motor when multiple muscles are recorded. The
patterns. As a practical example, rapid elbow patient should be asked about any type of adhe-
flexion may result from either a brief, isolated sive allergies that he/she has had. After the
contraction of the biceps muscle or prolonged skin has been cleansed and mildly abraded,
activity of the biceps and triceps muscles. In the electrodes are placed 2–3 cm apart over
this example, identification of the underlying the motor point of the muscle and oriented
motor pattern may distinguish myoclonus from parallel to the course of the muscle fibers. Spe-
dystonia. cial care should be given to older individuals
Noninvasive clinical neurophysiology tech- with thin skin and those who are anticoagu-
niques provide information that complements lated. The iliac crest provides a relatively inac-
and extends the clinical examination. Surface tive site for the ground electrode. Electrode
electromyography (EMG) is the most impor- impedances should be below 5 k to get high-
tant of these techniques, but other useful quality recording, but higher impedances may
tests include electroencephalography (EEG), be acceptable if the surface EMG signal is
EEG–EMG polygraphy with back-averaging, devoid of artifact (e.g., 60 Hz).
and elicited responses that include evoked A technical limitation of surface EMG stud-
potentials and certain other reflex responses. ies is the lack of selectivity. The activity of
Multichannel EMG studies map the temporal a single muscle is never actually recorded
pattern of movement with a resolution mea- because adjacent muscles inevitably contribute
sured in milliseconds. Furthermore, the sur- “cross talk” to the signal through volume con-
face EMG reflects not only alpha motor neuron duction. This effect is minimized by use of
activity but also, by inference, specific abnor- short interelectrode distances and by record-
mal central commands that underlie a move- ing from relatively superficial and isolated
ment disorder. Patterns of abnormal and nor- muscles, such as the biceps, deltoid, quadri-
mal findings for certain movement disorders ceps, tibialis anterior, or first dorsal interosseus.
are well described, and these characteristics At times, a group of muscles, such as the
can be used as supportive evidence for a more forearm flexors or extensors, are intentionally
specific movement disorder diagnosis and/or recorded.
origin. The quality of the surface EMG signal
must be assessed carefully before analysis. This
Purpose and Role of Movement Disorder signal represents the interference pattern of
Clinical Neurophysiology multiple motor units with high frequencies
• Additional sensitivity and accuracy to filtered out by the intervening skin and sub-
cutaneous tissue. Deep muscles, such as the
clinical assessment of movement
gluteus maximus or any muscle in an obese per-
disorders.
• Typical neurophysiology pattern confirms son, may produce a signal that is too degraded
for analysis. The frequency spectrum of the
type of certain movement disorders.
• Identifies the source of certain movement signal contains power throughout the range
between 1 and 1000 Hz, with maximal power
disorders.
at approximately 100 Hz. In practice, a low-
frequency filter cutoff of 1–30 Hz is used to
eliminate the unwanted effects of DC potential
and low-frequency movement artifact. A high-
TECHNIQUES frequency filter setting of 200–3 kHz passes
the important high-frequency components of
Surface EMG the signal. The amplification factor is set arbi-
trarily to display a maximal voluntary con-
Surface EMG studies are noninvasive and are traction that fills the amplifier range with-
within the capability of any EMG laboratory. out blocking. After the EMG signal has been
Surface EMG recording can be performed collected, it may be displayed as the raw
with any high-quality disk electrodes. We find interference pattern or digitally processed to
Movement Disorders 553
display a full-wave rectified signal or smoothed the movement disorder, but paroxysmal abnor-
EMG envelope. In addition to correlating malities (e.g., epileptiform) are particularly
abnormal movements with EMG activity, it relevant.
is important to note discharge duration, vari-
ability, and timing relationships between mus-
cles. The amplitude of the bursts is extremely
variable and rarely useful in routine clinical EEG–EMG Polygraphy with
studies. Back-Averaging
If a study demands highly selective record-
ing, intramuscular electrodes must be used. The neurophysiologic evaluation of movement
Electrodes fashioned from fine wire are use- disorders is enhanced by simultaneous record-
ful for this purpose. Pairs of wires are inserted ing of EEG, surface EMG, and other modal-
into the selected muscle through a hypoder- ities. This allows the potential detection of
mic needle that is then withdrawn. After the specific relationships between different types
electrodes are in position, they remain sta- of physiological activity. Reflex activation by
ble for many hours and resist displacement by touch, deep tendon reflex, mixed nerve stimu-
even vigorous body movement. When selective lation (median, tibial), light, sound, and digital
recording is needed for only short recording nerve stimulation can be examined this way,
periods, standard concentric or monopolar and it is useful to add a channel to mon-
needle recording may be suitable. In any situa- itor the stimulus production. If available,
tion, the improved selectivity of intramuscular back-averaging can be performed by marking
recording must be balanced against the added the beginning of an EMG event or stimulus.
discomfort to the patient. For some movement This may be done online or offline. Epochs are
disorders (e.g., tremor), accelerometry is a use- then defined with time included both before
ful addition to EMG recording for frequency and after the event marker (or trigger). The
measurements. averaging of these epochs reduces the signal-
Recording samples should be taken dur- to-noise ratio and allows detection of time-
ing rest, postural activation (e.g., arms out- locked relationships between waveforms of the
stretched), kinetic activation (e.g., finger to same or different modalities. The larger num-
nose), functional tasks (e.g., drinking from a ber of epochs used in the calculation, the more
cup, handwriting), and mental activation (e.g., likely a smaller waveform will be discernable. A
counting backward), and while performing minimum of 100 epochs to show a time-locked
associated movements (e.g., contralateral repet- relationship between EEG, EMG, and other
itive hand movements). The condition or events is usually adequate. However, the result
state(s) that are known to bring out the move- of the waveform averaging should be critically
ment disorder should be emphasized. This may evaluated for its reproducibility, signal-to-noise
include sleep. ratio, and its ability to be interpreted across all
electrodes being used.
EEG
EEG electrodes are recorded from the stan- Elicited Responses
dard 10–20 positions. These same positions
may be used as for routine clinical record- Evoked potentials and surface EMG res-
ings, but if some positions are sacrificed in ponses to stimulation may be performed by
order to make room for surface EMG chan- standard techniques. The reader is referred
nels, then the frontal, central, parietal, and to the chapters dealing with these topics. The
midline positions should be a minimum as choice of what type of elicited response to
they are usually the most active in movement test for is dictated by the known abnormalities
disorders. EEG is very useful for correlat- in various movement disorders. The somato-
ing the state of consciousness with abnor- sensory evoked potential (SEP) and long-
mal movement. Any EEG abnormality seen latency surface EMG reflexes are the most
may have important diagnostic implications for common.
Key Points of 50–100 ms duration and a silent period in

the agonist. A final agonist burst usually com-
• Multiple neurophysiology techniques are
pletes the EMG activity. The triphasic pattern
available to noninvasively study move- appears to be fundamental to the motor con-
ment disorders. trol of ballistic limb movements.1 The theory
• The particular techniques used will depend
explaining this is that the initial agonist burst
on the movement disorder being studied
scales the size of the ballistic movement, the
and the presentation of that movement
antagonist burst brakes the limb, and the final
disorder in that specific patient.
agonist burst determines its destination. Bal-
listic (or phasic) and tonic patterns are best
thought of as two poles on a spectrum of vol-
untary movement. Most of the time, normal
SURFACE EMG: NORMAL surface EMG activity consists of varying super-
PATTERNS imposed combinations of ballistic or phasic and
tonic muscle activation at any one instant and is
Three normal patterns of surface EMG activ- coordinated as needed across different muscle
ity are recognized: reflex, tonic, and ballis- segments. Movement disorders reflect abnor-
tic (Fig. 33–1). Reflex activity, such as the mal involuntary partial or complete alteration
monosynaptic tendon jerk, produces brief, of these patterns
synchronized discharges of alpha motor neu-
rons. The surface EMG appearance of this Key Points
discharge is a short (10–30 ms) burst of activ-
ity in agonist and, at times, antagonist muscles. • The three basic types of normal human
This reflex pattern is an involuntary response to movement are reflex, tonic, and ballistic.
stimulation. Indeed, most persons are unable • These basic types do appear in move-
to generate voluntarily bursts that have this ment disorders but in abnormal ways and
short duration. However, voluntary movement circumstances.
produces two types of EMG patterns. When a
person moves a limb slowly or holds it in a static
posture, a tonic pattern results. This pattern
consists of a continuous and steady EMG dis- TREMOR
charge, often with cocontraction of agonist and
antagonist muscles. Cocontraction is a normal Tremor is an oscillation of one body part in rela-
mechanism of motor control that increases the tion to another body part that is approximately
stiffness, or resistance, across a joint. By con- sinusoidal and rhythmic. The three sources
trast, when one wills a very rapid movement of tremor are mechanical, reflex, and a cen-
of a joint, a ballistic pattern develops. An ini- tral oscillator. All physical objects oscillate at
tial agonist burst of 50–100 ms duration leads a resonant frequency related to their inertia
the pattern, followed by an antagonist burst and stiffness. Because body parts are physical
Figure 33–1. The three normal surface EMG patterns: A, reflex pattern; B, triphasic pattern; and C, tonic pattern. Upper
trace, agonist muscle; lower trace, antagonist muscle.
objects, they have a resonant frequency. In silent, in an alternating or reciprocal pattern. In

addition, muscles are connected to the central other tremors, the pair may fire simultaneously
nervous system through peripheral nerve reflex in a synchronous or cocontracting pattern.
loops, which oscillate at varying frequencies. Occasionally, the EMG pattern may shift from
Also, areas of the central nervous system oscil- one pattern to another during the period of
late spontaneously, possibly producing rhyth- recording. It is more common to see different
mic motor activity in the related body part. patterns in different conditions, such as recip-
The contribution of all three of these sources rocal during rest and more synchronous firing
can contribute to the clinical phenomenon of during muscle activation. Long-term EMG-
tremor. based automated analysis may be helpful in
In tremor, surface EMG discharges reflect separating parkinsonian tremor from essential
the oscillating positive and negative influ- tremor.4
ences on the intended voluntary activation or
resting state. The surface EMG of tremors
records the grouping of motor unit poten- Abnormal Patterns
tials as discrete bursts of activity.2 Analysis of
these EMG bursts helps to establish whether a Abnormal patterns of neurophysiologic find-
movement is truly tremulous. Disorders such ings are seen with the various types of tremor
as phasic dystonia may appear regular on (Fig. 33–2 and Fig. 33–3). The clinical and
visual inspection but be shown to be irreg- electrophysiologic features of the different
ular when measured on an EMG recording. forms of tremor are summarized in Table 33–1.
Conversely, a low-frequency tremor with some
irregularity may appear so jerky as to be
EXAGGERATED PHYSIOLOGIC
myoclonus.
TREMOR
Even normal subjects have a fine, often imper-
Recording Techniques ceptible tremor when holding part of the body
in a static posture. Physiologic tremor is pri-
Electrodes are placed over the agonist and marily a mechanical tremor, but in some per-
antagonist pairs of muscles both including sons a central 6–12 Hz oscillator may have
and just beyond the tremor distribution. For a role.5, 6 The mechanical portion of normal
hand tremor, forearm flexors and extensors physiologic tremor shows no oscillating surface
are most active and care must be taken to EMG discharges. However, placing weights
eliminate cross talk between the signals by on the body part shifts the frequency of the
electrode positioning, because this technical mechanical component lower. Under circum-
error may lead to misinterpretation. Ampli- stances that increase catecholaminergic secre-
tude of the tremor bursts is assessed in tion, such as anxiety, fatigue, hypothermia,
various limb positions. Tremors are grouped hypoglycemia, thyrotoxicosis, or pheochromo-
into those occurring at rest and those occur- cytoma, physiologic tremor increases in ampli-
ring with action. Action tremors are fur- tude and becomes visible. In this state, known
ther subdivided into postural tremors (a as exaggerated physiologic tremor, underlying
body part maintains a position against grav- surface EMG activity emerges through reflex
ity), isometric tremors (muscle contraction activation. This may consist of a low-grade
against a stationary object), simple kinetic interference pattern or synchronous bursts
tremors (nontarget-directed voluntary move- of 50–100 ms duration and a frequency of
ment), and intention tremors (at the termi- 8–12 Hz. This pattern is often best recorded in
nation of a target-directed movement).3 Task- distal upper extremity muscles, such as finger
specific tremors are those seen during a defined extensors and flexors.
activity such as handwriting. The frequency
should be noted for every distribution and acti- ESSENTIAL TREMOR
vation state. Another variable used in tremor
analysis is the pattern of agonist–antagonist fir- Essential, or familial, tremor is the most
ing. One muscle may fire while the other is common movement disorder. It is usually
Table 33–1 Classification of Tremor by Localization and Electrophysiologic

Features
Tremor type Activation Dominant Pattern of Other features
frequency firing
Physiologic Postural 6–12 Hz Regular Exacerbated anxiety,

fatigue, hypothermia,
hypoglycemis,
thyrotoxicosis,
pheochromocytoma
Essential Postural or kinetic 4–12 Hz (5–8 Hz Regular
most common)
Parkinsonian Rest (may be 4–7 Hz Regular
postural after
20 seconds of
activation)
Cerebellar Kinetic <5 Hz Irregular Most prominent
toward termination of
movement
Holmes Rest to postural <4.5 Hz Irregular
activation or
posture to kinetic
movement
Task-specific Specific tasks Irregular
(e.g., writing)
Orthostatic (shaky Standing or 13–18 Hz Regular May be present in arms
leg syndrome) stretching legs
Palatal Rest 1–3 Hz Regular
Psychogenic Widely variable Widely variable Irregular
a bilateral, symmetrical, postural, or kinetic slower in frequency and poorly responsive to

tremor involving the upper extremities. Addi- propranolol. Milanov9 has reported that the
tional or isolated head tremor may also occur.3 alternating pattern is more common than pre-
The voice can also be affected. Essential viously believed. This dictates caution in diag-
tremor is thought to result from a central nosing tremor on the basis of a single EMG
oscillator (an often suggested but unproven characteristic. Additionally, although essential
candidate involves cerebellar and brain stem tremor is commonly thought of as a pos-
pathways).6 The surface EMG usually shows tural tremor, electrophysiologic studies have
bursts of activity that are synchronous in ago- shown that the kinetic component may be
nist and antagonist muscles, with a possible more prominent than a static postural compo-
frequency range of 4–12 Hz and 5–8 Hz being nent.10 The surface EMG findings in neuro-
the most common. Of interest, the frequency pathic tremor are identical to those in essential
of essential tremor declines with age.7 When tremor.
the tremor is severe, less prominent activity
while the limb is at rest may be recorded.
Sabra and Hallett8 have reported patients PARKINSONIAN TREMOR
with essential tremor who clearly had an
alternating EMG pattern. This group of Tremor commonly accompanies akinetic-rigid
patients appeared to have a tremor that was syndromes such as Parkinson’s disease.
Figure 33–2. Essential and parkinsonian tremor recorded from antagonistic muscles. Note that in essential tremor the
muscle contractions are simultaneous, while in parkinsonian tremor they are alternating. (Courtesy J. Matsumoto.)
Clinically, parkinsonian tremors appear max- rather a series of inaccurate, irregular ballistic
imal at rest and attenuate with action. The movements.
dominant frequency is 4–7 Hz. Surface EMG
studies of parkinsonian tremors demonstrate
an alternating pattern of contraction that HOLMES TREMOR
is constant through the period of record- Holmes tremor (also known as rubral tremor,
ing (Fig. 33–2). Burst durations are typically midbrain tremor, thalamic tremor, or myorhy-
in the range of 50–100 ms. The frequency of thmia) increases while going from rest to
the bursts varies little through the period of postural activation and from postural activa-
the recording. The regularity of the tremor tion to kinetic or intention movement. This
increases as the disease progresses.11 Static tremor occasionally has an irregular presenta-
postures initially attenuate the burst ampli- tion. Holmes tremor frequency is usually less
tudes; however, when such postures are held than 4.5 Hz and often associated with a recog-
for 20 seconds or longer, the tremor bursts nized pathologic insult (e.g., infarct or demyeli-
may reappear (“re-emergent tremor”). The fre- nating plaque), in which case a delayed onset
quency of the postural or kinetic component (weeks to 2 years) is common.3
is usually the same as the rest component,
but sometimes it can be substantially (1.5 Hz)
higher. This may be the case in patients who TASK- AND POSITION-SPECIFIC
have essential tremor alone and in those with TREMOR
parkinsonism.3 Several tremors occur only with specific tasks
or positions. The classic example is primary
writing tremor. Although this tremor is pre-
CEREBELLAR TREMOR
dominant during writing, it often spills over
Cerebellar tremor occurs with kinetic and into other activities such as eating or grooming.
intention limb movements. It is most promi- The surface EMG correlate consists of bursts
nent toward the termination of the move- of 100 ms duration that occur maximally in the
ment. The tremor frequency is mainly less than pronator teres, supinator, or wrist flexors and
5 Hz, and the distribution may be distal, prox- extensors. The pattern may be synchronous or
imal, or both. Postural tremor may also be alternating. Taps to the forearm, particularly
present. Serial dysmetria is often confused with in a direction that produces supination of the
intention tremor. In dysmetria, surface EMG forearm, may stimulate bursts of tremor. Iso-
studies indicate that the terminal movements lated voice tremor is another example. It is not
are not the regular oscillations of tremor but known whether these tremors represent a form
of essential tremor or dystonia or have another PALATAL TREMOR

cause.3
Palatal tremor is a rhythmic, involuntary move-
ment that causes palatal elevation. At times,
it may spread to involve tongue, neck, facial,
ORTHOSTATIC TREMOR (SHAKY and even limb muscles. Palatal EMG activity
LEGS SYNDROME) is detected with surface electrodes linked to
Heilman12 has described a distinctive tremor, the mastoid processes or placed on the anterior
called orthostatic tremor, that occurs predom- neck. Bursts recur at a regular frequency
inately in the elderly. Patients may complain of approximately 1–3 Hz. Palatal tremor can
of quivering, vibration, or “shaking” in their stem from a symptomatic lesion, with associ-
legs shortly after they stand. This may cause ated olivary hypertrophy, or it can be essen-
a sense of instability so severe that walk- tial in nature. Rhythmic movements of the
ing becomes difficult. The surface EMG pat- levator veli palatini muscles cause the symp-
tern is distinctive and displays high-amplitude tomatic palatal tremors, but the essential
13–18 Hz tremor bursts (Fig. 33–3). Because palatal tremors involve mainly the tensor veli
of the rapid frequency, this diagnosis can palatini. Involvement of the latter muscle often
be difficult to make on the basis of clin- causes an audible ear click to the patient. The
ical observation. A surface EMG study is term palatal myoclonus is reserved for when
often the only way to make the definitive the movements are irregular and do not resem-
diagnosis. The bursts are recorded in the ble a sinusoidal pattern.
legs and paraspinal muscles, with the patient
standing. The agonist–antagonist relationship PSYCHOGENIC TREMOR
may vary during the recording. At rest, the
legs are electrically silent. Tremor discharges Occasionally, tremor may have a pattern that
may be present with the legs outstretched is atypical, and a conversion disorder or malin-
while sitting and also may be present in gering is suspected. Surface EMG can provide
the arms. supportive evidence only by demonstrating a
Anterior tibial
Gastrocnemius
Vastus lateralis
Hamstrings
Lumbar paraspinals
Thoracic paraspinals
Biceps
Triceps
1 mV
1 second
Figure 33–3. Orthostatic tremor. Surface EMG activity recorded with the patient standing and rectified for display.
Tremor activity at approximately 15 Hz is recorded maximally from the leg muscles.
pattern that does not correspond to any of • Orthostatic tremor (shaky leg syndrome)
those described above. Psychogenic tremors demonstrates high-amplitude bursts at
tend to be paroxysmal, have inconsistent acti- 13–18 Hz frequency.
vation states, and seldom display a dominant • Psychogenic tremor tends to be parox-
frequency throughout a prolonged recording. ysmal, has inconsistent activation states,
Indeed, the tremor frequency and amplitude and rarely displays a dominant frequency
tend to vary widely with time, change of posi- throughout a prolonged recording.
tion, or distraction. However, diagnostic proof
for a psychogenic or voluntary origin for a
tremor cannot be offered. When such clas-
sic psychogenic features are not present, psy- MYOCLONUS
chogenic tremors can show surface EMG pat-
terns that overlap with those patterns discussed Myoclonus is a clinical sign defined as sud-
above, thus limiting the diagnostic utility of den, brief, shock-like, involuntary movements
surface EMG for psychogenic tremors. caused by muscular contractions or inhibi-
tions.13 Muscular contractions produce “posi-
Key Points tive myoclonus,” whereas muscular inhibitions
produce “negative myoclonus” or asterixis.
• Surface EMG discharges record bursts of Myoclonic movements have now been recog-
motor unit potentials that reflect the oscil- nized to have many possible etiologies, anatom-
lating positive and negative influences on ical sources, and pathophysiologic features.2
the intended voluntary activation or rest- Myoclonus may be classified by examination
ing state. findings, presentation, etiology, and clinical
• Surface EMG can demonstrate or confirm neurophysiology testing. Important exam find-
regularity and agonist–antagonist relation- ings are activation state(s) and distribution.
ships more clearly than visual inspection. They may be focal (involving only a single limb
• Surface recording of tremor uses elec- or area), multifocal (affecting more than one
trodes placed over agonist and antagonist body part in a random, independent fashion),
muscles; their frequency and pattern of generalized (involving all body parts simultane-
agonist–antagonist firing are used to clas- ously), or segmental (involving only muscles of
sify the tremor. a given cranial or spinal segment).
• Different clinical types of tremor demon- The best method to organize the possi-
strate characteristic frequency but there is ble clinical presentations and etiologies of
considerable overlap. myoclonus is by using the major clinical
• Exaggerated physiologic tremor consists syndrome categories of Marsden et al. and
of bursts of 50–100 ms duration at a fre- their corresponding lists of known etiologies.14
quency of 8–12 Hz, most often recorded The Marsden et al. clinical syndrome cate-
in the distal upper extremity. gories are (1) physiologic myoclonus includes
• Essential tremor demonstrates bursts of motor phenomena such as hiccups, hypnic
activity at 4–12 Hz frequency. jerks, and the startle response, which have the
• Parkinsonian tremor is recorded maxi- appearance of myoclonus but occur in normal
mally at rest and has a dominant frequency subjects; (2) essential myoclonus designates
of 4–7 Hz; the burst frequency varies little disease in which abnormal muscle jerks are
during the recording. the primary feature of the illness; (3) epileptic
• Cerebellar tremor is usually less than myoclonus refers to myoclonus in the setting
5 Hz in frequency and is most prominent of epilepsy; and (4) symptomatic myoclonus
when recording toward the termination of represents all other disease states in which
movement. myoclonus occurs as a sign, often in the setting
• Holmes tremor (midbrain tremor) increases of diffuse neuropathology.
during activation of the limb and is occa- The goal for clinical neurophysiology study
sionally irregular, with frequencies less of myoclonus is to classify its origin and phys-
than 4.5 Hz. iological properties15 (Table 33–2). If suc-
• Task-specific tremor occurs only in certain cessful, this information is valuable in the
positions. diagnosis and treatment of myoclonus. Major
Table 33–2 Classification of Myoclonus by Localization and Electrophysiologic
Features
Localization Clinical Burst duration EEG Time-locked Etiologies
features abnormalities cortical
potentials
CORTICAL
Cortical Focal, 50 ms Yes Yes—Precedes Posthypoxia
Reflex multifocal, or myoclonus by Lance–Adams
Myoclonus generalized short latency syndrome
Focal spike Enlarged Progressive
with EEG cortical SEP myoclonic
back-averaging waves epilepsy
(precedes Reflex-induced syndromes
myoclonus by myoclonus Toxic and drug
6–22 ms in induced
upper Enhanced
extremity) long-latency
EMG response
(40–60 ms) to
electrical
stimulation
Cortical Focal, 50 ms Yes No enlarged Parkinson’s
origin multifocal, or Focal transient cortical SEP disease
myoclonus generalized with EEG waves Dementia with
without reflex back-averaging No Lewy bodies
activation reflex-induced Drugs
myoclonus
No
long-latency
EMG response
to electrical
stimulation
Focal motor Focal <100 ms Yes
seizures
Focal spike,
spike-and-
wave, sharp
wave, rhythmic
theta or delta,
or PLEDs
Myoclonus in Multifocal or <100 ms Yes, but not Variable
Alzheimer’s generalized always enlarged
disease time-locked cortical SEP
waves
Variable
long-latency
EMG response
to electrical
stimulation
(Continued)
560
Table 33–2 (Continued)
potentials
Myoclonus in <50 ms Yes; variable Variable

Creutzfeldt–Jakob correlation of enlarged
disease sharp waves cortical SEP
with myoclonus waves
Variable
long-latency
EMG response
to electrical
stimulation
Myoclonus with >200 ms Yes; high-voltage
subacute sclerosing polyphasic sharp
panencephalitis and slow wave
complexes
(500–2000 ms
duration, every
4–15 seconds)
Myoclonus in 25–50 ms No Long-latency
corticobasal EMG response
degeneration at 40 ms to
median nerve
stimulation
Asterixis 50–200 ms
(EMG
silence)
CORTICAL– Occurs in Yes Absence
SUBCORTICAL paroxysms at seizures
rest Primary
generalized
myoclonic
seizures
Generalized,
bilaterally
synchronous,
focal, or
multifocal
SUBCORTICAL– Longer No No
SUPRASEGMENTAL
Essential myoclonus Agonist only 50–200 ms No No
or agonist–
antagonist
cocontraction
Irregular in
amplitude,
duration, and
timing
Reticular reflex Generalized <50 ms Yes; variable No Posthypoxia
myoclonus jerks at rest
(Continued)
561

potentials
Stimulus Simultaneous, Uremia

sensitive bilaterally
rostral and
caudal
recruitment
Opsoclonus– Multifocal < 100 ms No No
myoclonus Predominantly Occur in
induced by trains
action
Propriospinal Trunk flexion 50–300 ms No No
myoclonus or extension
At rest or
stimulus-
activated
SEGMENTAL Frequency of 50–500 ms No No Palatal myoclonus
0.5–3 Hz
PERIPHERAL Focal Variable No No Hemifacial spasm
categories of the physiological classification the full pattern of any epileptogenic discharges
used here primarily refer to the neuroanatomic and other EEG abnormalities. Ideally surface
source of the myoclonus physiology types. Fur- EMG should be combined with simultaneous
ther subdivision is based on other physio- EEG recording (EEG–EMG polygraphy) for
logical properties as well as the clinical syn- correlation and EEG–EMG back-averaging.
drome and/or the specific disease in which Offline trigger placement for averaging is ideal
the myoclonus occurs. In practical terms, to choose epochs with typical myoclonus EMG
the classification for a particular example of discharges and no artifact. Online averaging
myoclonus is derived from the clinical neuro- (“jerk-locked averaging”) may be performed
physiological findings as well as an appreciation with a motion detector (e.g., accelerometer)
of the clinical context in which they occur. or a rectified EMG window detector. Besides
The main physiological classification categories assessment of individual myoclonus EMG dis-
for myoclonus are (1) cortical, (2) cortical– charges, it is useful to inspect the average
subcortical, (3) subcortical–suprasegmental, rectified myoclonus EMG discharges to deter-
(4) segmental, and (5) peripheral.16 One should mine time-locked relationships between dif-
be aware that multiple myoclonus physiology ferent muscles. All clinically apparent acti-
types might occur in the same patient. Physio- vation states should be collected for further
logical classification can assist with distinguish- analysis. Evoked potentials, at least median
ing types of etiologies between and within the SEPs, should be obtained. Long-latency EMG
major clinical syndrome categories. reflexes to median and/or tibial nerve stimula-
tion should be done.
Abnormal Patterns
Multiple clinical neurophysiology techniques
should be applied to the patient with myoclo- Abnormal patterns of neurophysiologic find-
nus. A routine EEG is appropriate to capture ings are seen with the various types of
myoclonus. The clinical and electrophysiologic patients usually have reflex jerks, it is usually
features vary according to the generator sites of easier to collect many myoclonus events by
the myoclonus (Table 33–2). muscle activation of the limb. The back-
averaged transient is typically a focal, bipha-
sic, or triphasic spike beginning with a pos-
CORTICAL
itive deflection that precedes the onset of
The cerebral cortex is the most common origin the myoclonic discharge by 6–22 ms in the
for myoclonus. The jerks are most often mul- upper extremity: the more distal muscle
tifocal; but focal, segmental, and generalized the myoclonus is recorded from, the longer
myoclonus can also occur. Action myoclonus the time interval.17 The duration of the tran-
is very common in these patients and provides sient is 15–40 ms. The conduction of the
most of the disability. At rest, myoclonus usu- spike to motor neuron pools is presumed
ally will be less prominent. Myoclonus induced to occur by corticospinal (pyramidal) path-
by reflex stimulation is common, and its char- ways. The maximum of the transient is usu-
acterization is important for physiological clas- ally located over the sensorimotor cortex at
sification. A common presentation is to have the central or centro-parietal electrode accord-
myoclonus with a combination of action and ing to anatomical somatotopic mapping, con-
reflex precipitants, and presence at rest. The tralateral to the myoclonus EMG discharge
vast majority of cortical myoclonus patients (Fig. 33–4).
have one or more of the three major cortical Enlargement of the cortical SEP P25–N33
physiology types: (1) cortical reflex myoclonus, parietal wave from median nerve stimulation
(2) cortical origin myoclonus without reflex is an important evidence for cortical reflex
activation, and (3) focal motor seizures. More myoclonus physiology. The enhanced early cor-
unusual physiological descriptions have been tical components of the giant SEPs may be
reported for Alzheimer’s disease, Creutzfeldt– generated by the somatosensory and primary
Jakob disease, subacute sclerosing panen- motor cortices.18 The establishment of normal
cephalitis (SSPE), corticobasal degeneration, values for a particular laboratory is encour-
and asterixis. All of these physiological descrip- aged with consistent methods and electrode
tions of cortical origin myoclonus will be dis- derivations being used. Shibasaki et al. pub-
cussed below. lished an upper limit for P25–N33 amplitude
at the postcentral electrode of 8.6 μV using an
Cortical Reflex Myoclonus ear reference, Ugawa used 10.8 μV with Fz
reference, and the normal upper limit value
This type of cortical myoclonus physiol- in our laboratory is 11.1 μV with Fz refer-
ogy is the predominant type in posthypoxic ence.19, 20 Sometimes the cortical SEP waves
myoclonus or Lance–Adams syndrome, pro- are “giant” and deviate from the morphol-
gressive myoclonus epilepsy syndromes, toxic ogy and distribution seen in normal individu-
and drug-induced myoclonus, and many other als. The definition of “giant” SEP is arbitrary
etiologies. Cortical reflex myoclonus is defined but >20 μV is a commonly used value. In
by the demonstration of a focal time-locked addition to the P25–N33 wave, the parietal
cortical transient that precedes the myoclonus N20–P25 and/or frontal P22–N30 are enlarged
by a short latency (<40 ms for arm) in asso- less often.
ciation with evidence for exaggerated reflex In most cases, abnormal long-latency EMG
cortical phenomena. This may include one or discharges are elicited by median nerve stim-
more of the following: (1) enlarged cortical ulation that shows EMG discharges at 50-ms
SEP waves, (2) reflex-induced myoclonus, and latency or greater (range 40–60) from the
(3) enhanced long-latency EMG responses to stimulus artifact trigger mark.17 Repetitive dis-
electrical nerve stimulation. charges may be seen, at intervals of 20–40 ms.21
Although spikes and/or sharp waves are At rest, in a normal individual, no response
sometimes present in the gross EEG, back- should be present. Care must be taken that
averaging of the EEG–EMG polygraph is the arm muscles are relaxed so as to avoid a
the preferred method for demonstrating a false positive response. Brown et al. found that
time-locked cortical transient preceding a intrahemispheric and interhemispheric spread
myoclonus EMG discharge. Although such in a grossly somatotopic fashion from a focus
Right
wrist
flexors
Right
wrist
extensors
+
500 μV
–
200 ms
0 1 2
C3 CZ C4
–10
–μV 0
+10
–200 0 +200 0 0
ms
Right wrist extensors

+400
–μV 0
–400
–200 0 +200
ms
Figure 33–4. Top, EMG polygraphy of right forearm muscles showing tonic EMG followed by myoclonus EMG dis-
charges (arrows). Bottom, EEG–EMG back-averaging of myoclonus EMG discharge. Back-averaged focal EEG transient
present over left sensorimotor cortex area (electrode C3).
564
in one hemisphere can produce these bilat- Alzheimer’s disease.25 The gross EEG can
eral and/or generalized jerks.22 Because of the show background slowing and abnormal slow
fast spread, the clinical jerking appears almost waves. Focal sharp waves or sharp and slow
synchronous. Cortical tremor refers to rela- waves may occur. Periodic or quasiperiodic
tively rhythmic distal upper extremity EMG sharp waves sometimes occur with similarity
discharges during action at approximately 9 Hz to Creutzfeldt–Jakob disease. The relationship
and duration around 50 ms.23 Despite the phe- of these gross EEG changes and events to
notypic designation of “tremor,” these dis- myoclonus is usually not clear. The myoclonus
charges were found to fit all the criteria of EMG discharges are less than 100 ms, and
cortical reflex myoclonus physiology. may occur in an agonist-only pattern or with
cocontraction in antagonists and other muscles.
Enlargement of the cortical SEP waves and
Cortical Origin Myoclonus Without
the presence of long-latency EMG responses
Reflex Activation
to median nerve stimulation are variable.
The establishment of cortical reflex myoclonus The most commonly reported instance is a
as a distinct cortical physiology was an impor- focal contralateral central negativity with onset
tant step for the study of myoclonus mecha- 20–40 ms premyoclonus EMG latency and
nism. However, it has become apparent that 40–80 ms duration, but longer premyoclonus
myoclonus may have a focal time-locked cor- EMG latencies with longer and more widely
tical transient that precedes the myoclonus but distributed EEG waves occur.
is unassociated with clinical reflex myoclonus,
enhanced long-latency EMG responses to elec- Creutzfeldt–Jakob Disease
trical nerve stimulation, or enlarged cortical
SEP waves. This physiology has been seen The myoclonus in Creutzfeldt–Jakob disease
with myoclonus occurring in Parkinson’s dis- can occur in early, middle, or late stages. EEG
ease, dementia with Lewy bodies, hereditary findings often show abnormal slow and/or sup-
diffuse Lewy body disease, drugs, and other pressed background and generalized periodic
conditions.24 sharp wave discharges. The EMG duration is
<50 ms and an agonist-only pattern or with
cocontraction in antagonists and other mus-
Focal Motor Seizures
cles is observed. There is a variable correlation
There are a variety of ictal EEG changes between the timing of the myoclonus and the
that may be seen in the contralateral motor sharp wave discharges on routine EEG. When
area corresponding to a focal motor seizure back-averaging is used, a broadly distributed
manifestation. Repetitive focal spike, spike- contralateral negative transient is seen.26 This
and-wave, sharp wave, rhythmic theta or delta EEG correlate has 100–160 ms duration and
activity, or desynchronization may occur. In latency to the myoclonus EMG discharge of
many cases, no grossly observable EEG activ- 50–85 ms. Enlargement of the cortical SEP
ity is seen, and back-averaging may uncover waves and enhanced long-latency reflexes is
a transient in some of those cases. In the variable.
case of epilepsia partialis continua, the above-
mentioned transients will be periodic and may
Subacute Sclerosing Panencephalitis
even occur in the pattern of periodic lateral-
izing epileptiform discharges (PLEDS). The These patients can show periodic movements
EMG discharge duration is usually less than that appear as a jerk followed by a momen-
100 ms. tary sustained position and then gradually melt
away to the static position. These movements
Alzheimer’s Disease often occur in the upper extremities. An EMG
burst duration of greater than 200 ms can be
The myoclonus in Alzheimer’s disease is usu- seen for this “dystonic myoclonus” of SSPE.
ally multifocal, although it can be general- In contrast to Creutzfeldt–Jakob disease, the
ized. A few different electrophysiological pat- jerks have a consistent relationship to periodic
terns of myoclonus have been described in complexes on routine EEG. These complexes
consist of high voltage (300–1500 μV), repet- at rest are not present. The small-amplitude
itive, polyphasic, and sharp and slow wave myoclonus arises from an abnormal discharge
complexes ranging from 500 to 2000 ms in from the sensorimotor cortex.29
duration, usually recurring every 4–15 seconds
or sometimes longer. The complex nature of
the discharge makes it difficult to measure Asterixis
latency between the EEG discharge and the
Negative myoclonus refers to a decrease
jerk EMG discharge. The complexes are typ-
in tonic EMG activity. The term asterixis
ically widespread and synchronous. The SEP
is considered to be equivalent to nega-
and long-latency EMG responses have not
tive myoclonus. Negative myoclonus correlates
been adequately studied in SSPE. A slow neg-
with an average EMG silence duration of
ative potential shift has been found to precede
50–200 ms. Three types of EMG patterns have
the jerk and EEG complex.27
been described.30 Type I consists of abrupt
onset and offset of EMG silence during vol-
Myoclonus of Corticobasal untary muscle activation. A type I primarily
Degeneration and Parkinson’s negative event that is associated with a brief,
Disease discrete burst of EMG activity that precedes
the silence characterizes type II. Type III
Myoclonus is an important feature of cor-
silent periods are those that follow typical
ticobasal degeneration and occurs in 50%
positive myoclonus, especially in trains. The
of cases. Its clinical presentation parallels
EMG silences of negative myoclonus usually
that of the overall syndrome including a
have a multifocal distribution. The EEG cor-
focal distribution in the arm (sometimes
relate of type II negative myoclonus is sim-
leg) associated with other focal limb man-
ilar to that of positive cortical myoclonus.
ifestations that can include apraxia, rigid-
Type I negative myoclonus does not have an
ity, dystonia, and alien limb phenomenon.
EEG correlate and may have a subcortical
The myoclonus is prominent with muscle
generator.
action and often has reflex activation to
cutaneous stimuli and deep tendon reflexes
of the affected limb. The EMG discharge CORTICAL–SUBCORTICAL
duration for the myoclonus in corticobasal
degeneration is 25–50 ms with a cocontrac- There is strong evidence that some gener-
tion pattern. The myoclonus physiology in alized seizure phenomena arise from parox-
corticobasal degeneration has distinctive fea- ysmal abnormal and excessive oscillation in
tures.28 The EEG shows no correlate to the bidirectional connections between cortical and
myoclonus EMG discharges, even when back- subcortical sites. The term cortical–subcortical
averaging is performed. Parietal SEP cortical myoclonus refers to myoclonus arising from
waves are either normal or poorly formed, this type of physiology and other similar phe-
and frontal P22–N30 components are usually nomena. For these entities, the abnormal influ-
intact. Median nerve stimulation reveals an ence of the subcortical input is critical. Despite
EMG reflex at about 40 ms, and there can be a the subcortical involvement, the cortical dis-
response to digital nerve stimulation at 50 ms. charge precedes and drives the myoclonus
This response is thought to be quite character- event. This myoclonus usually occurs in parox-
istic, but probably not specific, for corticobasal ysms from rest and can be associated with
degeneration. other seizure phenomena that may even be
In Parkinson’s disease, multichannel sur- more clinically significant than the myoclonus
face EMG recording during muscle activation itself. The myoclonus is often generalized or
has shown irregular, multifocal, brief (<50 ms) bilaterally synchronous, but focal or multifo-
myoclonus EMG discharges. Back-averaging cal distributions occur as well. This classifica-
showed focal, short-latency, electroencephalo- tion includes absence and primary generalized
graphic transients prior to the myoclonus myoclonic seizures. These electrophysiology
EMG discharge. Cortical SEP waves are not patterns are described in other chapters of this
enlarged, and long-latency EMG responses book.
SUBCORTICAL–SUPRASEGMENTAL Opsoclonus–Myoclonus Syndrome

The clinical and neurophysiological charac- The opsoclonus–myoclonus syndrome consists
teristics of subcortical myoclonus are more of opsoclonus, myoclonus, and variably other
variable than for those in cortical or cortical– manifestations such as ataxia, tremor, and
subcortical myoclonus. The myoclonus EMG behavior problems. The myoclonus is usu-
duration may be longer than that in corti- ally multifocal and predominantly induced by
cal or cortical–subcortical myoclonus. Simul- action. The EMG discharges are less than
taneous rostral and caudal recruitment of the 100 ms, often occur in trains, and can show
myoclonus in the EMG channels supports a agonist-only or agonist–antagonist cocontrac-
subcortical generator. As one might expect, tion. The EEG is usually normal and if
there is no evidence for abnormal cortical abnormalities do exist, they have no relation-
excitability (e.g., EEG spikes, enlarged corti- ship to the myoclonus. Back-averaging of the
cal SEP waves) in subcortical myoclonus that myoclonus shows no EEG correlate.32 The
is tightly correlated to the myoclonus. Assign- SEP cortical waves and long-latency EMG
ment of a case to the subcortical category can responses are normal.
be problematic if it is based largely on absence
of evidence for abnormal cortical excitability
or circumstantial findings rather than direct Propriospinal Myoclonus
evidence. This myoclonus occurs as trunk flexion or
extension with axial muscle activation. Prox-
imal limb muscles are often involved in the
Essential Myoclonus jerk bilaterally, but the predominant action is
in the axial muscles.33 These jerks can occur
In most cases, the myoclonus EMG discharge
from rest and/or activated by stimuli such as
duration is 50–200 ms. Longer discharge dura-
touch, deep tendon reflex, or muscle stretch.
tion values have been described, especially in
Single or repetitive jerks may occur. The
those cases with dystonia. The EMG patterns
EMG discharge lasts typically 50–300 ms but
can be agonist-only or cocontracting agonist–
sometimes longer. Both reciprocal and cocon-
antagonist. Often, the discharges are irregular
tracting agonist–antagonist relationships have
with respect to amplitude, duration, and tim-
been observed. A major characteristic of the
ing between agonist and antagonist. The SEP
EMG discharges is a simultaneous bilateral
and long-latency EMG responses are normal.
rostral and caudal recruitment originating from
The routine EEG is normal. No focal EEG
the area of the spinal cord origin. The activa-
correlates have been reported, and most EEG–
tion speed of consecutive muscles is thought to
EMG back-averaging has been unrevealing.
be slower than for the corticospinal (pyramidal)
pathway and is likely propriospinal. No EEG
abnormalities in the routine recording or with
Reticular Reflex Myoclonus back-averaging have been reported. SEPs are
This myoclonus appears clinically as general- normal.
ized jerks that occur at rest and are stimulus
sensitive. The posthypoxic state and uremia are
SEGMENTAL
known causes. The EMG discharge duration
is less than 50 ms.31 A major characteristic of Segmental myoclonus has its generator at a
the EMG discharges is a simultaneous bilateral particular segment or contiguous segments of
rostral and caudal recruitment that originates the brain stem and/or spinal cord. This seg-
from the area of the medullary reticular for- mental generator produces movements at a
mation. Axial and proximal limb muscles are particular segment or contiguous segments.
primarily involved. The EEG may show that Palatal myoclonus is the most common type
epileptiform changes are associated, but they of segmental myoclonus. The EMG usually
follow the first EMG activation, vary from jerk shows synchronous activation of the affected
to jerk, and do not show a time-locked relation- muscles. The typical frequency is in the
ship to the myoclonus. The SEP is normal. range of 0.5–3 Hz and the typical EMG
discharge duration varies widely between may be technically challenging or inadequate,

50 and 500 ms. The EEG and SEP are absence of the BP should not be used to
normal. indicate that voluntary mechanisms have been
ruled out.
PERIPHERAL
Peripheral myoclonus refers to myoclonic jerks
that are driven from a peripheral site. The
STARTLE DISORDERS
best-documented example is hemifacial spasm.
The startle reflex is a whole body jerk that
Such EMG discharges are characterized by
commonly occurs in response to sudden unex-
marked duration variability from discharge to
pected noise or touch. Characteristic EMG
discharge. The appearance and timing of EMG
onset latencies to loud noise are well defined,
discharges, which are supplied by the same
with the orbicularis oculi invariably leading
nerve, are similar. In peripheral myoclonus,
activation at 30–40 ms and the sternocleido-
the spectrum of EMG discharge duration
mastoid following at 55–85 ms. Limb muscles
may merge continuously with those EMG dis-
are less consistently active, with the biceps
charges that are responsible for movements
activated at 85–100 ms and leg muscles at
that are longer lasting.
100–140 ms.36 Burst durations range from 50 to
400 ms. The reflex habituates rapidly. As a nor-
mal phenomenon, startle represents another
PSYCHOGENIC JERKS form of physiologic myoclonus.
Exaggerated startle has numerous causes,
Jerks that arise from a psychogenic mech- including inflammatory brain stem lesions,
anism may appear quick enough to over- anoxic injuries, psychiatric illnesses, and drug
lap with myoclonus. Some authors would intoxication. Hereditary hyperekplexia is an
maintain that myoclonus only pertains to autosomal dominant condition characterized
involuntary movements. Nevertheless, the by exaggerated startle to unexpected stimuli.
term psychogenic myoclonus has gained some A major form consisting of exaggerated startle
acceptance. Clinical neurophysiology may offer followed by prolonged stiffness has been linked
the clinician some assistance in evaluating to a point mutation in the gene encoding the α1
these patients, but one must exercise cau- subunit of the glycine receptor.37 A minor form,
tion when interpreting the results. It is con- consisting of exaggerated startle alone, has no
sidered unwise to use such techniques alone recognized cause. The audiogenic myoclonic
to “prove” or “rule out” a psychogenic basis jerks in hyperekplexia clearly correspond to the
for myoclonus. Results must always be used startle pattern described previously. However,
in their clinical context and the possible the startle reflex is increased in magnitude and
pitfalls should be realized. A psychogenic is poorly habituating in this disorder.
myoclonus EMG discharge may show the
known voluntary reciprocal biphasic/triphasic
agonist–antagonist/agonist pattern with the
duration of each burst being 50–100 ms. PERIODIC LIMB MOVEMENTS
Longer EMG discharges are possible in psy- OF SLEEP
chogenic myoclonus and the discharge pat-
tern may have a nonstereotyped appearance. Periodic jerks of the legs may interrupt sleep
Stimulus-evoked jerks or jumps with a mean and cause insomnia or excessive daytime som-
latency in excess of 100 ms suggest voluntary or nolence. Such periodic limb movements of
psychogenic jerks.34 The bereitschaftspotential sleep commonly accompany restless legs syn-
(BP) is a back-averaged negative EEG cortical drome.38 They may also appear after spinal
potential that occurs before self-paced, volun- cord trauma or vascular injury, implicating
tary phasic movements. Terada et al. found damage to descending inhibitory pathways.
the presence of a BP preceding psychogenic During the transition between drowsiness and
myoclonus to be supportive evidence for a psy- light sleep, the movements begin their cyclic
chogenic etiology.35 However, because record- occurrence, with an average period lasting
ing movements from a psychogenic individual between 30 and 45 seconds. These movements
do not resemble myoclonus; thus, the previous the duration varies widely between 50 and
designation of nocturnal myoclonus has been 500 ms; the EEG and SEP are normal.
abandoned. The surface EMG pattern varies. • The startle reflex demonstrates character-
Most often, the burst durations are longer istic EMG onset latencies to loud noise
than 500 ms. The earliest and most actively with the orbicularis oculi invariably lead-
involved muscle is often the anterior tibial ing activation at 30–40 ms and the ster-
muscle. Although the jerks may appear unilat- nocleidomastoid following at 55–85 ms.
eral, bilateral asynchronous EMG activation is Limb muscles are less consistently active,
the common occurrence. with the biceps activated at 85–100 ms
and leg muscles at 100–140 ms.36 Burst
Key Points durations range from 50 to 400 ms.
• In periodic limb movements of sleep, the
• Useful neurophysiological techniques burst durations are longer than 500 ms.
for examining myoclonus include EEG,
EMG, EEG–EMG polygraphy with back-
averaging, evoked potentials, and long-
latency EMG reflexes.
DYSTONIA
• Main physiological classification cate-
Dystonia is a syndrome of involuntary sus-
gories for myoclonus are cortical, cortical–
tained muscle contractions that produce abnor-
subcortical, subcortical, segmental, and
mal postures, twisting, and repetitive move-
peripheral.
• Cortical myoclonus is typically multifo- ments. It may be focal or generalized. The
most common focal dystonia is cervical dysto-
cal, but may be focal, segmental, or
nia or torticollis. Blepharospasm, oromandibu-
generalized.
• Cortical reflex myoclonus is defined lar dystonia, and writers’ or occupational
cramps are other common focal dystonias.
by the demonstration of a focal time-
Generalized dystonia is usually a manifesta-
locked cortical transient that precedes the
tion of hereditary torsion dystonia. Generally,
myoclonus by a short latency (<40 ms
neurophysiologic studies are most helpful in
for arm) in association with evidence for
evaluating the focal dystonias, sometimes as a
exaggerated reflex cortical phenomena,
prelude to therapeutic injections of botulinum
including enlarged cortical SEP waves,
toxin.
reflex-induced myoclonus, or enhanced
long-latency EMG responses to electrical
nerve stimulation.
• Enlargement of the cortical SEP P25–N33 Recording Techniques
parietal wave from median nerve stimu-
lation is important evidence for cortical The physiologic hallmark of dystonia is intense
reflex myoclonus. cocontraction of agonist and antagonist mus-
• In most cases of cortical reflex myoclonus, cles, producing a marked increase in stiffness
abnormal long-latency EMG discharges across the joint and abnormal posturing. Thus,
are elicited by median nerve stimulation, muscles acting across the postured joint should
which shows EMG discharges at 50 ms be studied to look for simultaneous interfer-
latency or greater (range 40–60 ms) from ence patterns. Intramuscular electrodes often
the stimulus artifact trigger mark. are needed to ensure selective recordings.
• In myoclonus of corticobasal degener- Whereas cocontraction is not specific for dys-
ation, the EMG discharge duration is tonia, it does rule out joint contractures or
25–50 ms with a cocontraction pattern hysteria, in which abnormal limb posture is
and the EEG shows no correlate to the unaccompanied by EMG activity. The EMG
myoclonus EMG discharges. discharges may be tonic or occur in a repetitive
• In essential myoclonus (subcortical– rhythmic or arrhythmic pattern called phasic
suprasegmental myoclonus), the EMG dystonia. This pattern may distinguish itself
discharge duration is 50–200 ms. from tremor by the variability of the burst
• In segmental myoclonus, typical discharge durations and the frequent intrusion of tonic
frequency is in the range of 0.5–3 Hz and dystonia.
Abnormal Patterns reflects dystonic activity in the lateral pterygoid

and digastric muscles. Jaw-closing dystonia
TORTICOLLIS reflects activity in the temporalis, masseter, and
medial pterygoid muscles.
Deuschl and associates39 have described the
patterns of EMG discharge in spasmodic torti-
collis. With rotational torticollis, the contralat- SPASMODIC DYSPHONIA
eral sternocleidomastoid and the ipsilateral
splenius capitus are most often active. In retro- Spasmodic dysphonia may be of the adduc-
collis, all posterior neck muscles are active, tor or abductor type. Patients with adductor
and in laterocollis, the ipsilateral splenius capi- spastic dysphonia present with a strained or
tus and sternocleidomastoid muscles are active. tremulous voice caused by dystonia of the thy-
We have found similar patterns, but variations roarytenoid muscles. These muscles can be
in a particular pattern of muscle activity are recorded by needle examination performed
common. For this reason, performing multi- percutaneously or by direct laryngoscopy. With
channel EMG recording with intramuscular either method, the participation of an experi-
electrodes in some patients before injecting enced otorhinolaryngologist is advised. In con-
botulinum toxin may be useful (Fig. 33–5).40 trast, abductor spastic dysphonia is manifested
as a whispering voice and the dystonic activ-
ity is in the posterior cricoarytenoid muscles,
OROMANDIBULAR DYSTONIA
which are inaccessible to routine examination.
In oromandibular dystonia, patients may be Botulinum toxin injection is useful for these
unable to eat or speak because of abnormal disorders and simultaneous needle EMG is
jaw posturing. Jaw-opening dystonia frequently recommended.
Left trapezius
Left spienius capitus
Left cervical paraspinals
Right sternocleidomastoid
1 mV
1 second
Figure 33–5. Dystonia. Electromyographic activity recorded with intramuscular electrodes in a patient with spasmodic
torticollis. Both tonic and irregular phasic EMG bursts are present.
WRITERS’ AND OCCUPATIONAL this movement is performed as fast as pos-

CRAMPS sible, the triphasic pattern appears and its
configuration can be examined. Abnormali-
In various circumstances, repetitive skilled ties have been reported in various diseases.
motions may become complicated by painful The ballistic movements in Parkinson’s dis-
and disabling dystonia of the hand or wrist. ease are characterized by low amplitude of
There is no typical posture, and a com- the initial agonist burst. This results in a
bination of flexion and extension dystonia small-amplitude movement that the patient
may occur. Intramuscular EMG recordings compensates for by making sequential small
show the individual pattern of phasic and/or triphasic bursts. The movements of cere-
tonic spasms in multiple cocontracting forearm bellar hypermetria have been ascribed to a
muscles. delayed onset of the antagonist burst41, 42 or
a normal onset but slow rise of antagonist
Key Points activity.43 With pyramidal lesions, the first ago-
• Cocontraction of agonist and antagonist nist or antagonist burst is prolonged. Finally,
muscle is characteristic but not specific for in patients with athetosis caused by cere-
dystonia. bral palsy, excessive activity occurs in mus-
• Intramuscular EMG may be useful in per- cles not normally involved in the main action,
forming botulinum toxin injections. and agonist and antagonist muscles often
cocontract.
TICS, CHOREA, AND ATHETOSIS

SUMMARY
Surface EMG recording is of limited useful-
ness in the evaluation of tics, chorea, and Surface EMG, EEG, and elicited response
athetosis. Although these involuntary move- results provide a simple and noninvasive means
ments are clinically distinct, the surface EMG of studying movement disorders. These tech-
patterns are nonspecific and may appear niques are particularly helpful in classifying
similar in all three. Burst durations can be involuntary movements such as tremor and
100–300 ms and can resemble reflex, tonic, or myoclonus. In addition, EMG can assist with
ballistic patterns. designing and performing botulinum toxin
injections.
Key Points
• Surface EMG is nonspecific in tics,
chorea, and athetosis. REFERENCES
• Burst durations in tics, chorea, and atheto-
sis range from 100 to 300 ms. 1. Hallett, M., B. T. Shahani, and R. R. Young. 1975.
EMG analysis of stereotyped voluntary movements in
man. Journal of Neurology, Neurosurgery, and Psychi-
atry 38:1154–62.
VOLUNTARY MOVEMENT 2. Milanov, I. 2001. Electromyographic differentiation of
ABNORMALITIES tremors. Clinical Neurophysiology 112:1626–32.
3. Deuschl, G., P. Bain, and M. Brin. 1998. Consen-
sus statement of the Movement Disorder Society on
Techniques are available for recording volun- Tremor. Ad Hoc Scientific Committee. Movement
tary movement by surface EMG analysis. True Disorders 13(Suppl 3):2–23.
quantitative evaluation in this realm requires 4. Breit, S., S. Spieker, J. B. Schulz, and T. Gasser. 2008.
additional equipment to record movement Long-term EMG recordings differentiate between
parkinsonian and essential tremor. Journal of Neurol-
position or velocity. Qualitative evaluation is ogy 255(1):103–11.
helpful occasionally but must be interpreted 5. Deuschl, G., P. Krack, M. Lauk, and J. Timmer. 1996.
with caution. Such evaluation can be per- Clinical neurophysiology of tremor. Journal of Clinical
formed by having the patient make a short Neurophysiology 13:110–21.
6. Hallett, M. 1998. Overview of human tremor physiol-
ballistic movement such as elbow flexion while ogy. Movement Disorders 13(Suppl 3):43–8.
recordings are made from agonist and antag- 7. Elble, R. J. 2000. Essential tremor frequency
onist muscles. Under most conditions when decreases with time. Neurology 55:1547–51.
8. Sabra, A. F., and M. Hallett. 1984. Action tremor with 25. Wilkins, D. E., M. Hallett, A. Berardelli, T. Walshe,
alternating activity in antagonist muscles. Neurology and N. Alvarez. 1984. Physiologic analysis of
34:151–6. the myoclonus of Alzheimer’s disease. Neurology
9. Milanov, I. 2000. Clinical and electromyographic 34:898–903.
examinations of patients with essential tremor. The 26. Shibasaki, H., S. Motomura, Y. Yamashita, H. Shii, and
Canadian Journal of Neurological Sciences 27: Y. Kuroiwa. 1981. Periodic synchronous discharge and
65–70. myoclonus in Creutzfeldt–Jakob disease: Diagnostic
10. Brennan, K. C., E. C. Jurewicz, B. Ford, S. L. Pullman, application of jerk-locked averaging method. Annals of
and E. D. Louis. 2002. Is essential tremor pre- Neurology 9:150–6.
dominantly a kinetic or a postural tremor? A clinical 27. Oga, T., A. Ikeda, T. Nagamine, et al. 2000. Impli-
and electrophysiological study. Movement Disorders cation of sensorimotor integration in the generation
17(2):313–16. of periodic dystonic myoclonus in subacute scle-
11. Vaillancourt, D. E., A. B. Slifkin, and K. M. Newell. rosing panencephalitis (SSPE). Movement Disorders
2001. Regularity of force tremor in Parkinson’s disease. 15(6):1173–83.
Clinical Neurophysiology 112:1594–603. 28. Thompson, P. D., B. L. Day, J. C. Rothwell, P. Brown,
12. Heilman, K. M. 1984. Orthostatic tremor. Archives of T. C. Britton, and C. D. Marsden. 1994. The
Neurology 41:880–1. myoclonus in corticobasal degeneration. Evidence
13. Marsden, C. D., M. Hallett, and S. Fahn. 1982. The for two forms of cortical reflex myoclonus. Brain
nosology and pathophysiology of myoclonus. In Move- 117:1197–207.
ment disorders, ed. C. D. Marsden, and S. Fahn, 29. Caviness, J. N., C. H. Adler, T. G. Beach, K. L.
196–248. London: Butterworths. Wetjen, and R. J. Caselli. 2002. Small-amplitude
14. Caviness, J. N., and P. Brown. 2004. Myoclonus: Cur- cortical myoclonus in Parkinson’s disease: Physiology
rent concepts and recent advances. Lancet Neurology and clinical observations. Movement Disorders 17(4):
3(10):598–607. 657–62
15. Shibasaki, H., and M. Hallett. 2005. Electrophys- 30. Ugawa, Y., T. Shimpo, and T. Mannen. 1989. Phys-
iological studies of myoclonus. Muscle & Nerve iological analysis of asterixis: Silent period locked
31(2):157–74. averaging. Journal of Neurology, Neurosurgery, and
16. Caviness, J. N. 2003. Clinical neurophysiology of Psychiatry 52:89–93.
myoclonus. Movement disorders. Handbook of clinical 31. Hallett, M., D. Chadwick, J. Adam, and C. D.
neurophysiology, ed. M. Hallett, Vol. 1, 521–48. Marsden. 1977. Reticular reflex myoclonus: A phys-
17. Shibasaki, H. 2000. Electrophysiologic studies of iological type of human post-hypoxic myoclonus.
myoclonus. AAEE minimonograph #30. Muscle & Journal of Neurology, Neurosurgery, and Psychiatry
Nerve 23:321–35. 40:253–64.
18. Hitomi, T., A. Ikeda, R. Matsumoto, et al. 2006. Gen- 32. Gwinn, K. A., and J. N. Caviness. 1997. Electro-
erators and temporal succession of giant somatosen- physiological observations in idiopathic Opsoclonus–
sory evoked potentials in cortical reflex myoclonus: Myoclonus syndrome. Movement Disorders 12(3):
Epicortical recording from sensorimotor cortex. Clini- 438–42.
cal Neurophysiology 117(7):1481–6. 33. Brown, P., P. D. Thompson, J. C. Rothwell, B. L.
19. Shibasaki, H., Y. Yamashita, and S. Tsuji. 1977. Day, and C. D. Marsden. 1991. Axial myoclonus of
Somatosensory evoked potentials. Diagnostic crite- propriospinal origin. Brain 114:197–214.
ria and abnormalities in cerebral lesions. Journal of 34. Brown, P., and P. D. Thompson. 2001. Electro-
Neurological Sciences 34:427–39. physiological aids to the diagnosis of psychogenic
20. Ugawa, Y., K. Genba, T. Shimpo, and T. Mannen. 1991. jerks, spasms, and tremor. Movement Disorders 16:
Somatosensory evoked potential recovery (SEP-R) in 595–9.
myoclonic patients. Electroencephalography and Clin- 35. Terada, K., A. Ideda, P. C. Van Ness, et al. 1995. Pres-
ical Neurophysiology 80:21–5. ence of Bereitschaftspotential preceding psychogenic
21. Caviness, J. N., and M. Kurth. 1997. Cortical myoclonus: Clinical application of jerk-locked back
myoclonus in Huntington’s disease associated with an averaging. Journal of Neurology, Neurosurgery, and
enlarged somatosensory evoked potential. Movement Psychiatry 58:745–7.
Disorders 12(6):1046–51. 36. Wilkins, D. E., M. Hallett, and M. M. Wess. 1986.
22. Brown, P., B. L. Day, J. C. Rothwell, P. D. Audiogenic startle reflex of man and its relationship to
Thompson, and C. D. Marsden. 1991. Intrahemi- startle syndromes. A review. Brain 109:561–73.
spheric and Interhemispheric spread of cerebral cor- 37. Tijssen, M. A., R. Shiang, J. van Deutekom, et al.
tical myoclonic activity and its relevance to epilepsy. 1995. Molecular genetic reevaluation of the Dutch
Brain 114:2333–51. hyperekplexia family. Archives of Neurology 52:
23. Ikeda, A., R. Kakigi, N. Funai, R. Neshige, Y. Kuroda, 578–82.
and H. Shibasaki. 1990. Cortical tremor: A variant of 38. Coleman, R. M., C. P. Pollak, and E. D. Weitzman.
cortical reflex myoclonus. Neurology 40:1561–5. 1980. Periodic movements in sleep (nocturnal
24. Caviness, J. N., C. H. Adler, T. Beach, K. Wetjen, myoclonus): Relation to sleep disorders. Annals of
and R. J Caselli. 2002. Myoclonus in Lewy body Neurology 8:416–21.
disorders. In Myoclonus and paroxysmal dyskinesias, 39. Deuschl, G., F. Heinen, B. Kleedorfer, M. Wagner,
ed. S. Fahn, M. Hallet, S. Frucht, and D. Truong, C. H. Lucking, and W. Poewe. 1992. Clinical and
Advances in Neurology, Vol. 89, 23–30. New York: polymyographic investigation of spasmodic torticollis.
Lippincott Williams & Wilkins. Journal of Neurology 239:9–15.
40. Van Gerpen, J. A., J. Y. Matsumoto, J. E. Ahlskog, 42. Gilman, S., J. Bloedel, and R. Lechtenberg. 1981.
D. M. Maraganore, and P. G. McManis. 2000. Util- Disorders of the cerebellum. Contemporary neu-
ity of an EMG mapping study in treating cervical rology series, Vol. 21. Philadelphia: FA Davis
dystonia. Muscle & Nerve 23:1752–6. Company.
41. Flament, D., and J. Hore. 1986. Movement and 43. Manto, M., E. Godaux, J. Jacquy, and J. Hildebrand.
electromyographic disorders associated with cere- 1996. Cerebellar hypermetria associated with a selec-
bellar dysmetria. Journal of Neurophysiology 55: tive decrease in the rate of rise of antagonist activity.
1221–33. Annals of Neurology 39:271–4.
Chapter 34
Vertigo and Balance

David A. Zapala and Robert H. Brey
INTRODUCTION Vestibular Evoked Myogenic Potentials

Functional Anatomy and Physiology Computerized Dynamic
Clinical Features of Vestibulopathy Posturography
and Central Compensation
CLINICAL APPLICATIONS OF
Laboratory Vestibular Test Methods
VESTIBULAR TESTING:
LABORATORY EXAMINATION: ASSESSING SENSORINEURAL
VOR-BASED MEASURES SYNDROMES OF THE LABYRINTH
Recording Method Superior Nerve Syndrome
Preparation for Testing Superior SCC Dehiscence
Videonystagmography Test Battery Basement Syndrome
Computerized Rotary Chair Tests Posterior Syndrome
Subjective Visual Vertical (SVV) Split Syndrome
Assessment Global Syndrome
LABORATORY EXAMINATION: VESTIBULAR REHABILITATION
VSR-BASED MEASURES SUMMARY
INTRODUCTION These test methods do not stand alone. To

interpret test results meaningfully an appre-
Complaints of dizziness and imbalance can ciation of how normal central nervous system
be caused by a bewildering array of disor- (CNS) compensation and abnormal vestibu-
ders. Some disorders are easily recognized lar reflex function interact during the recovery
while others can remain perplexingly obscure. process is required. Only then can correlations
This chapter presents several electrophysio- between patient presentation and vestibular
logic methods that might be used to investigate test results be placed in a meaningful context.
vestibular function. These methods include This chapter summarizes how electrophysio-
videonystagmography (VNG) or electronystag- logic vestibular test information can establish
mography (ENG), rotational tests, subjective the site of lesion and reflect central compensa-
visual vertical (SVV) measurements, vestibu- tion status. Relationships between test results
lar evoked myogenic potentials (VEMPs), and and nerve-based labyrinthine syndromes are
computerized dynamic posturography (CDP). also presented.
575
A careful history and physical examination often, these tests confirm or challenge clini-
are by far the most important tools in the cal hypotheses developed through the history
assessment of vertigo and imbalance.1–4 A clear and physical examination. There are several
understanding of the patient’s complaint, as factors that complicate the interpretation of
well as mapping the onset and progression of electrophysiologic measures of vestibular func-
symptoms, will go a long way in converging tion. First, some forms of vestibulopathy pro-
on a diagnosis. However, even in experienced duce fluctuating output from the vestibular end
hands there are challenges. As humans, we per- organ. Examples include Meniere’s syndrome,
ceive and navigate within our environments migraine-associated vertigo, or perilymph fis-
with innate ease. This ease masks the complex- tula. Depending upon the presence and sever-
ity by which visual, somatosensory, and vestibu- ity of symptoms at the time of study, an
lar sensory information are integrated into an impaired ear may produce normal test results.
understanding of our location and movement Second, current test methods measure reflexes
in the environment. In addition, exquisite cas- from two or possibly three of the five vestibu-
cades of reflexive, semiautomated, and voli- lar receptors in each labyrinth. So, a normal
tionally refined movements support bipedal study cannot exclude a fluctuating or obscure
stance and locomotion. Most of these processes vestibulopathy.
require little conscious attention or effort. As Vestibular assessment methods also differ
a result, patients are often only able to pro- from other electrophysiologic methods in that
vide vague, nonspecific descriptions of what vestibular tests measure reflexes that are eas-
they mean by “dizziness” or “imbalance” when ily modified by higher forms of behavior. For
things go wrong. example, vestibular-induced eye movements
To simplify the differential diagnosis, most can be suppressed volitionally. Abnormal pos-
clinicians will use a set of heuristic rules to tural sway from reduced vestibulo-spinal tone
organize patient descriptions of dizziness or is easily modified with attention to visual and
imbalance.2, 5 Complaints suggestive of ver- proprioceptive feedback. Reliable assessment
tigo (circular vection of the self or visual thus requires careful attention and control of
surround) are taken to imply an otogenic psychophysical and cognitive variables to guard
vestibular problem. Complaints of disequilib- against inadvertent suppression of vestibular-
rium (unsteadiness when standing or walking) induced reflexes. Therefore, evaluator skill is
suggest a neuromuscular cause. Lightheaded- another important factor in test accuracy.
ness suggests a neurovascular problem. Not Finally, the relationship between vestibu-
all complaints can be shoehorned into these lar end organ output and central vestibular
categories, and even when there is a rea- processing is dynamic. With time, changes in
sonable linguistic match, classification errors end organ output result in adaptive changes in
occur. For example, pilots, sailors, and gym- the way aberrant vestibular signals are inter-
nasts may not report vertiginous sensations in preted within the central vestibular system. As
the face of obvious vestibular disease due to a result, some abnormal vestibular reflexes are
their prior training suppressing these sensa- only observable when lesions are acute, while
tions. Further, some vestibular disorders do other reflexes may remain persistently abnor-
not induce sensations of vertigo, nausea, or mal. The ability to correlate the onset and
imbalance. Examples would include deficits progression of symptoms with the pattern of
from slow-growing vestibular schwannomas or abnormal test results can give the examiner
disorders of the vestibular maculae. In cases insight into the underlying disease process and
where vertigo is not reported and codeveloping the state of central vestibular system compen-
otologic or neurologic symptoms are lacking, sation. This is an important factor in planning
heuristically based classification schemes may medical management and vestibular rehabilita-
misdirect subsequent diagnostic inquires. tion services.
Electrophysiologic measures of vestibular In the primary care setting, vestibular test-
function augment the traditional history and ing may not be routinely warranted. Most
physical examination.6, 7 They can identify and complaints of dizziness and imbalance will
quantify reduced vestibular output, leading have well-recognized causes and most forms
to assurance that a vestibular cause under- of vestibular dizziness will be self-limiting in
pins the patient’s complaints. However, they this setting. On the other hand, in specialty set-
seldom yield a diagnosis in isolation. More tings, where patients are sent with persistent
Vertigo and Balance 577
problems after failing primary medical care, types of sensory organs within the labyrinth:
special tests may play an invaluable role. The the semicircular canals (SCCs) and the otolith
value of recognizing the obscure vestibulopa- organs, the saccule and utricle (Fig. 34–1).
thy is that, with a few exceptions, end organ– These organs are sensitive to three differ-
based vestibular deficits are not life threat- ent types of head movements: angular head
ening and can be managed with vestibular movements, head tilts, and linear translations.
rehabilitation, medication, and rarely surgery. In response to these head movements, the
Even when treatment cannot restore balance vestibular system drives two primary reflexes:
function to pre-insult levels, the psychologi- the vestibulo-ocular reflex (VOR) and the
cal benefits of a comprehensive evaluation are vestibulo-spinal reflex (VSR).
noteworthy.
ANGULAR HEAD MOVEMENTS
Purpose and Role of Vestibular Testing
AND THE VOR
• To identify and quantify reduced vestibu-
Angular head movements occur when the head
lar output.
• Used to provide insight into the underly- is turned about the pivot point of the cervi-
cal spine and are predominantly transduced by
ing disease process and the state of central
the SCCs. Endolymph within each SCC will
vestibular system compensation.
demonstrate the characteristic of inertia when
the head turns in the plane of a SCC. That
is, the endolymph will lag behind the move-
Functional Anatomy and ment of the skull during these movements,
Physiology causing the cupula within the SCC ampulla to
bend. The magnitude of cupular deflection is
The anatomy and physiology of the vestibu- determined by the magnitude of acceleration
lar mechanism and certain functional relation- and the degree to which the head movement
ships are important to keep in mind when is aligned with the plane of the stimulated
evaluating the vestibular system. There are two canal.
Figure 34–1. Anatomy of the membranous labyrinth and vestibulocochlear nerve branches. Note that the vestibular
(Scarpa’s) ganglion has superior and inferior portions, serving the superior and inferior vestibular nerve branches.
In humans, the VOR is strongly driven by variation of reciprocal innervation. Each SCC
angular head movements. The primary goal of is coplaner with a SCC on the contralateral
the VOR is to maintain a visual targets’ position side. The horizontal canals are coplaner. The
on the fovea despite head movement. A sec- right anterior and the left posterior canals are
ondary goal is to maintain the orientation of the coplaner, and the left anterior and right pos-
horizontal meridians of the eye with the hori- terior canals are coplaner. An angular head
zon. There are three SCCs oriented in three movement to the right (in any canal plane) will
nearly orthogonal planes in each labyrinth. increase the firing rate of the right vestibu-
They are positioned to respond to any head lar nerve and decrease the firing rate of the
movement about the pivot point of the cervi- left vestibular nerve. That is, for any angular
cal spine. They are also nearly coplaner with head movement, the leading canal will always
insertion planes of the ocular motor muscles, excite the eighth nerve afferents while the
facilitating the rapid generation of compen- lagging canal will always inhibit the eighth
satory VOR eye movements. Stimulation of nerve afferents. From the perspective of the
a single SCC will reflexively provoke an eye central vestibular system, differences in out-
movement in the plane of that SCC. For exam- put of coplaner ampullary nerve afferents are
ple, when the head is turned horizontally to interpreted as angular head movements in the
the right, the VOR drives the eyes horizon- direction of the leading ear.
tally to the left to maintain visual target posi- One consequence of the resting discharge
tion on the fovea. Similarly, when the head is rate and reciprocal innervation organizational
rolled clockwise so that the right ear is moved themes within the vestibular system is that
toward the right shoulder, the eyes will roll VOR-induced eye movements are driven by
counterclockwise in the orbit to maintain the changes in the output of both ears when
relationship between the horizon and the hor- head movements are slow. However, when
izontal meridian of the retina. This would be angular head accelerations reach a certain crit-
described as a torsional vestibular-induced eye ical magnitude, the nerve firing rate of the
movement. It is important to remember that lagging ear will drop to zero spikes per sec-
VOR-mediated eye movements are reflected ond. Beyond this critical acceleration point,
in the slow phase of vestibular-induced nys- the leading ear alone controls the velocity
tagmus. However, by convention, the direc- of vestibular-induced eye movements. This
tion of nystagmus is described by the fast characteristic, known as Ewald’s second law,
phase. underpins the head thrust test (a bedside
Sensory transduction within the vestibular test of SCC function) and is important in
labyrinth is accomplished by specialized hair understanding some of the rotational step test
cell systems. Underneath each cupula resides results.
the crista, a bed of hair cells with the cilia
embedded into the cupula. Within the otolithic Key Points
organs, the maculae also contain hair cells
beds. Hair cell systems have certain charac- • Angular head movements occur when the
teristics that will influence functional perfor- head is turned about the pivot point of the
mance. First, hair cell systems have a resting cervical spine.
discharge rate. Resting discharge rates allow • Three SCCs are positioned to respond to
hair cell systems to encode periodic or har- angular head movements.
monic movements of the hair cell cilia accu- • The VOR is driven primarily by angular
rately. Deflecting the cilia in one direction head movements, (SCC-mediated).
increases the nerve firing rate, while deflecting • Stimulation of one SCC will provoke eye
the cilia in the opposite direction decreases the movements in the plane of that canal.
firing rate. In this way, both accelerating and • Specialized hair cells coupled to each SCC
decelerating angular head movements (that cupula transduce angular head move-
would bend the cupula in different directions) ments. The resting discharge rate of each
can be encoded by changes in discharge firing hair cell allows them to encode periodic
rates of each ampullary nerve. movements.
An additional characteristic of vestibular • Deflection of the cilia of the hair cells
hair cell systems is that they demonstrate a changes the firing rate in relation to
acceleration or deceleration of the head • Nerve firing rates asymmetries induced by

movements. head tilts drive changes in the VSR and
• Reciprocal innervation of the vestibular produce subtle changes in the VOR.
hair cell systems causes a decrease in the
firing rate of one vestibular nerve when LINEAR TRANSLATIONS AND
there is an increase in the firing rate of the ACCELERATIONS
contralateral nerve.
• When head movements are slow (low The ear responds to linear translations or
acceleration), the VOR is driven by accelerations as well. Otoconia, because of
changes in the output of both ears; when their specific gravity, will also demonstrate the
movements are fast (high acceleration), characteristic of inertia. As a result, linear head
only the leading ear controls the VOR (this movements will produce a shear force across
is known as Ewald’s second law). the otolithic organs, depending on the relation-
ship between the planes of the maculae and
the direction of movement. In contrast to the
HEAD TILTS nerve firing rate pattern generated by head
tilts (where head tilts produce an asymmetry
The vestibular labyrinth will also respond to in nerve output), linear accelerations may pro-
tilting of the head away from gravitational ver- voke the same change in firing rate between
tical (head tilts). The otolith organs serve this sides. This signal is important in resolving
function. Otoconia are attached to a gelatinous ambiguous visual situations where it is difficult
matrix affixed to the macula of each otolithic to determine if the head or the visual surround
organ. Otoconia have a high-specific gravis moving. An example is the false sensation
ity relative to endolymph and, consequently, of movement experienced by a person sitting
respond to the pull of gravity. Hair cell cilia in in an automobile when a large vehicle parked
the macula are loosely attached to the gelati- alongside begins slowly to pull forward. The
nous matrix. As a result, changes in head atti- visual input to the person is consistent with
tude will produce a change in the gravity vector the sensation of the car rolling backward; the
acting on the otoconia. This, in turn, will pro- reflexive action is to step on the brakes. The
duce a change in the shear force across the absence of a vestibular signal consistent with
macula, changing the resting discharge rate of linear acceleration helps the CNS to resolve
the associated nerve afferents. Differences in the sensory conflict by determining that it was
firing rates between ears are interpreted as a the larger vehicle that was moving and that the
head tilt. Head tilts about the horizontal plane visual information was incorrect.
would primarily be encoded by the utricle. Vertical linear translations or accelerations
These asymmetries drive changes in VSR stim- are encoded primarily by the saccule and
ulation to the neck, spine, and large antigravity are important in navigating steps and possi-
muscles. They also induce subtle changes in bly evoking protective reflexes when falling. A
the VOR (see Subjective Visual Vertical Test clinical test based on this reflex, the VEMP,
section below). has proven particularly valuable in the clinical
setting.
Finally, there are a few anatomical rela-
Key Points
tionships that are helpful to remember when
• Head tilts are movements away from grav- interpreting vestibular test results. The eighth
itational vertical and are served by the nerve separates into three main branches as
otolith organs. it courses distally along the internal auditory
• Changes in head attitude produce changes canal. The branches are the superior vestibular,
in the gravity effect acting on the otoconia, inferior vestibular, and cochlear eighth nerve
and change the resting discharge rate of branches. The superior vestibular nerve fur-
nerve afferents. ther branches to innervate the horizontal or
• Difference in firing rates between ears are “lateral” SCC ampulla, the anterior or “supe-
interpreted as head tilt. rior” SCC ampulla, the utricle, and a small part
• Head tilts about the horizontal plane are of the saccule (Voit’s branch). This can eas-
encoded by the utricle. ily be remembered by the pneumonic “LSU
on top” (the Lateral and Superior semicircular It may be helpful to illustrate how compen-
canal ampullae and Utricle are innervated satory influences impact vestibular test results.
by the “on top” superior vestibular nerve). A prototypical example of an acute vestibu-
The inferior vestibular nerve branch innervates lopathy is superior nerve neurolabyrinthitis. In
the posterior SCC ampulla and the major- this condition, the patient experiences a sud-
ity of the saccule. It also carries the efferent den onset of vertigo and vestibular ataxia that
cochlear bundle. The blood supply to the lasts for over one day, with symptoms gradu-
vestibular labyrinth follows the innervation ally diminishing over the course of several days
pattern. The anterior vestibular artery sepa- or weeks. There are no accompanying otologic
rates from the labyrinthine artery and supplies or neurologic symptoms. There is a sudden
“LSU.” After separating from the ante- diminishment of superior vestibular nerve rest-
rior vestibular artery, the labyrinthine artery ing firing rates associated with the lateral and
becomes the common cochlear artery. The superior SCCs (LSU distribution). This pro-
common cochlear artery further divides into duces an asymmetry between ears within the
the main cochlear artery and the vestibulo- central vestibular nuclei that is interpreted as
cochlear artery. The latter forms the posterior an angular head turn toward the intact ear.
vestibular artery that supplies the posterior Nystagmus, with the fast phase of the hori-
canal and saccule. zontal component beating toward the intact
ear, will be evident with the eyes opened and
Key Points fixated.
• Linear head movements produce similar In the setting of an acute unilateral loss
of vestibular tone, three cardinal behaviors
forces across the otolith organs on both
of vestibular-induced spontaneous nystagmus
sides.
• Vertical linear accelerations are encoded can be appreciated. First, the slow compo-
nent (slow phase) of the nystagmus follows the
primarily by the saccule.
• The superior vestibular nerve branch planes of the involved SCCs. In our example,
the observed nystagmus will be a combination
innervates the superior and horizontal
of a horizontal and a torsional nystagmus—
SCCs, the utricle, and a small part of the
reflecting the involved lateral (horizontal) and
saccule.
• The inferior vestibular nerve branch superior (anterior) canals. Further, the direc-
tion of the horizontal nystagmus component
innervates the posterior SCC and the
remains fixed regardless of eye position. This
larger portion of the saccule.
follows from the fact that the brain interprets
the asymmetry in vestibular output from the
ears as fixed angular rotation.
Clinical Features of The second cardinal behavior is that the
Vestibulopathy and Central observed spontaneous nystagmus will follow
Compensation Alexander’s Law. Alexander’s Law states that
when gaze is directed toward the nystagmus
The signs and symptoms of vestibulopathy will fast phase (in the direction of the intact or
depend on two general factors. First, there is “strong” ear), the horizontal nystagmus will
a relationship between the pattern of damage increase in magnitude. In contrast, when the
to sensory structures in the vestibular labyrinth fast phase is directed toward the weak ear,
and the abnormal reflex behaviors. Second, the nystagmus appears less intense. While a
the degree to which abnormal reflex behaviors physiological explanation of this phenomenon
manifest is influenced by the degree of central is beyond the scope of the present chap-
compensation. The central vestibular pathways ter, the concept can be intuitively appreciated
are by nature plastic and begin the process of by remembering that in normal individuals,
compensating for changes in vestibular tone end point nystagmus can be observed at the
within hours of an onset of acute vestibulopa- extreme limits of lateral gaze. This nystagmus
thy. Further, vestibular-driven reflexive behav- has a fast phase that beats toward the direction
iors are modifiable by higher levels of behav- of gaze. On extreme gaze to the right, there
ior. As a consequence, signs and symptoms of is a right-beating end point nystagmus. On
vestibulopathy evolve over time. extreme gaze to the left, there is a left-beating
nystagmus. This nystagmus is described as the is intact—an expected observation in the face
result of “leaky” gaze-holding circuits. When of a sudden unilateral loss of vestibular end
there are sudden acute changes in vestibular organ tone.
tone (for just about any reason), gaze-holding As hours go by following the onset of supe-
circuits become even more “leaky.” Physiologic rior vestibular nerve neurolabyrinthitis, the
end point nystagmus becomes more easily pro- vestibular cerebellum may “clamp” the out-
voked and interacts with the direction-fixed put of the vestibular system overall, in an
spontaneous vestibular nystagmus. When gaze effort to increase the influence of pursuit and
is directed toward the fast phase, the two nys- gaze-holding signals on eye position. Vestibu-
tagmus signals sum. When gaze is directed lar reflexes measured in this state may be
away from the fast phase, the two nystagmus bilaterally reduced and mask the underlying
signals cancel each other. asymmetry in output. As hours turn to days,
Gaze nystagmus that is direction fixed and spontaneous nystagmus observable with the
follows Alexander’s Law can be classified into eyes opened and fixed slowly diminishes. How-
3rd degree, 2nd degree, and 1st degree. A ever, nystagmus present with visual fixation
3rd degree Alexander’s nystagmus is present denied may persist for several months, often
in all positions of gaze and is greater when developing head position–dependent changes.
gaze is directed toward the unaffected ear. For example, the nystagmus may be present
As the CNS physiologically compensates for only when the head is turned to one side
the deficit caused by the peripheral lesion, or when lying down. This is termed a posi-
the nystagmus progresses to the 2nd degree tional nystagmus and its persistence relates
(i.e., nystagmus is only observed with gaze to the degree to which central vestibular cir-
away from the lesion and at midline) and cuits have reorganized following the initial end
finally to the 1st degree (i.e., nystagmus is only organ insult.
observed with gaze away from the side of the It is important to recognize that vestibu-
lesion). lar nystagmus may emanate from SCC or
Along these same lines, the torsional com- macular tone asymmetries. In general, cen-
ponent of the spontaneous nystagmus will be tral compensation changes will passively seek
more evident with the eyes directed away to diminish the effect of tonal asymmetries
from the involved ear, perpendicular to the when the head is at rest. Spontaneous nys-
plane of the involved vertical canal. So with tagmus, and to some extent positional nys-
the eyes directed toward the fast phase, and tagmus, will diminish accordingly (so-called
away from the involved ear, the nystagmus static compensation). However, the vestibular
will appear as a combination of horizon- system will also attempt to relearn how to
tal and torsional components, all in keeping interpret signals from the impaired vestibular
with Alexander’s Law and the planes of the end organ with head movement. This requires
involved SCCs. experience (so-called dynamic compensation).
The final cardinal behavior of vestibular- Movement induced sensory conflicts between
induced nystagmus is the effect of visual the impaired and the intact ear signals, as
fixation on nystagmus magnitude. Vestibular- well as conflicts among vestibular, visual, and
induced eye movements are easily modified somatosensory signals are important stimuli in
by higher level signals from the gaze-holding, the process of achieving dynamic compensa-
saccadic, and pursuit systems (among others). tion. Reorganizing neural networks to resolve
In the face of an acute unilateral vestibulopa- these conflicts is the basis of vestibular rehabil-
thy, the flocculus of the cerebellum can selec- itation.
tively increase the control pursuit and gaze VSRs follow a similar time course in the
holding circuits have on eye position, thereby recovery process. Initially, superior vestibular
suppressing the magnitude of spontaneous nys- nerve neurolabyrinthitis will provoke a ten-
tagmus with the eyes opened and fixed. How- dency to stand with the center of mass shifted
ever, when visual fixation is denied, vestibular toward the side of the involved ear. The
nystagmus increases in magnitude. So observ- patient may notice a tendency to veer to the
ing that spontaneous nystagmus suppresses involved side. This is thought to reflect the
with visual fixation assures the observer that acute loss of vestibular tone to the descend-
at least part of the central vestibular system ing vestibulo-spinal tracts. Over the course of
days, many of these symptoms subside as voli- Laboratory Vestibular Test

tional and then more automatic movement pat- Methods
terns emerge to compensate for the loss of
vestibular tone. There are five clinically available electrophys-
The process of completely adjusting to iologic measures that assess vestibular reflex
changes in vestibular end organ function function either directly or indirectly. Three
occurs over time. Patients may still complain measures, ENG (or the older VNG), rotary
of symptoms even while the overt signs of chair tests, and measures of SVV, focus on
vestibulopathy have receded. The presence VOR behaviors. The remaining two measures,
or absence of abnormal vestibular reflexes VEMPs and CDP, focus on behaviors associ-
will depend on time between the symptom ated with vestibulo-spinal function or the voli-
onset and the compensation state at the time tional use of vestibular information for postural
of testing. Many of the electrophysiologic stability and control.
measures described below are designed to
provoke abnormal reflex patterns that may
be occult on physical examination but still LABORATORY EXAMINATION:
problematic from a dizziness and imbalance
perspective. VOR-BASED MEASURES
Key Points Recording Method

• Patterns of abnormal reflex behaviors There are two recording methods currently
will follow the pattern of damage to available in clinical settings to measure eye
vestibular end organ structures, but movements: electrode-based electrooculogram
are modified by central compensatory recording (EOG) or infrared camera-based
processes. video-recording methods. In electrooculogra-
• Three cardinal behaviors of abnormal phy, the corneoretinal potential (approximately
vestibular-induced eye movements: (1) 1 mV, with the front of the eye positive rel-
the slow component of pathologic vestibu- ative to the back of the eye) serves as the
lar nystagmus follows the planes of the source signal recorded. Electrodes are placed
involved SCCs; (2) when gaze is directed at the outer canthi of each eye to measure
toward the nystagmus fast phase, hori- the horizontal component of eye movement,
zontal nystagmus will increase (Alexan- and electrodes placed above and below one
der’s Law); and (3) vestibular-induced eye eye are used to measure vertical movements.
movements are modified by higher level A common, or ground, electrode is placed on
signals from gaze-holding, saccadic, and the forehead just above the nose. As the eye
pursuit systems. moves horizontally or vertically, the corneo-
• Gaze nystagmus that is direction fixed retinal source dipole changes its orientation
(Alexander’s Law) is classified as 1st, 2nd, relative to the recording electrodes. Changes
or 3rd degree. The 3rd degree is present in the field potential are measured across elec-
in all positions of gaze, 2nd degree is trode pairs to estimate eye position. To use
present with gaze away from the lesion the EOG method, there must be a corneo-
and at midline, and 1st degree is present retinal potential. If a patient is blind in both
only with gaze away from the side of the eyes and has no corneoretinal potential, elec-
lesion. trooculography cannot be used to measure eye
• They lawfully reflect the planes of the movement.
involved SCCs. Infrared video-recording systems have
• When vestibular nystagmus is observable quickly emerged as the preferred method of
with the eyes opened, the horizontal com- recording eye movements. The value of these
ponent will always beat in the same direc- systems is that eye movements can be observed
tion regardless of eye position and will and recorded for offline analysis. Addition-
follow Alexander’s Law. ally, eye position can be controlled, avoiding
• Visual fixation suppresses vestibular-induced measurement errors when the eyes deviate
nystagmus. significantly away from the central position.
Infrared video-recording systems may employ the slow component is calculated in degrees
slow frame rates, limiting the recording res- per second (termed slow component velocity
olution of faster eye movements. Faster dig- or SCV) and serves to quantify the nystagmus
itization speeds may overcome this limitation magnitude.
and allow for more refined measurement of There are many types of nystagmus: pen-
high velocity saccadic and vestibular-induced dular, rotatory (torsional), jerk (horizontal and
eye movements. Additionally, pattern recog- vertical); and causes of nystagmus: vestibu-
nition schemes are being developed to allow lar, congenital, cerebellar as in rebound
for the measurement of torsional eye move- nystagmus, ocular motor as in internuclear
ments by recognizing changes in the orien- ophthalmoplegia or gaze evoked. Other eye
tation of the iris muscle pattern. The ability movement behaviors, such as inappropriate
to record torsional eye movements increases square waves, disconjugate eye movements,
the sensitivity of video-recording methods and ocular dysmetria such as saccadic over-
to changes in vertical canal mediated eye or undershoots, may also reflect pathologic
movements. conditions. Leigh and Zee8 offer an exhaus-
Regardless of which recording system is tive compendium of most types of nystag-
used, certain conventions have been developed mus and other eye movement abnormalities.
to display eye movement tracings for anal- Recording abnormal eye movement behav-
ysis. The conventions for plotting horizontal iors may help distinguish between the dif-
and vertical nystagmus over time are shown in ferent types and causes. However, this does
Figure 34–2. For horizontal and vertical eye not emerge automatically from ENG or VNG
movements, “pen up” defines the movements records. Rather, it remains for the exam-
that are up or to the right and “pen down” iner to be well tutored in recognizing these
defines those that are down or to the left. The eye movements and appreciating their impli-
chart is read from left to right. Nystagmus cations. Nystagmus evoked by vestibulopathy
direction is defined by the fast, or jerk, com- will be limited to horizontal, vertical, or tor-
ponent, which is generated by the CNS. The sional “jerk” forms of nystagmus, correlating
slower component is in the opposite direction with the involved sensory organs within each
and is generated by the VOR. The speed of labyrinth.
Figure 34–2. Electrooculographic recordings of horizontal and vertical nystagmus. A, Right-beating horizontal nystagmus
and up-beating vertical nystagmus. B, Left-beating horizontal nystagmus and down-beating vertical nystagmus. On the
latter, tracing is an example of how the velocity (slope, rise/run) of the slow component is measured as 10◦ /second.
Key Points undergoes vestibular testing. Patients should

refrain from smoking tobacco, because nico-
• The corneoretinal potential (approximately
tine constricts blood vessels and, thus, impairs
1 mV, with the front of the eye positive the blood supply to the vestibular mechanism.
relative to the back of the eye) is the As a stimulant, caffeine can also adversely
source potential recorded during elec- affect the vestibular reflexes. Alcohol alters
trooculography. the chemical balance of the perilymph and
• Electrodes are placed at the outer can-
endolymph and induces positional nystagmus.
thi of each eye to measure the hor- Geotropic (beating toward the earth) positional
izontal component of eye movement, nystagmus can be induced within one-half hour
and electrodes placed above and below of alcohol ingestion and continue up to 4 hours.
one eye are used to measure vertical Ageotropic (beating away from the earth) posi-
movements. tional nystagmus can be observed from 5 to 24
• Changes in the field potential are mea-
hours after alcohol is ingested. Consequently,
sured across electrode pairs to estimate alcohol should be avoided for at least 24 hours
eye position. before testing.
• Infrared video-recording systems are the
Use of anticonvulsants, long-acting SSRIs,
preferred method of recording eye move- and medications for vascular regulation should
ments, since eye movements can be not be discontinued before testing. Patients
observed and recorded for offline analysis with diabetes should eat a light meal at least 2
and eye position can be controlled, avoid- hours before testing and should not avoid tak-
ing measurement errors when the eyes ing their insulin. Leigh and Zee8 provide an
deviate significantly away from the central
in-depth discussion of the effects of drugs on
position.
• By convention for horizontal and verti- eye movement.
cal eye movements, “pen up” deflections Key Points
define eye movements that are right-
ward (horizontal vector) or upward (ver- • Medications, such as vestibular suppres-
tical vector) and “pen down” deflections sants, sedatives, tranquilizers, short-acting
indicate eye movements that are leftward antidepressants, and pain relievers, should
(horizontal vector) or downward (vertical be discontinued 24–48 hours before the
vector). patient undergoes vestibular testing.
• The speed of the slow component of nys- • Patients should refrain from smoking
tagmus is calculated in degrees per second before vestibular testing, since nicotine
(termed slow component velocity or SCV), constricts blood vessels and impairs blood
and quantifies the nystagmus magnitude. supply to the vestibular mechanism.
• Nystagmus evoked by vestibulopathy will • Caffeine can adversely affect the vestibu-
be limited to horizontal, vertical, or tor- lar reflexes.
sional “jerk” forms of nystagmus, corre- • Alcohol alters the chemical balance of the
lating with the involved sensory organs perilymph and endolymph and induces
within each labyrinth. positional nystagmus and should be avoi-
ded for at least 24 hours before testing.
Preparation for Testing Videonystagmography Test Battery

A carefully documented history should rule
Purpose and Role of the ENG/VNG Test
out any preexisting condition (e.g., congenital Battery
nystagmus and use of medications or drugs)
that could influence test results. Many medi- • Inventory visually guided eye movements,
cations, such as vestibular suppressants, seda- such as saccadic, pursuit, and optokinetic-
tives, tranquilizers, antidepressants, and pain based eye movements, are used to detect
relievers, have side effects related to dizziness. visual and central eye movement deficits
When possible, these medications should be and ensure that the ocular motor final
discontinued 24–48 hours before the patient common pathway is intact and adequate
for interpreting vestibular-induced eye overshooting occurs consistently, it may be

movements without artifact. related to posterior fossa involvement. New
• Test for pathologic vestibular-induced computerized systems produce random visual
nystagmus provoked by head movements targets [typically light-emitting diodes (LEDs)]
or static head positions without visual subtending a visual arc of ±30◦ on a light bar.
fixation. The accuracy, latency, and velocity of the eye
• Compare the strength of the horizontal movements relative to the stimulating signal
SCC-induced VOR evoked from each ear are calculated and compared with age- and
using the bilateral, bithermal caloric test. sex-matched normal values. As a very gen-
Vestibulopathy that decreases the caloric eral rule, slow saccadic velocities are associ-
response from one ear results in a caloric ated with ocular motor final common pathway
asymmetry termed a unilateral weakness deficits. Impaired accuracy may be seen in final
(UW). A UW score of 25% or more usually common pathway or cerebellar involvement.
indicates a significant asymmetry in SCC Saccadic onset latency delays, when unilateral,
output. may imply a deficit at any level in the functional
unit. There are several technical variables
that influence the accuracy of these parame-
INVENTORY OF VISUALLY ters, including the sampling rate used in the
GUIDED EYE MOVEMENTS digitization process. Video-recording systems
Assessing visually guided eye movements (sac- sampling at 30 Hz may not be adequate for
cades, smooth pursuits, and optokinetic reflex reliable saccadic velocity measurements. Fur-
movements) contributes to the assessment of ther, before abnormal latency and accuracy
the dizzy patient in three ways. First, from values are considered significant, visual impair-
a historical perspective, measuring visually ments such as cataracts, retinopathy, central
guided eye movement behaviors can detect scotoma, or poor patient cooperation must be
ocular motor and supranuclear CNS disease. ruled out.
Improved clinical knowledge and modern elec-
trophysiologic and imaging studies have largely Key Points
supplanted the use of eye movements as pri-
mary indicators of CNS disease. Nevertheless, • During saccadic eye movement testing,
recording abnormal visually guided eye move- the patient looks back and forth at targets
ments can help validate eye movement abnor- positioned at least ±10◦ off midline, pro-
malities observed or perhaps overlooked on the viding an opportunity to assess for under-
bedside examination. shoot or overshoot dysmetria.
The second contribution to assessing visu- • The accuracy, latency, and velocity of
ally guided eye movement measurements, the eye movements relative to the stim-
along with bedside ocular motor testing, is ulating signal are calculated and com-
to ensure the ocular motor final common pared with age- and sex-matched normal
pathway is intact and adequate for interpret- values.
ing vestibular-induced eye movements without • Slow saccadic velocities are associated
artifact. Finally, in the elderly, inaccurate and with ocular motor final common path-
visually guided eye movements can correlate way deficits.
with impaired visual perception, making it dif- • Impaired accuracy of saccadic eye move-
ficult for patients to accurately identify obsta- ment testing may be seen in final common
cles while attempting to ambulate through an pathway or cerebellar involvement.
environment. • Saccadic onset latency delays, when uni-
lateral, may imply a deficit at any level in
the functional unit.
Saccadic Eye Movement Testing • Before abnormal latency and accuracy val-
During calibration, the patient looks back ues during saccadic eye movement testing
and forth at targets positioned at least are considered significant, visual impair-
±10◦ off midline, which provides the exam- ments such as cataracts, retinopathy, cen-
iner an opportunity to look for undershoot tral scotoma, or poor patient cooperation
or overshoot dysmetria. If undershooting or must be ruled out.
Smooth Ocular Pursuit Testing findings include asymmetry between right and
leftbeating or an inability of the patient to
The smooth ocular pursuit test can be con- increase eye speed with increased stimulus
ducted by having the patient hold the head still speed appropriately. Asymmetric OKN may be
and follow a pendulum with the eyes. Newer seen in the presence of strong spontaneous
computerized systems produce a range of fre- nystagmus. OKN asymmetries recorded with-
quencies, using pendular-like signals, by turn- out the presence of a positional nystagmus
ing on and off adjacent LEDs on the light bar. or caloric weakness may imply supratentorial
Ocular pursuit operates well up to a frequency involvement. Perceptually, normal individuals
of 1 Hz. The computerized pursuit tests usu- exposed to a moving full-field optokinetic stim-
ally cover a range from 0.2 to 0.7 Hz. The test ulus experience a strong sensation of circu-
is sensitive to medication effects, poor vision, lar vection. In contrast, some individuals with
and poor patient cooperation. Several trials closed head injury fail to experience self-
may be needed to ensure that the patient is movement despite normal reflexive nystagmus,
trying his or her best. Pursuit performance suggesting a disconnection between centers
also diminishes with age. Age-matched nor- associated with the perception of self-motion.
mal reference values are required to distin-
guish between normal aging effects and true
pathologic pursuit. When these confounding Key Points
variables are excluded, abnormal test results
• OKN tests smooth ocular pursuit or CNS
suggest CNS dysfunction. The most common
abnormality is “cogwheeling” pursuit, in which function.
• OKN asymmetries recorded without the
the eyes are continually making saccadic move-
ments to catch up with the target. Unilateral presence of a positional nystagmus or
loss of pursuit or saccadic intrusions may be caloric weakness may imply supratentorial
observed with saccadic dysmetria in unilateral involvement.
cerebellar disease.
TESTING FOR PATHOLOGIC
Key Points NYSTAGMUS
• The smooth ocular pursuit test can be con-
Positioning Induced Nystagmus
ducted by having the patient hold the head
still and follow a pendulum or electronic Positioning induced nystagmus is nystagmus
pendular-like signals with the eyes. that is provoked following head movement.
• Smooth ocular pursuit testing is sensitive Benign paroxysmal positioning vertigo (BPPV)
to medication effects, poor vision, and is by far the most common cause of position-
poor patient cooperation. ing induced vertigo.9, 10 BPPV can be provoked
• Smooth ocular pursuit performance also from any SCC. However, the posterior canal
diminishes with age. is most commonly involved. It follows that the
• Unilateral loss of pursuit or saccadic intru- most common provocative maneuver is the
sions may be observed with saccadic dys- Dix–Hallpike test, or Nylen maneuver, which
metria in unilateral cerebellar disease. moves the patient in the plane of the right (or
left) posterior SCC.
Optokinetic Nystagmus To perform a Dix–Hallpike maneuver, the
patient is seated with the head turned 45◦
Optokinetic nystagmus (OKN) is another test to the right or the left and is then placed
of smooth ocular pursuit or CNS function. The in the corresponding head-hanging position
stimulus is usually generated as a series of light (with the head still turned 45◦ to the side).
and dark vertical bars that move from right to The eyes should remain open so the exam-
left or left to right of the patient at 20◦ , 40◦ , or iner can observe any torsional nystagmus. In
60◦ per second. Ideally, the entire visual field of the past, it was thought that the movement
the patient should be filled with these stimuli should be brisk. However, more comfortable
in a darkened room. Less acceptable alterna- movements can still provoke BPPV symptoms
tives are small handheld rotating drums with so long as the movement is along the plane of
black and white stripes or a series of LEDs that the involved canal. The maneuver is illustrated
appear to move across a light bar. Abnormal in Figure 34–3.
Figure 34–3. Dix–Hallpike maneuver (Nylen maneuver) for the head-hanging right position. A, In the sitting position,
the patient turns the head 45◦ to the right with the eyes open and fixed. The patient is then put quickly into the supine
position. B, The patient in the head-hanging right position with eyes open and fixed. Observe for nystagmus and dizziness
for at least 30 seconds. The classic positive response is counterclockwise nystagmus with the head down and turned to the
right and clockwise nystagmus with the head down and turned to the left. The nystagmus is reversed with sitting up but is
less intense.
When positive, a brisk nystagmus will the affected ear. Herdman17 has provided an
develop after a brief delay of 5–40 seconds. excellent flowchart describing the various types
The nystagmus will be identical to the nor- of BPPV and their treatments.
mal nystagmus provoked while moving into the EOG recording techniques have little value
provocative position (i.e., the eye moves in the in the assessment of BPPV because this
plane of the posterior canal). The nystagmus method cannot record torsional eye move-
will crescendo and then gradually diminish, ments. However, video-based systems allow
typically within 10–40 seconds of onset. The the examiner to more carefully evaluate eye
provoked vertigo will duplicate the patient’s movements offline. This makes detecting hor-
symptoms. When the patient is returned to izontal and anterior canal variations easier to
the sitting position, a second burst of nystag- detect.
mus may develop. Returning the patient to the Central positioning nystagmus is a rare
head-hanging position will often provoke a less abnormality associated with posterior fossa dis-
intense response. ease.18, 19 Positioning nystagmus may take any
Epley11 postulated the existence of float- form, but is often vertical or oblique. Impor-
ing particles in the posterior SCC as the tantly, central positioning nystagmus is not
cause of these symptoms and proposed a associated with vertigo or dizziness and does
physical maneuver, the canalith reposition- not demonstrate visual fixation suppression.
ing procedure, to move the particles out of In certain lesions around the fourth ventricle,
the involved canal. Several investigators have severe nausea may accompany the nystagmus.
reported excellent success rates with this pro-
cedure.12–16 Key Points
Variants of BPPV involve loose particles in
the horizontal or anterior SCCs. Treatment for • BPPV is by far the most common cause of
horizontal canal BPPV involves a procedure positioning induced vertigo.
in which a supine patient is rotated 360◦ to • The most common provocative maneuver,
the right or left, in the direction away from the Dix–Hallpike test, moves the patient
in the plane of the right (or left) poste- right, supine head left, lateral right (no neck
rior SCC. torsion), lateral left (no neck torsion), head
• When the Dix–Hallpike test is positive, a hanging, and supine with the upper torso ele-
brisk nystagmus will develop after a brief vated by 30◦ (Fig. 34–4). The purpose for test-
delay of 5–40 seconds. The nystagmus will ing upper torso elevated by 30◦ is to have a
crescendo and then gradually diminish, reference for establishing the amount of spon-
typically within 10–40 seconds of onset. taneous nystagmus present during caloric irri-
• Central positioning nystagmus is a rare gation (see below). Recordings are made for
abnormality associated with posterior 20–60 seconds. Visual fixation and volitional
fossa disease. suppression of nystagmus are common pitfalls
of the positional test.
Gaze Testing Normal patients may demonstrate some
positional nystagmus.20 Typically, this nystag-
Gaze testing is performed with the patient’s mus will not exceed 6◦ per second in the hor-
eyes fixed on a visual target in the primary izontal vector or 9◦ per second in the vertical
gaze position and ±30◦ from midline, both vector in a single head position. Nystagmus
horizontally and vertically. Gaze-evoked nys- that exceeds 6◦ per second in the horizontal
tagmus (nystagmus that changes direction with vector is consistently present in four or more
eye position) typically implies CNS involve- positions with an SCV greater than 4◦ per sec-
ment. Intoxication or centrally acting medi- ond, or nystagmus that changes direction in a
cation effects will commonly cause this eye single head position does not often occur in
movement behavior. Often, this nystagmus will otherwise normal patients.21, 22
not be observed when visual fixation is denied. Abnormal findings exceeding the limits
Peripheral vestibulopathy may also produce described above are categorized as (1) direction-
abnormal gaze nystagmus with the eyes open fixed or (2) direction-changing. Direction-
and fixed. Importantly, this nystagmus will changing can be geotropic (fast phase beats
demonstrate the three cardinal behaviors of toward the floor in lateral head positions),
vestibular nystagmus mentioned above. The ageotropic (fast phase beats toward the
nystagmus will be direction-fixed, will follow ceiling in lateral head positions), or
Alexander’s Law, and will suppress with visual direction-changing in a single head position.
fixation. Eye movement recordings enable cal- Although the findings are nonlocalizing, there
culation of nystagmus slow-component veloc- are some general rules. Direction-fixed posi-
ities so that the effects of eye position, visual tional nystagmus is often observed in chronic
fixation, and nystagmus direction can be objec- unilateral hypofunction and usually beats
tively established. toward the unaffected or less affected ear. With
an irritative lesion such as in Meniere’s syn-
Key Points drome, nystagmus can beat toward the dis-
• Gaze-evoked nystagmus typically implies eased ear. Direction-changing nystagmus may
CNS involvement. be seen with otolith organ involvement, though
• Peripheral vestibulopathy may also pro- not exclusively so. The intensity of positional
duce abnormal gaze nystagmus with nystagmus is also useful in monitoring physio-
the eyes open and fixed. When the logic compensation during the disease process,
observed nystagmus follows Alexander’s because the nystagmus diminishes as compen-
law, suppresses with visual fixation, and sation occurs.
corresponds with an underlying posi- Direction-changing nystagmus in a single
tional/spontaneous nystagmus, it is almost head position (excluding positioning induced
certainly vestibular induced. nystagmus) is usually a sign of CNS dysfunc-
tion and is referred to as periodic alternat-
Positional Testing Without Fixation ing nystagmus. Periodic alternating nystagmus
may change direction every two minutes or
The purpose of positional testing without fix- so. Persisting vertical or oblique nystagmus
ation is to measure eye movement with the may also be of central origin, particularly if
patient’s head held in various static posi- the nystagmus does not suppress with visual
tions, such as sitting, supine, supine head fixation.
Figure 34–4. Head positions used in the static positional test. The patient’s eyes remain closed, but the patient must
be alert.
Key Points which orients the lateral SCCs vertically. The

• The purpose of positional testing without stimulus used is water (±7◦ C relative to nor-
mal body temperature) or air (±13◦ C relative
fixation is to measure eye movement with
to normal body temperature). The ear is irri-
the patient’s head held in various static
gated with water for 30–40 seconds or with air
positions. for 60–70 seconds.
• Normal patients may demonstrate some
The theory behind the caloric test is that
positional nystagmus, which does not
warming or cooling the bone surrounding
exceed 6◦ per second in the horizontal vec- the lateral SCC creates a convection current
tor or 9◦ per second in the vertical vector in the endolymph that causes utriculopetal
in a single head position. or utriculofugal flow. The change in rest-
• Direction-fixed positional nystagmus is
ing discharge rate is then registered within
often observed in chronic unilateral hypo- the vestibular nuclei, where the asymmetry
function and usually beats toward the in ampullary nerve output between the hori-
unaffected or less affected ear. zontal SCCs drives VOR-induced eye move-
• Direction-changing nystagmus in a single
ments. Bithermal irrigations are used to induce
head position is usually a sign of CNS eye movements in both directions from each
dysfunction and is referred to as periodic ear. This minimizes the effect of underlying
alternating nystagmus. spontaneous nystagmus on caloric asymmetry
calculations.
The left lateral SCC at rest and after warm
CALORIC IRRIGATION
(44◦ C) and cool (30◦ C) caloric stimulation is
The caloric test is the bedrock of the shown in Figure 34–5. Warm stimuli cause the
VNG/ENG battery. It is one of the few tests nystagmus to beat toward the stimulated ear,
that allow each labyrinth to be tested in and cool stimuli cause it to beat toward the
isolation. Caloric irrigation is performed with nonstimulated ear, thus the acronym COWS
the patient’s head in a 30◦ supine position, (cold opposite, warm same).
Figure 34–5. A, Lateral semicircular canal at rest (i.e., no caloric stimulation or fluid rotation). Kinocilium and stereocilia
are vertical, producing a normal resting potential in vestibular nerve. B, Stimulation with warm water (44◦ C) causes upward
convection current (utriculopetal endolymph flow). Stereocilia bend toward kinocilium, depolarizing the dendrites at base
of hair cell, thus increasing firing rate of the vestibular nerve. C, Stimulation with cool water (30◦ C) induces downward
convection current (utriculofugal endolymph flow), causing kinocilium to bend toward stereocilia, which hyperpolarizes
dendrites at base of hair cell, thus decreasing firing rate of the vestibular nerve.
590
The goal of the caloric test is to compare 30%–40%, the result is abnormal and indica-
the strength of each horizontal canal in an tive of a CNS lesion. A summary of ENG test
effort to identify the weak or paretic ear. battery results is given in Table 34–1.23
This is accomplished by the Jonkees UW Caloric irrigation is the standard test for
formula. Unilateral weakness is determined determining the laterality of lesions involving
by comparing the nystagmus generated on the horizontal canal or superior vestibular
each side: nerve branch. However, certain pitfalls must
be avoided. Small variations in technique can
(RW + RC) − (LW + LC) influence the performance of the caloric test.
UW = × 100
(RW + RC + LW + LC) Careful equipment calibration and laboratory-
specific reference values are necessary. The
where RW is the right warm-peak SCV, RC patient must not be allowed to fixate visually
is the right cool-peak SCV, LW is the left but must remain mentally alert. Also, con-
warm-peak SCV, and LC is the left cool-peak genital nystagmus or any drugs or medica-
SCV. The result is a percentage difference in tion that could influence the results must be
nystagmus magnitude between ears. In most ruled out.
testing laboratories, a unilateral weakness of The results in patients with perforated tym-
at least 25% is needed to be clinically signif- panic membranes may also be misleading. If
icant. An example of a left peripheral weak- the middle or external ear is wet because of
ness, 49% weaker on the left side, is shown in infection and drainage, a warm air stimulus ini-
Figure 34–6. tially cools rather than heats the bone and the
Directional preponderance (DP) compares nystagmus beats in the direction produced by a
the magnitude of rightward- and leftward- cold stimulus. This could be misinterpreted as
directed nystagmus during the caloric test. It a CNS abnormality.
is calculated as follows:
Key Points
(RW + LC) − (LW + RC)
DP = × 100 • Visually guided eye movement tests
(RW + RC + LW + LC)
include measurement of saccadic, pur-
The difference between the two directions suit, and optokinetic-induced nystagmus.
must be at least 30% to be considered signif- When medication effects, poor vision, and
icant. DP is nonlocalizing. It usually accom- poor patient cooperation can be excluded,
panies a direction-fixed positional nystagmus abnormalities may point to CNS involve-
and, like positional nystagmus, reflects central ment. Abnormal tests require clinical cor-
compensation state. relation to establish significance.
Bilateral weakness is suggested when the • Positioning nystagmus is nystagmus that
sum of the peak nystagmus SCVs from all four develops as the result of prior move-
irrigations is less than 28◦ per second. An exam- ment. BPPV is the most common form
ple of this is shown in Figure 34–7. The total and is easily treated with repositioning
nystagmus generated (in degrees) is 0 + 4 + 0 + maneuvers.
3 = 7◦ per second. The caloric test is not well • Positional nystagmus that is persistent
suited to detecting less severe forms of bilat- and suppresses with visual fixation is
eral vestibular weakness. Test variability is such commonly of peripheral origin and ref-
that, in some cases, it is difficult to distinguish lects incomplete central compensation.
between true bilaterally weak responses and Positional nystagmus is likely significant
low SCVs evoked from normal patients. Rotary when it exceeds 6◦ per second in the hor-
chair tests are better suited to determining true izontal vector or 9◦ per second in the ver-
bilateral vestibular weakness. tical vector in a single head position; or is
Another aspect of the caloric test is the mea- greater than 4◦ per second and is consis-
surement of visual fixation suppression. Shortly tently present in four or more positions.
after the eyes reach their maximal velocity, • Positional nystagmus that changes direc-
vision is reestablished and the patient is asked tion in a single head position and is not
to fixate on a visual target. If the patient associated with BPPV is likely a cen-
does not suppress the nystagmus by at least tral, periodically alternating nystagmus.
Figure 34–6. Responses to caloric stimuli in a patient with left unilateral vestibular weakness. A, Calculated response of
the nystagmus over time. The small boxes represent peak eye velocity values averaged for each of the four irrigations. These
responses show a 49% left peripheral weakness and a 30% right-beating directional preponderance. B, Raw data obtained
during peak eye velocities. Note the weak responses obtained by stimulating left ear. SPV, Slow phase velocity.
592
Figure 34–7. Responses to caloric stimuli in patient with peripheral vestibular weakness bilaterally. A, Calculated
response of nystagmus over time. The small boxes represent peak eye velocity values averaged for each of the four irriga-
tions. These responses show a total of 7◦ of right-beating nystagmus, which is probably caused by the positional nystagmus
present whenever eyes are closed. B, Raw data obtained during peak eye velocities. Note weak responses with stimulation
of both ears. SPV, slow phase velocity.
593
Table 34–1 Electronystagmography Test Battery: Correlation of

Abnormal Findings and Suspected Site of Lesion∗
Test Type of abnormality Suspected site of lesion
Saccade Ipsilateral dysmetria Cerebellopontine angle

Bilateral dysmetria Cerebellum
Decreased velocity Throughout the central nervous
system, muscle weakness or
peripheral nerve palsy
Internuclear ophthalmoplegia Medial longitudinal fasciculus
Pursuit Break-up Brainstem or cerebrum
Saccadic Cerebellum
Gaze Direction-fixed and horizontal Peripheral vestibular
Direction-changing and vertical Brainstem
Up-beating Brainstem or cerebellum
Down-beating Cervicomedullary junction or
cerebellum
Rotary Vestibular nuclei/brainstem
Failure of fixation Less than 40% decrease Brainstem or cerebellum
suppression
Positional Direction-fixed Nonlocalizing or peripheral
Direction-changing Nonlocalizing or central
Dix-Hallpike Classic Peripheral
vestibular—undermost ear
Nonclassic Nonlocalizing
Caloric Unilateral or bilateral weakness Peripheral vestibular
Directional preponderance Nonlocalizing
∗ Exceptions to the rule may occur.
From Cyr.19 By permission of Allyn & Bacon.
Persisting vertical or oblique nystagmus In some protocols, the rotary chair chamber
may also be of central origin, particularly is darkened or deliberately illuminated. The
if it does not suppress with visual fixation. resulting compensatory eye movements devel-
• Asymmetries greater than 25% on the oped as the chair turns reflect either VOR or
bilateral, bithermal caloric test are com- visually enhanced VOR (VVOR) respectively.
monly associated with unilateral weak- There are several advantages to the rotary chair
ness in the ear with lower nystagmus test. First, for most test protocols, patients tol-
peak SCVs. erate rotary chair testing better than caloric
• DP asymmetries greater than 30%, while irrigation tests. Second, computer-controlled
not diagnostic, may reflect central com- angular rotation is a more consistent stimulus
pensation status when unilateral weakness than caloric irrigation, which produces more
is also detected. reliable data in the form of gain, phase, and
• Bilateral vestibular weakness may be symmetry measurements. Further, the con-
present when the sum of the caloric SCVs trolled rotational stimulus appears to be bet-
is less than 28◦ per second. ter for monitoring changes over time than the
caloric test. Third, bilateral horizontal canal
weakness can be efficiently quantified. Finally,
Computerized Rotary Chair Tests small children can be tested without difficulty.
A major shortcoming is that slow accelera-
Computerized rotary chairs assess the VOR tion rotary testing stimulates both lateral SCCs
while the patient is turned in a precisely simultaneously. Techniques for isolating ear-
controlled manner. There are several test specific VOR responses are developing, using
strategies that can be employed with this tool. high-acceleration chair movements (following
Ewald’s second law). However, these methods they are open. As the chair rotates, the com-
are not currently in wide use. Caloric irriga- puter digitizes the analog signals from the eyes
tion is still the primary test for evaluating each and compares the eye movement with the chair
horizontal SCC independently. rotation. The algorithms compare the veloc-
ity, phase, and gain of the two signals. At low
Purpose and Role of Computerized Rotary frequencies (e.g., 0.01 Hz), normal eye veloc-
Chair Testing ity leads chair velocity by as much as 45◦ . As
the chair frequency increases and approaches
• Slow harmonic acceleration (SHA) is use- 0.64 Hz, the phase difference approaches zero.
ful in detecting bilateral weakness, detect- With the patient rotating in the dark, the gain
ing subtle VOR deficits, and testing infants (ratio of eye velocity to chair velocity) is low
or elderly patients at risk for bilateral at low frequencies and increases at higher
weakness. frequencies. The relationships of phase, gain,
• With vision, SHA measures the contribu- and symmetry of chair velocity and eye veloc-
tion of vision to VOR-induced eye move- ity are shown in Figure 34–8. Normal phase
ments. Abnormalities on this task pre- gain and symmetry are shown for a patient in
dict oscillopsia (perceived instability of the Figure 34–9.
visual scene from loss of stabilizing eye The data from a patient with left periph-
movements). eral vestibular weakness, as indicated by a
• Fixation suppression tests detect central 59% caloric difference between the two ears
vestibular impairment. and a 22% right-beating DP, are shown in
• High-acceleration step tests may demon- Figure 34–10. Note that the gain is normal
strate reduced VOR gain when rotating from 0.01 to 0.32 Hz (0.64 Hz was not tested).
toward the weak ear, and shortened time Phase is abnormal or borderline abnormal
constants indicate vestibular deficits. from 0.01 to 0.16 Hz. Asymmetry, although
within normal range, is slightly below the
The test used most often is the low-frequency line, indicating that right-beating vestibu-
Slow harmonic acceleration (SHA) test, with lar nystagmus is greater than left-beating
the patient kept in total darkness. To under- nystagmus. This patient record demonstrates
stand SHA, it is important to recognize that the signs of central compensation for an under-
normal horizontal head movements about the lying peripheral vestibular lesion. Despite the
pivot point of the cervical spine occur at fre- persisting left caloric weakness, there is very
quencies between 1 and 4 Hz. Below 1 Hz, little abnormality noted in SHA test results due
VOR-induced eye movements will not be suf- to the effects of central compensation. Asym-
ficient to keep the eye focused on a visual metries, phase leads, and gain reductions are
target. Consequently, when head movements common in acute lesions. Over time, abnor-
occur at lower frequencies (slower accelera- mal gain and asymmetry values trend into the
tion and deceleration speeds) VOR-induced normal range as central compensation occurs.
eye movements must be augmented by pursuit Phase (reflecting VOR timing relationships)
eye movements. can remain abnormal, particularly if there is
SHA testing consists of accelerating and complete loss of function on one side.
decelerating the chair from 0◦ to 50◦ or 80◦ Slowly developing deficits, such as a vestibu-
per second in a sinusoidal fashion from 0.01 to lar schwannoma, may not show any abnormal-
0.64 Hz. This is below the optimal frequency ity at all on SHA, due to the slow growth of
range of the horizontal VOR. VOR-driven eye the mass and ample opportunity for central
movements are systematically reduced in gain circuits to compensate for reductions in end
and demonstrate timing distortions (expressed organ output. Because both ears can contribute
as phase lead in degrees). When patient to evoked eye movements, SHA test results
responses demonstrate lower than normal gain tend to reflect central compensation status
reductions, higher than normal phase leads, or more than reductions in peripheral vestibular
tend to provoke a stronger nystagmus in one output.
direction, vestibulopathy is very likely. Patients with total bilateral vestibular weak-
Most modern systems use an infrared cam- ness have poor gain at all frequencies and
era to monitor the patient’s eyes to ensure that no response to caloric irrigation. Phase and
A
Phase Phase difference
Eyes lead chair
Peak chair
Velocity
To right
50
Velocity
Degrees/ 0
second
Slow phase peak
–50
Eye velocity
180° 360°
B Peak chair velocity
Gain To right Slow phase peak
50
Eye velocity
Velocity 0
Degrees/
second
–50
Gain = Eye velocity/Chair velocity 180° 360°
C Peak chair velocity

Symmetry Slow phase To right Slow phase
Eye velocity Eye velocity
Right beating Left beating
50
Velocity
Degrees/ 0
second
–50
180° 360°
Symmetry = Right beating – Left beating
Figure 34–8. Measurement of phase (A), gain (B), and symmetry (C) using a computerized rotary chair. Sine waves
represent fast Fourier analysis of the velocity of the chair and slow phase movement of the eyes, as indicated.
symmetry values are meaningless in this situ- vestibular cerebellum produce a loss in the
ation because there is no eye velocity to use ability to suppress nystagmus in one direction.
for comparison. Lesser degrees of vestibular Bilateral loss of fixation suppression can also
weakness may provoke near normal caloric occur from central vestibular deficits. As with
responses, but demonstrate bilateral weakness other visually guided eye movement tests, poor
on SHA measurements. This commonly occurs vision and patient attention problems must be
in the elderly. The caloric test, in these cases, excluded before the possibility of CNS disease
may underestimate the role played by the can be entertained.
vestibular system in multifactorial imbalance of The rotary chair affords the ability to investi-
the aged. gate VVOR interactions. The patient, with eyes
Fixation suppression of vestibular-induced open, is rotated in a lighted room. Thus, visual
eye movements can be measured with great and vestibular clues are available. In normal
precision during SHA testing. A visual fixa- subjects, the test produces gain measurements
tion target is presented that moves at the same that approach 1.0 and phase measurements
velocity as the chair. Normal patients are able that approach 0◦ . Acceptable VVOR perfor-
to suppress vestibular nystagmus by gazing at mance requires normal vestibular and visual
the visual target. Unilateral disorders of the pursuit integration and results in stable vision
Figure 34–9. Normal rotary chair test results for phase, gain, and symmetry obtained with patient rotating in the dark.
Results in shaded areas are abnormal.
Figure 34–10. Results of rotary chair test (conducted in darkness) in patient with right peripheral weakness. Phase is
abnormal, but gain and symmetry are normal. Shaded areas are abnormal.
597
with fast or slow head movements. Low gain change in the vestibular mechanism induced
in high-frequency movements may be seen by ototoxic medications. The rotary chair also
in cases of severe bilateral vestibulopathy. provides the best environment for optokinetic
Deficits with low-frequency movements may testing, because most of the visual field can be
be seen when pursuit is inadequate to produce filled with the moving visual stimuli.
stable vision. Deficits in the middle frequen-
cies result from a combination of pursuit and Key Points
vestibular weakness. Thus deficit performance • SHA methods are the test of choice
on this test predicts oscillopsia with routine
for identifying bilateral vestibular weak-
head movements.
nesses.
The step test provides the information nec- • Abnormal timing relationships demon-
essary to assess the time constant (i.e., the time
strated in phase measures are sensitive
it takes for nystagmus to decay to 37% of its
measures of vestibulopathy.
maximum after stimulation has stopped). The • Abnormal VVOR results indicate risk of
patient is quickly accelerated from 0◦ to 60◦ ,
oscillopsia.
100◦ , or 240◦ per second in 0.5 second. The • Fixation suppression can be measured
chair then continues to rotate at 60◦ , 100◦ , or
with great precision using the SHA
240◦ per second for 1 minute. As the flow of the
paradigm.
endolymphatic fluid in the SCCs approaches • Rotary chair tests are good for measuring
the velocity of the head, the nystagmus decays.
changes in vestibular output over time.
Because of the elasticity of the cupula, the • Because both horizontal SCCs contribute
response time for it to bend and return to
to VOR-induced eye movements on most
its resting state is approximately 4–7 seconds.
rotary chair tests, the method is not opti-
However, the nystagmus continues for 10–30
mal for identifying the side of vestibular
seconds, which is attributed to central velocity
involvement. At the same time, the test
storage. When the chair is stopped suddenly
is ideal for measuring dynamic compensa-
(0.5 second), the fluid continues to move rel-
tion status.
ative to the head but in the opposite direc-
tion, thus generating nystagmus in the opposite
direction. The time constant during this period
is also measured. Either CNS dysfunction or Subjective Visual Vertical (SVV)
the inability of the peripheral system to send Assessment
the appropriate information for integration can
cause abnormal findings. Purpose and Role of SVV Assessment
There are several emerging rotational test • SVV is thought to reflect asymmetries in
methods that warrant mention. High- utricle tone. As such, this is the only
acceleration step tests may initially provoke eye measure of utricular function currently
movements from the leading ear. Measuring available.
nystagmus accurately during high-acceleration
movements is technically difficult. However, Ocular tilts can be mediated in part by
when refined, this may provide ear-specific asymmetric utricle output. Lesions that cause
asymmetry data compatible with caloric test reduced utricular tone on the involved side
results. Unilateral centrifugation, a rotational may provoke an ocular tilt reaction. In the ocu-
test where the patient is positioned so that the lar tilt reaction, the patient’s head tilts toward
axis of rotation is just over one utricle, can pro- the side of the lesion with skewed deviation and
voke utricular-mediated VOR eye movements. ocular torsion. The reaction may be observed
Measuring these can help detect utricular dis- in acute lesions of the peripheral or cen-
orders. These methods remain in evolution at tral utricular pathways. The reaction typically
the time of this writing. extinguishes quickly. However, a residual ocu-
The consistent stimulus and reliable nature lar torsion may persist longer than the head tilt,
of rotary chair testing make it the best reflecting a residual bias difference between
choice for monitoring changes in the VOR the two utricles. SVV tests measure how well a
over time. Rotary chair testing is valuable patient can volitionally set a projected line per-
when monitoring physiologic compensation or fectly vertical in an otherwise darkened room.
Normal subjects are able to manipulate a ver- proximity to the stapes footplate within the
tical line within 2◦ of true vertical or true labyrinth. Although still under study, the most
horizontal (subjective visual horizontal or SVH likely neural pathway underpinning the VEMP
measurement) without any additional visual is from the saccule to the medial and/or lat-
reference. Patients with unilateral lesions may eral vestibular nuclei via the inferior vestibular
be off by as much as 15◦ acutely, and often carry nerve. Descending vestibulo-spinal tracks then
a residual tilt of 5◦ –6◦ following compensation. carry the signal to the spinal accessory nuclei
Incorporating these measurements with con- within the anterior horn cells of cervical spine
stant velocity rotary chair rotation may enhance levels C1–C6 and ultimately to the SCM mus-
the sensitivity of SVV or SVH. With constant cle. There is likely more than one pathway for
velocity rotation, a shear force develops over this reflex. VEMPs tend to be recorded ipsilat-
the utricle, potentially increasing the asymme- erally over neck muscles, but may be seen con-
try between normal and impaired sides. tralaterally when recorded from eye muscles.
It is important to understand that patient Nevertheless, the ability to use VEMP mea-
performance on SVV or SVH tests may surements to asses inferior vestibular nerve
vary across trials. Therefore, the average of function has been an important addition to
several trials, using a controlled psychophys- VOR-based tests of horizontal canal or superior
ical method, should be used to capture vestibular nerve function. Moreover, combin-
best performance. Further, other causes of ing caloric and VEMP test results can lead
ocular tilt, including ocular motor disorders, to a view of ascending VOR and descend-
must be excluded. When this can be accom- ing VSR networks within the brain stem. This
plished, SVV and SVH measurements consti- provides a more comprehensive assessment
tute a unique vehicle for measuring utricle of central vestibular function and can aid in
function. localization.
Key Points Purpose and Role of VEMPs

• Acute vestibular lesions may produce tilts • The VEMP reflects the function of the
in SVV measurements as high as 15◦ . saccule, inferior vestibular nerve branch,
• Residual tilts of 5◦ –6◦ toward the weak and descending vestibular spinal path-
side may be the sequela of peripheral ways.
labyrinthine impairment.
• SVV tilts may require multiple trials to
RECORDING METHODS
establish reliable results.
• Other causes of ocular tilt need to be Clinical techniques for recording VEMPs are
excluded. not yet standardized. As a consequence, several
techniques have been described in the litera-
ture. Careful construction of normal reference
LABORATORY EXAMINATION: values using the specific technique applied in
the laboratory is required to optimize interpre-
VSR-BASED MEASURES tation of clinical data. Several of the variables
that influence VEMP responses are reviewed
Vestibular Evoked Myogenic below.
Potentials
Stimuli
VEMPs are early, biphasic field potentials that
can be recorded from the muscles of the VEMPs may be elicited from air-conducted
head, neck, and spine following presentation (sound travels through the middle ear conduc-
of an intense transient acoustic stimulus.24 tive mechanism) or bone-conducted (sound
The response can be recorded using standard travels through the temporal bone directly to
evoked potential methods and is commonly the inner ear) stimuli. So long as the evoking
recorded over the sternocleidomastoid (SCM) stimulus is intense and abrupt enough to elicit
muscles. The VEMP is thought to reflect stim- a synchronous discharge of the hair cell bed of
ulation of the saccule, which lies in close the macula sacculae, a sharp transient change
in the muscle field potential may be detected some authors use a large patch electrode over
through signal averaging. Acoustic clicks can the SCM so that the electrode remains close
be generated by most commercial signal aver- to the belly of the muscle during head move-
aging systems and are thus a convenient, but ments.28
not necessarily optimal, stimuli. Air-conducted The myogenic signal is amplified × 5,000,
VEMPs are optimally evoked by tone bursts in band-pass filtered from 1–5 Hz to 250–1000 Hz,
the frequency range between 500 and 1500 Hz. digitized, and signal-averaged. Epochs typically
The high-frequency energy contained in click include a 20-ms prestimulus interval and an 80-
stimuli will do little to evoke a VEMP, but will ms poststimulus interval. Importantly, artifact
make the stimulus subjectively louder to the rejection is turned off. Amplifier gain may need
patient. to be adjusted so that EMG activity does not
Stimulus rise time directly controls VEMP saturate the amplifiers.
latency. Slower stimulus rise times produce
longer VEMP latencies. Thus for tone bursts Procedure
with a single cycle rise time, a 1000-Hz tone
burst will provoke an earlier VEMP than a To obtain a VEMP, the target muscle must be
500-Hz tone burst. Click stimuli, which have contracted. There are several ways to accom-
a near instantaneous rise time, will provoke an plish this. In some laboratories, the patient
earlier VEMP than a 1000-Hz tone burst. is in the sitting position, flexing their head
For reasons that are not clearly understood, to contract the SCM muscle.29 In our labora-
bone-conducted stimuli tend to produce ear- tory, the patient lies on an examination table
lier VEMP responses than air-conducted stim- with the upper torso elevated 30◦ . During the
uli. The optimal tone burst frequency for bone- signal averaging process, the patient rotates
conducted stimuli will be somewhat lower than his or her head 45◦ so that the test ear is
for air-conducted stimuli, reflecting middle ear up. The head is then lifted off the table by
filter effects. A current limitation of bone- approximately 1 inch, causing contraction of
conducted stimuli is that effective stimulus the SCM muscle. In some laboratories, muscle
intensities are difficult to produce using cal- contraction is controlled by online measure-
ibrated bone vibrators. Skull taps, and using ment of a rectified EMG signal.25 Signal aver-
triggering reflex hammers, may get around this aging only occurs when a rectified EMG signal
limitation. falls within preset parameters. This technique
may improve test accuracy. However, defini-
tive study of these different techniques has not
Signal Averaging been accomplished at the time of this writing.
There are two electrode montages commonly Because muscle contraction is required to
used to record a VEMP from the SCM. In the observe a VEMP, muscle fatigue becomes an
first method, the noninverting lead is placed issue. This is particularly important in elderly
on the upper half of the SCM muscle, with patients. In our laboratory, we signal between
an inverted lead and ground placed on the 40 and 120 epochs per average, stopping when
sternum.25, 26 In this orientation, the first pos- the biphasic VEMP amplitude is three times
itive peak (P1 or P13 when click stimuli are larger than the wavelets in the prestimulus
used) will be plotted upward on the averaged interval. Between three and six subaverages are
waveform. In the second method, the nonin- then combined into a superaverage for analysis
verting lead and the ground are placed on the purposes. An example of VEMP waveforms is
forehead, while the inverting lead is placed shown in Figure 34–11.
on the belly of the SCM muscle.27 In this
orientation, the first positive peak will be plot- Analysis
ted downward. Beyond the polarity difference,
there are subtle changes in VEMP waveform Following the labeling conventions of Akin and
using these two methods. Clinically, these dif- Murnane,25, 26 the first positive polarity peak in
ferences have not proven to be important to the composite average was labeled P1 (some
date. Electrode position over the SCM muscle authors refer to this as P13), and the following
is an important variable regardless of electrode negative peak was labeled N1 (also known as
montage employed. To decrease variability, N23). From each composite average, P1 and
Figure 34–11. Example of VEMP waveforms. Five subaverages (top) were obtained from electrodes placed at FPz (non-
inverting) and over the belly of the SCM muscle (inverting). The subaverages were averaged together and measurements
were made for P1 (16.3 ms), N1 (24.9 ms), and the P1–N1 amplitude (292 μV). Amplitude asymmetry ratios, absolute
latencies, and interaural latency differences are compared using normal limits in Table 34–3.
N1 peak latencies and the P1–N1 amplitude established for each ear, as the patient muscle
are measured. Amplitude and latency asym- strength allows.
metries between right and left sides are also
compared using the following asymmetry ratio
calculation: NORMAL VALUES
As mentioned above, recording technique
Asymmetry ratio(%) =
influences VEMP parameters. Table 34–2 sum-
(Right P1-N1 amplitude − Left P1-N1 amplitude) marizes the recording method used at the
× 100
(Right P1-N1 amplitude + Left P1-N1 amplitude) Mayo Clinic at the time of this writing.
Table 34–3 shows the 95% limits for each
This ratio is only calculated using stimulus parameter using the Mayo method. Amplitude
intensities greater than 90 dBnHL to ensure and absolute latency distributions were loga-
a suprathreshold stimulus. Threshold, defined rithmically distributed. Thus the 95% limits
as the lowest stimulus intensity level at which were calculated on transformed data. Abso-
one can reliably identify P1 and N1, is also lute amplitude measures show a great deal of
Table 34–2 VEMP Recording Parameters—Mayo Clinic Florida Protocol

Stimulus
Frequency 500 Hz tone burst (2–0–2 rise–plateau–fall)
Rate 5.1 burst/second
Presentation Monaural through ER-3A inserts
Recording
Montage FPz (+) to SCM muscle belly (−); ground = forehead
Filter 2–250 Hz bandpass
Amplification ×5000
Averaging
Window 100 ms
Prestimulus 25 ms
Epochs/Average 40–120
Averaging Stop Rule 120 epochs or when P1–N1 amplitude is three times greater than any
wave in the prestimulus interval
Superaverages Measurements taken from superaverage (3–6 subaverages)
Table 34–3 VEMP Normal Values—Mayo Clinic Florida Protocol
Patient age <30 30–39 40–49 50–59 60–69 70–79 80+ Age independent
limit
P1
Mean (ms) 16 16.1 16.2 16.3 16.4 16.6 16.8
95% absolute latency 13.8–18.4 13.9–18.5 14.1–18.6 14.2–18.7 14.4–18.8 14.5–18.9 14.8–19.1
limits (ms)
95% interaural 2.3
asymmetry limit (ms)
N1
Mean (ms) 24.0
95% absolute latency 20.8–27.6
limits (ms)
95% interaural 3.5
602
asymmetry limit (ms)

P1–N1 Amplitude
Mean (μV) 183 150 138 126 116 97
95% (μV) 58–578 42–535 40–471 38–416 37–367 33–285
95% interaural 40.2
asymmetry limit (%)
Threshold
Mean (dB HL) 76.6 78.3 80.1 82 83.9 85.8 87.8 70–100
95% (dB HL) 70–96 70–98 70–100 70–100 70–100 70–100 70–100
95% interaural 11.4
asymmetry limit
(dB HL)
Note: 95% confidence limits for P1, N1, P1–N1 amplitude and VEMP threshold by age, base of 495 normal studies (Olsholt and Zapala, in process). Distributions were logarithmically
distributed. Confidence limits were calculated based on transformed distribution estimates. The 95% interaural latency asymmetry limits represent the maximum expected latency
difference between ears. The 95% interaural amplitude asymmetry (P1–N1 amplitude) is calculated using the VEMP amplitude asymmetry ratio.
variation in younger age groups, which will Somatosensory information is the dominant
have implications for identifying pathologic input, followed by visual and vestibular inputs.
conditions associated with large VEMP ampli- The inputs from these three systems are inte-
tudes (such as superior SCC dehiscence—see grated, analyzed, and incorporated into a com-
Superior SCC Dehiscence section). plex network by the CNS for maintenance
of balance. For many years, physicians have
used subjective methods for assessing a per-
LESION EFFECTS
son’s ability to maneuver and to maintain bal-
Hearing impairment may affect VEMP results ance, with and without vision. Tests such as
depending on the type of hearing loss. Even the Romberg and tandem gait tests are two
subtle forms of conductive hearing loss can examples.
cause the VEMP to be absent or demon- CDP provides quantitative information on
strate reduced P1–N1 amplitude. In con- how a patient uses sensory information to
trast, sensorineural hearing loss will have maintain upright stability when standing. It
no effect on VEMP amplitude or latency. also measures the performance of automated
Lesions within the vestibular labyrinth will motor responses typically employed to avoid
cause decreased VEMP P1–N1 amplitudes and a fall when upright balance is disturbed.
elevate thresholds if saccule or inferior vestibu- The quantitative information developed dur-
lar nerve structures are involved. Lesions ing CDP testing can be used to describe the
affecting the superior vestibular nerve or asso- types of problems patients have with balance
ciated labyrinthine structures will have lit- in day-to-day activities. As such, it is helpful in
tle effect on the VEMP. Retrolabyrinthine determining deficit areas that might betray the
lesions, particularly focused in the cerebel- presence of a disorder, or might be targeted for
lopontine angle, may increase VEMP peak rehabilitation.
latencies.
Purpose and Role of CDP
Key Points
• Quantifies the contribution of sensory
• VEMPs reflect the function of the sac- information (visual, vestibular, and somato-
cule and inferior vestibular nerve branch sensory) and automatic motor responses
of cranial nerve (CN) VIII. in controlling upright stance.
• Conductive hearing loss can cause the
air-conducted VEMP to be absent. It This test consists of two major components,
must be excluded before abnormal VEMP each containing subtests. The first component
results can be interpreted. is a test for motor control to maintain balance.
• Sensorineural hearing loss (cochlear nerve
The second component is a test for measur-
branch) and lesions involving the supe- ing the patient’s use of sensory information as
rior vestibular nerve branch or related it relates to maintaining balance. The patient’s
structures will have no effect on VEMP anterior–posterior and lateral sway is moni-
results. tored by measuring vertical force with strain
• Abnormal VEMP results from labyrinthine
gauges that are mounted underneath the two
lesions affect VEMP P1–N1 amplitude platforms on which the patient stands (one
and VEMP threshold. foot on each platform). By analyzing the forces
• Abnormal VEMP results from retro-
developed over the platforms while the patient
labyrinthine lesions may also include peak stands, estimates of where they maintain their
latency delays. center of mass over their feet (base of support)
can be established. The relationship between
where a patient’s center of mass is held over
Computerized Dynamic their base of support allows for calculation of
Posturography sway. Additionally, how ankle and hip move-
ments contribute to maintaining upright stance
Balance is a complex function that requires can be estimated by measuring shear forces on
input from three major sensory systems. the platform.
CDP protocols are designed to challenge MOTOR CONTROL TEST

a patient’s ability to stand in the face of an
With the motor control test (MCT), the patient
unstable support surface or inaccurate sensory
is presented with forward and backward
data. The platform can move forward or
backward to produce perturbations of small, platform perturbations. The stretch reflex and
later occurring extra pyramidal activity work
medium, or large magnitude. It can also tilt up
or down, rotating at the axis below the patient’s to bring the center of mass back over the
ankle. The visual surround can be made to sway patients’ base of support. Loss of sensation
in the lower limbs, such as would occur with
in the same anterior–posterior direction as
the patient (sway-referenced visual surround). peripheral neuropathy, or aberrant reflexive
The platform can also be made to tilt con- movements, such as would occur with lower
motor neuron disease, may delay or abolish
currently with the patient’s anterior–posterior
sway (sway-referenced platform movement). normal corrective movements. Higher level
This forces the patient to ignore or to compen- lesions, involving the brain stem, cerebellum,
or cerebral motor pathways, may also produce
sate for the adverse sensory stimulation.
The results of the test are compared with abnormal responses, reflected in later occur-
those from control subjects matched for age. ring movements. Computer algorithms calcu-
Humans have a cone of stability of approx- late the latency of the corrective response to
imately 12.5◦ for anterior–posterior sway22 each perturbation. The symmetry and amount
of force and weight distribution are also mea-
(Fig. 34–12). Thus, a person who sways approx-
sured and displayed with the MCT.
imately 12.5◦ is at the limits of stability and is
In the test for adaptation, toes up or down,
likely to fall. A fall is scored any time a patient
the platform tilts up or down 8◦ to gener-
reaches out and touches the visual surround to
keep from falling or anytime he or she moves ate a stimulus analogous to that of walking on
the feet to keep the center of gravity over the uneven surfaces. It is expected that patients
base of support. will perform poorly on trial 1, but on tri-
als 2–5, they should adapt and perform nor-
mally. Abnormal performance on adaptation
testing can occur when ankle range-of-motion
12.5° is limited, tactical and proprioceptive inputs
from the lower limbs are disturbed, or when
patients are severely deconditioned. Cognitive
changes and movement disorders may also pro-
duce impaired performance, largely because
the effect of the initial stretch reflex serves
to move the patients’ center of mass further
off of their base of support. Volitional effort
is required to counter this effect. Poor per-
formance on this test contributes further evi-
dence that the patient under study may not
be able to coordinate ordinary recovery move-
ments on uneven surfaces, regardless of the
cause.
SENSORY ORGANIZATION TEST

The sensory organization test (SOT) consists
of six different conditions, with three trials
Center of area of foot support
possible for each condition. These conditions
Figure 34–12. The cone of stability is 12.5◦ and is the are divided into stable support surface trials
amount of sway against which anterior–posterior sway is
judged. (From EquiTest Systems. Installing EquiTest ver-
(conditions 1, 2, and 3) and unstable (sway-
sion 4.03. NeuroCom International, Clackamas, OR. By referenced) support surface trials (conditions
permission of NeuroCom International.) 4, 5, and 6). In conditions 1 and 4, the visual
A. Equilibrium score
Visual condition
100
Sway-
Fixed Eyes closed
referenced
75
Support condition
Fixed
1 2 3 50
25
NN NN
/ / / /
reference
SS SS
Sway-
Fall
4 5 6 1 2 3 4 5 6 Composite
84
Sensory conditions
Figure 34–13. The six sensory conditions of the SOT.

(From EquiTest Systems. Installing EquiTest version 4.03. B. Sensory analysis
NeuroCom International, Clackamas, OR. By permission
of NeuroCom International.) 100
surround is stable. In conditions 2 and 5, 75

the eyes are closed (vision denied). In con-
ditions 3 and 6, the eyes are opened, but
the visual surround is sway-referenced. Recall
that sway-referenced means that the visual 50
surround or platform is driven by the patient’s
anterior–posterior sway. Thus visual informa-
tion is inaccurate in conditions 3 and 6, and
25
the patient must either ignore or compensate
for the inappropriate information to remain
standing. These conditions are summarized in
Figure 34–13. 0
Test results for conditions 1–6 for a nor- SOM VIS VEST PREF
mal subject are shown in Figure 34–14. Scores Figure 34–14. A, Sensory organization equilibrium
approaching 100 indicate little sway, whereas scores of a normal subject for the six test conditions
those near zero represent a large amount of and the composite score. The latter is the mean of
sway (relative to the 12.5◦ of the cone of sta- 14 scores (the mean of conditions 1 and 2, and all 12
trials of conditions 3–6). N/S, no score (trial was not
bility). If a patient touches the wall or moves run). B, Normal sensory analysis scores. PREF, visual
the feet, the trial is scored as a fall (zero preference; SOM, somatosensory; VEST, vestibular; VIS,
points). Scores falling into the shaded area are visual. Results in shaded areas are abnormal. (From
abnormal (Fig. 34–14A). From these scores, a EquiTest Systems. Installing EquiTest version 4.03.
NeuroCom International, Clackamas, OR. By permission
sensory analysis is performed (Fig. 34–14B). of NeuroCom International.)
The scores (ratios) are derived from com-
parisons of the various conditions, as shown
in Figure 34–15. This information provides a performance on the SOT requires volitional
fairly complete picture of a person’s ability to movement. Poor performance thus reflects the
maintain stability. patient’s willingness or ability to use sensory
Abnormal results typical of patients with information for maintaining upright stance.
acute vestibular disorders are low scores on For example, patients with Parkinson’s dis-
conditions 5 and 6, with an abnormal vestibular ease will often perform poorly on conditions 4,
ratio. Patients who are unable to suppress inap- 5, and 6—not necessarily because they expe-
propriate visual information function poorly rience a sensory deficit, but rather because
on conditions 3 and 6. This has been they have a movement ignition deficit that is
demonstrated in patients with Alzheimer’s dis- challenged by the unstable platform in these
ease. It is important to keep in mind that conditions. Poor performance on conditions
Sensory analysis
Ratio Test Ratio
Significance
name conditions pair
Question: Does sway increase when

SOM Condition 2 visual cues are removed?
Somatosensory Condition 1
Low scores: Patient makes poor use of
somatosensory references.
2 1

VIS somatosensory cues are
Condition 4
Visual inaccurate?
Condition 1
visual references.
4 1
visual cues are removed
and somatosensory cues
VEST Condition 5
are inaccurate?
Vestibular Condition 1
vestibular cues, or vestibular
5 1 cues are unavailable.
Question: Do inaccurate visual cues

result in increased sway
PREF compared to no visual cues?
Condition 3+6
Visual
Condition 2+5
Preference Low scores: Patient relies on visual cues
even when they are
3+6 2+5 inaccurate.
Figure 34–15. Summary of sensory analysis for six sensory conditions and the significance of their outcomes. See
Figure 32–13 for abbreviations. (From EquiTest Systems. Installing EquiTest version 4.03. NeuroCom International,
Clackamas, OR. By permission of NeuroCom International.)
other than 5 and 6 should alert the clini- conditions 1 and 2. Results such as these are
cian that problems beyond the vestibular sys- physiologically inconsistent.
tem might be contributing to imbalance or
dizziness. Key Points
SOT performance is not diagnostic. There
is no clear relationship between the functional • CDP quantifies the contribution of sen-
performance and the presence of specific dis- sory and motor behaviors and abilities on
ease entities. However, the SOT does demon- functional balance.
strate a patient’s ability to use sensory infor- • Motor control and adaptation tests quan-
mation for maintaining postural control and tify automatic corrective movements to
reflects fall risk.30, 31 Information from this test bring the patient’s center of mass back
is helpful in identifying when patients have over the base of support following plat-
poor balance and can lead to appropriate reha- form perturbations or tilts.
bilitation strategies. • The SOT measures the patient’s ability or
One of the most effective uses of CDP is willingness to use sensory information for
testing patients who have functional abnormal- maintaining upright stance.
ities or at least functional overlays.32 Because • CDP is particularly useful in detecting
patients must cooperate to complete the test, functional behaviors that cannot be con-
they have ample opportunity to exaggerate sistent with organic disease.
their responses. Of interest, many patients with • Although not diagnostic, CDP is helpful in
functional problems perform relatively better identifying deficit areas that might betray
on the difficult tests, such as conditions 5 the presence of a disorder, or might be
and 6, and poorly on the easier tests, such as targeted for the need for rehabilitation.
CLINICAL APPLICATIONS OF schwannoma were calculated for each syn-

VESTIBULAR TESTING: drome type. The base rate (prevalence) for
ASSESSING SENSORINEURAL vestibular schwannoma during the study inter-
val was 1.27%.
SYNDROMES OF THE
LABYRINTH
Superior Nerve Syndrome
In patients complaining of vestibular symp-
toms, the test methods described above may The superior nerve syndrome is a common
be used to provide evidence of central com- syndrome, characterized by singular impair-
pensation status. Examples include most forms ment of structures associated with the supe-
of positional nystagmus, phase and asymme- rior vestibular nerve (the Lateral and Superior
try measurements on rotary chair slow har- semicircular canal ampullae and the Utricle).
monic acceleration, and performance on the Test results indicate a unilateral caloric weak-
sensory organization subtest of CDP. Other ness with normal audiological and VEMP stud-
tests, such as the bilateral, bithermal caloric ies. The prototypical condition for this syn-
test, or the VEMP test, offer direct measure- drome would be vestibular neuronitis (also
ments of superior and inferior vestibular nerve known as labyrinthitis or superior nerve
output. When these latter tests are combined neurolabyrinthitis—see example description in
with the audiological evaluation, the function the Introduction section of this chapter). The
of the three main branches of CN VIII can be superior nerve syndrome is strongly associ-
better assessed. ated with the complaint of vertigo and sub-
It can be profitable to group conditions jective handicap (as measured on the dizzi-
based on the function of these three nerve ness handicap inventory). However, relative to
branches: (1) the superior vestibular, (2) infe- other vestibular syndromes, performance on
rior vestibular, and (3) cochlear eighth nerve CDP SOT tends to be less impaired. Two
branches. Some disease processes tend to important associations were made with this
affect some branches while sparing others. syndrome. First, this syndrome is highly corre-
Further, at least preliminarily, we have been lated with subsequent development of BPPV
able to show relationships between patient’s (presumably due to utricular damage). Sec-
subjective complaints, postural control ability, ond, there was a very low association with
and the pattern of labyrinthine dysfunction. vestibular schwannoma. In 1253 consecutive
These syndrome classes, with the exception of patients referred for auditory and vestibular
the Superior Semicircular Canal Dehiscence testing, no cases of vestibular schwannoma
syndrome, were developed as part of a ret- were found in patients with superior nerve
rospective study of 1578 consecutive patients syndrome.
seen at the Mayo Clinic Florida for vestibu-
lar testing between 2002 and 2004.33 Patients
included in this study had these syndromes as Superior SCC Dehiscence
the sole explanation of their dizziness or imbal-
ance complaints. Patients with comorbidities Minor35 reported 17 cases of a dehiscence
that could have contributed to their complaints of the superior SCC identified with high-
of dizziness or imbalance were excluded. From resolution computed tomography. The patients
these data, correlations between patient’s had vertigo, oscillopsia, or both when presented
complaints and test results, medical diagno- with intense sounds or stimuli that produced
sis, SOT performance, co-occurring BPPV, and changes in middle ear or intracranial pres-
subjective report of dizziness handicap (as sure. These stimuli produced torsional eye
measured on the Dizziness Handicap Inven- movements commensurate with stimulating
tory34 ) were tallied. the affected canal. The VEMP contributes to
A second unpublished retrospective study, this diagnosis in that the P1–N1 amplitude
looking at vestibular schwannoma presenta- tends to be abnormally large on the involved
tions, was performed by analyzing 1253 consec- side and the threshold may be abnormally
utive cases seen in 2005 through 2006.28 In this low (<65 dBnHL using the Mayo protocol).
second study, likelihood ratios for vestibular Although the exact mechanism for this effect
is not definitively understood, it appears that postural control as measured on SOT are
some of the pressure exerted by the stapes associated with this syndrome. As might be
footplate in response to sound is shunted anticipated, BPPV on the involved side is
through the opening in the superior SCC. not associated with this condition. Vestibular
Along the way, the shunted pressure wave per- schwannoma can present as a posterior syn-
turbs the saccule. The audiological evaluation drome, but it is apparently rare. No cases were
may also assist in recognizing this condition.36 observed in 1253 consecutive cases. One case
There tends to be a conductive hearing loss was observed in the earlier 2003–2004 study.
on the involved side, with paradoxically present
acoustic reflexes present. Surgical plugging of
the dehiscence improves patients’ symptoms in Split Syndrome
most cases.37
The split syndrome is characterized by an
abnormal caloric weakness and co-occurring
Basement Syndrome hearing loss on the same side. Episodic ver-
tigo, as in Meniere’s syndrome, was strongly
In basement syndrome, there is unilateral hear- associated with this pattern. Patients were typ-
ing loss and an abnormal VEMP on the same ically observed between episodes. Vestibular
side. The implication is involvement of struc- schwannoma also presented with this group
tures associated with the cochlear and infe- (likelihood ratio = 2.1:1). Subsequent develop-
rior vestibular nerve branches. These nerves ment of BPPV was also common.
fill the inferior partition of the internal audi- This syndrome is intriguing for several rea-
tory canal and thus the designation of “base- sons. First, it is difficult to explain the syn-
ment syndrome.” Vertigo is not a strong com- drome based on proximity of nerve branches.
plaint in this syndrome. Rather patients tend Distally the superior vestibular nerve and
to complain of lightheadedness, heavy headed- cochlear nerve are separated by a bony shelf.
ness or vague, nondescript sensations. Dizzi- At the CN VIII root, the nerve branches are
ness Handicap Inventory scores tend to indi- in close proximity, but not any more so than
cate less self-reported handicap than observed the inferior vestibular nerve branch. So the
with syndromes involving the superior nerve. sparing of the VEMP is problematic. Similarly,
Yet there is a tendency for the group to have from a membrane point of view, the saccule
lower SOT scores (poorer balance). BPPV is sits between the cochlea and the pars superior
not commonly encountered in this group. (where the utricle and SCCs reside). Again, the
Approximately 50% of vestibular schwan- saccule appears spared, despite its interposi-
nomas emanate from the inferior vestibular tion between the cochlea and the pars superior.
nerve branch. However, only one vestibu- One possible explanation is that the VEMP
lar schwannoma case in 1253 patients was is a robust response that may persist despite
found to present with a basement syndrome. subcritical damage to the saccule.
A second patient with vestibular schwannoma The complaint of severe vertigo was com-
presented with an isolated abnormal VEMP mon in this group. Yet, caloric testing repli-
latency delay. Thus, the likelihood ratio for cated subjective complaints in less than 31% of
vestibular schwannoma given a basement syn- cases. Episodic vertigo cases represented more
drome was low (0.54:1). than 50% of cases in this group. It is possible
that the episodic vertigo patients experienced
vertigo so severe during their episodes that the
Posterior Syndrome vertigo provoked by the standard caloric test
was no longer provocative.
Posterior syndrome occurs when there is evi-
dence for both abnormal horizontal canal or
superior vestibular nerve and saccule or infe- Global Syndrome
rior vestibular nerve function. VEMP and
caloric responses are abnormal on the involved As the name implies, this syndrome is asso-
side. Strong complaints of vertigo and impaired ciated with hearing loss and co-occurring
abnormal VEMP and caloric responses. This underpinning the loss of function, and an
implies total involvement of CN VIII. It understanding of what sensory functions are
is rarely encountered in general practice. available, salvageable, and permanently lost.
While many disease states were found in this Formal testing will be necessary to accomplish
group, including idiopathic, viral labyrinthi- this.
tis, and Meniere’s syndrome, the likelihood of
a vestibular schwannoma in this group was Key Points
very high (likelihood ratio = 9.5:1). In gen-
eral, it seems that the probability of vestibular • Vestibular disorders can be based on the
schwannoma increases as the number of CN function of these three nerve branches—
VIII branches involved increases. This high- the superior vestibular, inferior vestibular,
lights the value of the VEMP. Without the and cochlear eighth nerve branches.
VEMP test, global and split syndromes appear • Audiological tests assess cochlear nerve
identical on vestibular testing. The likelihood branch function; bilateral, bithermal caloric
of vestibular schwannoma for patients who had tests assess superior vestibular nerve func-
hearing loss and caloric unilateral weakness tion and the ascending VOR pathways
(without considering VEMP results) was 5.0:1. in the brain stem; VEMP tests assess
So, the VEMP test helps stratify risk for retro- the inferior vestibular nerve branch and
labyrinthine involvement. To place this into descending VSR pathways in the brain
context, auditory evoked brain stem responses stem.
were estimated to have a likelihood ratio of • The superior nerve syndrome is a com-
6.4:1 in this same group. So being classified mon syndrome, characterized by singular
as having a global vestibular syndrome carried impairment of structures associated with
greater risk for vestibular schwannoma than an the superior vestibular nerve (the Lateral
abnormal evoked potential study. and Superior SCC ampullae and the Utri-
Three principles are suggested so far. First, cle). Vertigo and subsequent BPPV are
syndromes that reflect a superior vestibular commonly associated with this syndrome.
nerve distribution are associated with com- • Dehiscence of the superior SCC causes
plaints of vertigo. When inferior nerve func- vertigo, oscillopsia, or both when pre-
tion is present, these syndromes carry a higher sented with intense sounds or stimuli
probability for secondary BPPV. that produced changes in middle ear or
Second, syndromes that reflect an inferior intracranial pressure.
vestibular nerve distribution are less likely to • In basement syndrome, there is unilat-
be associated with complaints of vertigo and eral hearing loss and an abnormal VEMP
are only rarely associated with BPPV. There on the same side. Vertigo and subsequent
was a statistical trend for poorer performance BPPV are not associated with this condi-
on SOT, implying more difficulty with postu- tion. Fall risk is potentially increased.
ral control tasks. Variability on SOT was great • Posterior syndrome occurs when there is
however, and so poor performance may not be evidence for abnormal horizontal canal or
evident in individual cases. superior vestibular nerve and saccule or
Third, the addition of the VEMP test inferior vestibular nerve function. VEMP
improves the ability to recognize retro- and caloric responses are abnormal on the
labyrinthine involvement. As more branches involved side.
of CN VIII show involvement, the risk of • The split syndrome is characterized by
vestibular schwannoma increases. The global an abnormal caloric weakness and co-
syndrome, where all three CN VIII branches occurring hearing loss on the same side.
are involved, holds the greatest risk. • Global syndrome is associated with hear-
The syndrome approach mentioned here ing loss and co-occurring abnormal VEMP
holds promise for improving our ability to and caloric responses. This implies total
explain and anticipate patient complaints. Ulti- involvement of CN VIII. Global syndrome
mately, natural course of the problematic groups have a higher risk of vestibular
symptoms and fall risks will likely require schwannoma relative to other syndrome
both an understanding of the disease process types.
VESTIBULAR REHABILITATION vestibular reflexes. First, there is a relationship

between damaged sensory epithelia within the
The management of vestibular lesions can be membranous labyrinth and abnormal reflex-
medical, surgical, or rehabilitative. Medical ive behavior. However, CNS compensation will
treatments help with symptomatic care dur- modify and often extinguish abnormal reflexive
ing acute vestibular crisis and have a role in behavior on the standard physical examina-
prophylaxis. Surgery may help in those rare tion. When vestibulopathy is a possibility and
cases of perilymphatic fistula, superior canal the physical examination is unrevealing, the
dehiscence, or perhaps endolymphatic sac pro- electrophysiological measures described in this
cedures. However, in the vast majority of cases, chapter may be informative.
the CNS’s own adaptive abilities will play a ENG or VNG has been historically the
larger role in ameliorating troubling vestibular most commonly employed measure of vestibu-
symptoms. lar function. The bilateral, bithermal caloric
Vestibular rehabilitation has been available test provides a direct measure of vestibular
for many years. However, carefully developed, weakness involving the horizontal canal and
personalized treatments have been shown the superior vestibular nerve branch. As such,
to substantially improve compensation.17, 22, 31 it has localizing value. Additionally, screen-
These programs consist of habituation exer- ing for positioning induced nystagmus is help-
cises, postural control exercises, general con- ful in detecting benign paroxysmal positional
ditioning, and psychological support when vertigo. Gaze and positional nystagmus tests
needed. This relatively new area has begun help differentiate between vestibular-induced
to provide a mechanism for helping to treat spontaneous nystagmus and gaze nystagmus
the condition of patients who previously were from CNS involvement. Visually guided eye
told to live with their balance problem. Appro- movements may help characterize saccadic,
priate exercises prescribed by a physical ther- pursuit, or optokinetic deficits that may also
apist trained in vestibular rehabilitation can have been appreciated during the physical
help speed and improve the recovery of examination.
many patients. Appreciating what structures Rotary chair testing is particularly useful
are working, how they relate to subjective com- in detecting bilateral vestibular weaknesses.
plaints and function deficits should ultimately Rotary chair test methods employ precisely
help refine vestibular rehabilitation treatment controlled vestibular and visual stimuli, and
plans even further. are repeatable over time. This makes rotary
chair protocols attractive when serial testing is
required. In older patients, bilateral vestibu-
SUMMARY lar weakness may be missed on the standard
caloric test. Rotary chair SHA testing is useful
This chapter has summarized five clini- when assessing elderly and pediatric patients
cally available electrophysiologic measures for who may not tolerate the standard caloric test.
patients who complained of vertigo and imbal- SVV measures offer a measure of utricular-
ance. These measures will seldom provide a induced ocular tilt effects. The utricle is closely
diagnosis. Careful history and physical exam- associated with control of upright stance. The
ination, coupled with an understanding of ability to have a direct measure of utricu-
vestibular anatomy, physiology, and pathophys- lar function is thus an important addition to
iology, remain the bedrock for diagnosing the vestibular battery. Repeated measurements
problems of vertigo and imbalance. The elec- using an appropriate psychophysical method
trophysiologic measures described in this chap- may be necessary for accurate measurements
ter are merely extensions of the physical of ocular tilt using this method.
examination. Test data cannot be interpreted VEMPs offer a measure of saccule and infe-
effectively without the same understanding of rior vestibular nerve branch function. Like the
vestibular anatomy, physiology, and pathophys- caloric test, the VEMP reflects the function
iology that underpins the initial history and of each ear in isolation. It thus has localizing
physical examination. value.
There are several key concepts for the CDP consists of the SOTs and the MCTs.
clinician to keep in mind when measuring These tests are not diagnostic in the sense that
they point to specific sites of lesion. Rather, 4. Walker, M. F., and D. S. Zee. 2000. Bedside vestibular
they measure behaviors that underpin nor- examination. Otolaryngologic Clinics of North Amer-
mal control of upright stance in the functional ica 33:495–506.
5. Ruckenstein, M. J. 1995. A practical approach to dizzi-
sense. The MCTs quantify reaction times for ness. Questions to bring vertigo and other causes into
compensatory movements. The SOT measures focus. Postgraduate Medicine 97:70–2, 75–8, 81.
a patient’s ability or willingness to use visual, 6. Baloh, R. W., and V. Honrubia. 2001. Clinical neuro-
vestibular, and somatosensory information to physiology of the vestibular system, xx, 408. New York:
Oxford University Press.
remain standing. Vestibular deficits become 7. Eggers, S. D., and D. S. Zee. 2003. Evaluating
apparent when the test situation minimizes the dizzy patient: Bedside examination and labora-
visual and somatosensory information (SOT tory assessment of the vestibular system. Seminars in
conditions 5 and 6). Performance deficits and Neurology 23:47–58.
other SOT conditions should alert the exam- 8. Leigh, R. J., and D. S. Zee. 2006. The neurology of
eye movements, x, 762. New York: Oxford University
iner that deficits beyond the vestibular system Press.
may be contributing to impaired balance. CDP 9. Furman, J. M., and S. P. Cass. 1999. Benign parox-
is particularly helpful in identifying functional ysmal positional vertigo. New England Journal of
performance patterns that cannot be explained Medicine 341:1590–6.
10. Parnes, L. S., S. K. Agrawal, and J. Atlas. 2003. Diag-
by organic impairment. nosis and management of benign paroxysmal posi-
Patients can be meaningfully classified into tional vertigo (BPPV). Canadian Medical Association
vestibular syndromes based on bilateral, bither- Journal 169:681–93.
mal caloric test results, VEMP test results, 11. Epley, J. M. 1992. The canalith repositioning proce-
and audiological evaluation results. These dure: For treatment of benign paroxysmal positional
vertigo. Otolaryngology and Head and Neck Surgery
tests reflect superior vestibular nerve, inferior 107:399–404.
vestibular nerve, and cochlear nerve branch 12. Epley, J. M. 2001. Human experience with canalith
function. Syndromes that involve the supe- repositioning maneuvers. Annals of the New York
rior vestibular nerve branch tend to provoke Academy of Sciences 942:179–91.
13. Prokopakis, E. P., T. Chimona, M. Tsagournisakis, et al.
stronger complaints of vertigo, and a ten- 2005. Benign paroxysmal positional vertigo: 10-year
dency for benign paroxysmal positional vertigo. experience in treating 592 patients with canalith repo-
Syndromes that involve the inferior vestibu- sitioning procedure. Laryngoscope 115:1667–71.
lar nerve branch are less likely to provoke 14. Richard, W., T. D. Bruintjes, P. Oostenbrink, and R. B.
strong complaints of vertigo and have a low risk van Leeuwen. 2005. Efficacy of the Epley maneu-
ver for posterior canal BPPV: A long-term, controlled
for benign paroxysmal positional vertigo. The study of 81 patients. Ear, Nose, & Throat Journal
risk for retrolabyrinthine involvement, such 84:22–5.
as vestibular schwannoma, increases with the 15. Steenerson, R. L., G. W. Cronin, and P. M. Marbach.
number of CN VIII branches involved. 2005. Effectiveness of treatment techniques in 923
cases of benign paroxysmal positional vertigo. Laryn-
Vestibular rehabilitation is an important goscope 115:226–31.
addition to medical and surgical treatments for 16. White, J., P. Savvides, N. Cherian, and J. Oas. 2005.
patients with complaints of vertigo and imbal- Canalith repositioning for benign paroxysmal posi-
ance. Understanding how vestibular reflexes tional vertigo. Otology & Neurotology 26:704–10.
are impaired helps anticipate functional deficits. 17. Herdman, S. 2000. Vestibular rehabilitation, xxix, 597.
Philadelphia: Davis.
Electrophysiological measures of vestibular 18. Brandt, T. 1990. Positional and positioning vertigo and
function can help in this regard. nystagmus. Journal of Neurological Sciences 95:3–28.
19. Shoman, N., and N. Longridge. 2007. Cerebellar ver-
mis lesions and tumours of the fourth ventricle in
patients with positional and positioning vertigo and
nystagmus. The Journal of Laryngology and Otology
REFERENCES 121:166–9.
20. Sunami, K., R. Tochino, T. Zushi, et al. 2004. Posi-
1. Baloh, R. W. 1995. Approach to the evaluation of tional and positioning nystagmus in healthy subjects
the dizzy patient. Otolaryngology and Head and Neck under videonystagmoscopy. Acta Otolaryngologica
Surgery 112:3–7. 124(Suppl 554):35–7.
2. Labuguen, R. H. 2006. Initial evaluation of vertigo. 21. Barber, H. O., and G. Wright. 1973. Positional nystag-
American Family Physician 73:244–51. mus in normals. Advances in Oto-Rhino-Laryngology
3. Traccis, S., G. F. Zoroddu, M. T. Zecca, T. Cau, M. A. 19:276–83.
Solinas, and R. Masuri. 2004. Evaluating patients with 22. Shepard, N. T., and S. A. Telian. 1996. Telian, practical
vertigo: Bedside examination. Neurological Science management of the balance disorder patient, xi, 221.
25(Suppl 1):S16–19. San Diego: Singular Publishing Group.
23. Cyr, D. G. 1991. Vestibular system assessment. In patients with normal ENG results. Ear, Nose, & Throat
Hearing assessment, ed. W. F. Rintelmann, 2nd ed., Journal 84:212–14.
777. Austin, TX: Pro-ed. 31. Whitney, S. L., G. F. Marchetti, and A. I. Schade.
24. Welgampola, M. S., and J. G. Colebatch. 2005. 2006. The relationship between falls history and com-
Characteristics and clinical applications of vestibular- puterized dynamic posturography in persons with bal-
evoked myogenic potentials. Neurology 64:1682–8. ance and vestibular disorders. Archives of Physical
25. Akin, F. W., and O. D. Murnane. 2001. Vestibular Medicine and Rehabilitation 87:402–7.
evoked myogenic potentials: Preliminary report. Jour- 32. Mallinson, A. I., and N. S. Longridge. 2005. A new
nal of the American Academy of Audiology 12:445–52; set of criteria for evaluating malingering in work-
quiz 491. related vestibular injury. Otology & Neurotology 26:
26. Akin, F. W., O. D. Murnane, and T. M. Proffitt. 686–90.
2003. The effects of click and tone-burst stimulus 33. Zapala, D. A., and K. Hibbits. 2004. Sensorineural
parameters on the vestibular evoked myogenic poten- Syndromes of the Vestibular Labyrinth. In: American
tial (VEMP). Journal of the American Academy of Academy of Audiology’s 16th Annual Convention &
Audiology 14:500–9. Expo, Salt Lake City, Utah, 2004.
27. Zapala, D. A., and R. H. Brey. 2004. Clinical expe- 34. Jacobson, G. P., and C. W. Newman. 1990. The devel-
rience with the vestibular evoked myogenic poten- opment of the Dizziness Handicap Inventory. Archives
tial. Journal of the American Academy of Audiology of Otolaryngoly—Head and Neck Surgery 116:424–7.
15:198–215. 35. Minor, L. B. 2000. Superior canal dehiscence syn-
28. Zapala, D. A. 2007. The VEMP: Ready for the clinic. drome. American Journal of Otology 21:9–19.
The Hearing Journal 60:10–20. 36. Minor, L. B., J. P. Carey, P. D. Cremer, L. R. Lustig,
29. Wang, C. T., and Y. H. Young. 2006. Comparison of S. O. Streubel, and M. J. Ruckenstein. 2003. Dehis-
the head elevation versus rotation methods in eliciting cence of bone overlying the superior canal as a cause
vestibular evoked myogenic potentials. Ear & Hearing of apparent conductive hearing loss. Otology & Neu-
27:376–81. rotology 24:270–8.
30. Sataloff, R. T., M. J. Hawkshaw, H. Mandel, A. B. 37. Banerjee, A., A. Whyte, and M. D. Atlas. 2005. Supe-
Zwislewski, J. Armour, and S. Mandel. 2005. Abnormal rior canal dehiscence: Review of a new condition.
computerized dynamic posturography findings in dizzy Clinical Otolaryngology 30:9–15.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART F
Autonomic Function
The autonomic nervous system regulates vis- autonomic nervous system—from the hypotha-
ceral function and the internal environment lamus to the autonomic axons in the trunk and
of the body through its effects on the heart, limbs. Autonomic function is not measured as
gut and other internal organs, peripheral blood frequently as it should be. With better under-
vessels, and sweat glands (Chapter 35). Auto- standing of the clinical importance of mea-
nomic dysfunction has important implications suring autonomic function and with increas-
for health and disease yet is clinically under ing use of newly available tests of cardiovagal
recognized. Clinical signs of autonomic dysfunction, segmental sympathetic reflexes, pos-
function are easily overlooked, and neural tural hemodynamics, and power spectral anal-
activity in the autonomic nervous system is ysis, the tests and measurements of autonomic
difficult to record directly. Although sympa- function will be of greater benefit in patient
thetic nerve function in peripheral nerves can care.
be recorded with fine-tipped tungsten elec- Pain is mediated mainly through small
trodes, this technique is difficult to apply clin- nerve fibers, particularly in the autonomic ner-
ically. Therefore, the assessment of autonomic vous system. Measurements of their function
function depends primarily on measuring the can help elucidate the mechanisms under-
response of the autonomic nervous system to lying pain, especially peripheral pain. The
external stimuli. emerging modalities for assessment of pain
The measurements of sweating (Chapters 36 pathways include quantitative sensory tests,
and 38), cardiovascular activity and peripheral autonomic tests, microneurography, and laser-
blood flow (Chapters 37 and 39), and central evoked potentials (Chapter 40). Direct record-
autonomic-mediated reflexes provide insight ing of spontaneous electric activity in nerves
into the broad range of disorders that affect by microneurography is tedious but can be
the central and peripheral components of the particularly helpful.
615
Chapter 35
Autonomic Physiology
INTRODUCTION MUSCLE SYMPATHETIC

SYMPTOMS AND DISEASES ACTIVITY
GENERAL ORGANIZATION OF AUTONOMIC CONTROL OF
THE AUTONOMIC SYSTEM HEART RATE
Visceral Afferents CARDIOVASCULAR
Visceral Efferents REFLEXES
Arterial Baroreflexes
SYMPATHETIC FUNCTION Cardiopulmonary
Functional Anatomy of the Reflexes
Sympathetic Outflow Venoarteriolar Reflexes
Assessment of Sympathetic Ergoreflexes
Function in Humans
MAINTENANCE OF POSTURAL
SYMPATHETIC INNERVATION OF NORMOTENSION
THE SKIN SUMMARY
INTRODUCTION and coordinates such physiological functions

as blood pressure and heart rate, respiration,
The autonomic nervous system is essential to body temperature, sweating, lacrimation, nasal
maintaining physiologic homeostasis in health secretion, pupillary size, gastrointestinal motil-
and responding to illness. Its integrative func- ity, urinary bladder contraction, sexual physi-
tions coordinate input from peripheral and vis- ology, and blood flow to the skin and many
ceral afferent nerves to orchestrate a dynamic organs.
balance among organ systems. Its adaptive
functions react moment by moment to the var-
ious forms of stress the body experiences. It SYMPTOMS AND DISEASES
consists of sympathetic (thoracolumbar) and
parasympathetic (craniosacral) divisions. Both Disorders affecting the autonomic nervous
divisions contain general visceral afferent and system may be manifested by failure or hyper-
efferent fibers. The enteric nervous system, activity of one or many of the visceral effector
located in the wall of the gut, is considered a organs. Autonomic neuropathies that discon-
third division of the autonomic nervous system. nect central autonomic centers and auto-
The autonomic nervous system thus regulates nomic ganglia from their peripheral effectors
617
may result in deficits in autonomic function. or paraneoplastic pandysautonomia), or as a

Examples include orthostatic hypotension due slowly progressive disease (pure autonomic
to adrenergic failure, heat intolerance due failure).
to sudomotor failure, gastroparesis, hypotonic The presence of autonomic failure has
bladder, and erectile failure. Some auto- important implications for clinical manage-
nomic neuropathies result in localized deficits, ment as well as for prognosis. Disruption of
for example, Horner’s syndrome, Adie’s tonic autonomic function can influence the long-
pupil, and neurogenic bladder from sacral term risks of morbidity,1, 2 mortality,3–5 and
anterior horn cell involvement in poliomyeli- intraoperative mortality.6, 7
tis. Autonomic centers disconnected from Autonomic testing is indicated in patients
inhibitory influences may give rise to episodic (1) with any of the aforementioned symptoms
autonomic hyperfunction. Examples include suggesting autonomic failure, (2) with periph-
autonomic dysreflexia and hypertonic blad- eral neuropathies (particularly small fiber neu-
der following spinal cord trauma, diencephalic ropathies in which nerve conduction and elec-
syndrome following head injury, hypertensive tromyographic findings may be normal), or (3)
surges of baroreflex failure following irradi- with parkinsonism or cerebellar dysfunction
ation to the carotid sinuses, auriculotempo- in which MSA is suspected. Unlike the eas-
ral syndrome, and catecholamine storms in ily reproducible function of somatic motor or
pheochromocytoma. sensory nerves, autonomic nerve function is
Autonomic disturbances frequently accom- difficult to evaluate precisely in humans. In
pany neurologic illnesses affecting motor or general, evaluation of autonomic function has
sensory systems or may occur in isolation. The been restricted to noninvasive recordings of
presence of autonomic failure is sometimes heart rate, blood pressure, blood flow, or sweat
clinically obvious. More frequently, accurate production. The interpretation of the results
characterization, localization, and grading of of these tests may be difficult, because (1)
autonomic dysfunction require a careful his- the effector organs react slowly to variations
tory to elicit subtle symptoms, a neurological in neural input, (2) the interactions of sympa-
examination attentive to autonomic signs, and thetic and parasympathetic outputs at a single
testing in a clinical autonomic laboratory. target level are complex, and (3) autonomic
The differential diagnosis of autonomic responses are affected by pharmacologic,
failure includes the peripheral neuropathies hormonal, local chemical, and mechanical
(many of which involve autonomic fibers), cen- influences.
tral degenerative disorders, and medical disor- This chapter provides an overview of some
ders that impact the autonomic nervous sys- aspects of autonomic function that may help
tem. Peripheral autonomic failure frequently with interpreting the results of noninvasive
occurs in small fiber neuropathies, particularly autonomic tests commonly used clinically.
in diabetes mellitus and amyloidosis. Addi- Therefore, the focus is on sudomotor, cardio-
tionally, prominent autonomic dysfunction can vascular, and adrenergic functions. Gastroin-
occur in autoimmune diseases, such as autoim- testinal, bladder, and sexual functions are not
mune autonomic ganglionopathy associated discussed.
with antibodies to the ganglionic acetylcholine
receptor, Guillain–Barré syndrome, Sjögren’s
Purpose and Role of Autonomic Testing
syndrome, Lambert–Eaton myasthenic syn-
drome associated with antibodies to voltage- • Recognize the presence, distribution, and
gated calcium channels, and paraneoplastic severity of autonomic dysfunction.
disorders such as lung carcinoma associated • Recognize patterns of autonomic failure
with antineuronal nuclear antibody (ANNA-1 that can be related to specific syndromes,
or anti-Hu). Autonomic disturbances are com- for example,
mon also in botulism, diphtheritic neuropathy, ◦ Generalized autonomic failure
and Chagas disease. Central neurodegenera- ◦ Orthostatic hypotension
tive disorders associated with autonomic fail- ◦ Orthostatic intolerance
ure include multiple system atrophy (MSA), ◦ Reflex syncope
dementia with Lewy bodies, and Parkinson’s ◦ Distal small fiber neuropathy
disease. Isolated autonomic failure may occur ◦ Peripheral neuropathies that may have
acutely or subacutely (e.g., in inflammatory an autonomic component
Autonomic Physiology 619
◦ Neurodegenerative disorders that may GENERAL ORGANIZATION OF

have an autonomic component THE AUTONOMIC SYSTEM
◦ Organ-specific autonomic syndromes
such as gastroparesis that may have a The general organization of the autonomic ner-
more widespread autonomic component. vous system is composed of visceral afferent
• Recognize potentially treatable autonomic and efferent nerves (Fig. 35–1).
disorders.
• Recognize disorders that warrant further
evaluation. Visceral Afferents
• Diagnose benign autonomic disorders that
may mimic life-threatening conditions. Visceral receptors generally are slowly adapt-
• Quantitatively evaluate autonomic dys- ing mechanoreceptors or chemoreceptors that
function over time have a low level of spontaneous activity and are
◦ To define the progression or remission innervated by small myelinated and unmyeli-
of autonomic disease nated fibers.8 Visceral afferent signals may
◦ To assess the response to therapy. mediate local reflexes in peripheral organs or
Visceral
effectors
Cranial
Parasympathetic
preganglionic
ganglion
parasympathetic
neurons
ACh
NE
Sympathetic
ganglia
Preganglionic
sympathetic
neurons
(T1–L3)
Sacral
preganglionic Parasympathetic
parasympathetic ganglion
neurons
(S2–S4)
Figure 35–1. Organization of the sympathetic and parasympathetic outputs of the autonomic nervous system. ACh,
acetylcholine; NE, norepinephrine. (Permission from Mayo Clinic Medical Neurosciences, 5th ed., Mayo Clinic Scientific
Press 2008.)
provide collateral input to autonomic ganglion to the neurotransmitter. This phenomenon,

cells. Most visceral afferent fibers enter the called denervation supersensitivity, is evidence
central nervous system at a spinal or, in the of a lesion involving postganglionic neurons in
case of cranial nerves (CN), the medullary which the remaining neuroeffector receptors
level to initiate segmental or suprasegmen- are upregulated.
tal viscerovisceral reflexes or to relay visceral Activity of most autonomic effectors is mod-
information to higher centers.8–10 Spinal sym- ulated by dual, continuous sympathetic, and
pathetic afferents mediate visceral pain and parasympathetic influences.11, 12 Sympathetic
initiate segmental and suprasegmental viscero- control originates from preganglionic neurons
visceral reflexes. Visceral afferents that enter in segments T2–L1 of the spinal cord and pre-
at the medullary level are in the vagus (CN dominates in blood vessels, sweat glands, and
X) and glossopharyngeal (CN IX) nerves and cardiac muscle. Parasympathetic outflow orig-
synapse in the nucleus of the tractus solitarius inates from preganglionic neurons in nuclei
(NTS).9, 10 This nucleus is critically involved in of CNs III, VII, IX, and X and in seg-
respiratory and cardiovascular reflexes through ments S2–S4 of the spinal cord. It predomi-
its extensive connections with neurons in the nates in control of the salivary glands, sinoa-
so-called intermediate reticular zone of the trial node, gastrointestinal tract, and blad-
medulla.10, 11 der.11, 12 Sympathetic–parasympathetic interac-
tions are not simply antagonistic but are
functionally complementary. The sympathetic
Visceral Efferents and parasympathetic systems may interact at
several levels, including that of the central ner-
Sympathetic and parasympathetic autonomic vous system, autonomic ganglia, neuroeffector
outflow systems produce responses that are junction, and target organ.8–12
longer in latency and duration and more con-
tinuous and generalized than those mediated
by the somatic motor system. This reflects
important differences in the functional orga- Key Points
nization of autonomic and somatic efferents. • The autonomic nervous system consists of
Autonomic output involves a two-neuron path- sympathetic, parasympathetic, and enteric
way that has at least one synapse in an divisions.
autonomic ganglion.11 The preganglionic axons • Most visceral afferent fibers enter the
are small myelinated (type B) cholinergic central nervous system at a spinal or
fibers that emerge from neurons of the gen- medullary level to initiate segmental
eral visceral efferent column in the brain or suprasegmental viscerovisceral reflexes
stem or spinal cord and pass through periph- or to relay visceral information to higher
eral nerves to the autonomic ganglia.12 Post- centers.
ganglionic autonomic neurons send varicose, • Spinal sympathetic afferents mediate vis-
unmyelinated (type C) axons to innervate ceral pain and initiate segmental and
peripheral organs. Autonomic terminals con- suprasegmental viscerovisceral reflexes.
tain varicosities. • The vagus (CN X) and glossopharyngeal
At most sympathetic neuroeffector junctions (CN IX) nerves synapse in the NTS,9, 10
the primary postganglionic neurotransmitter is which is involved in respiratory and car-
norepinephrine, which acts on the various sub- diovascular reflexes.
types of α- and β-adrenergic receptors. In the • Sympathetic and parasympathetic auto-
sweat glands, sympathetic effects are mediated nomic outflow systems produce responses
principally by acetylcholine. The primary post- that are longer in latency and duration and
ganglionic neurotransmitter at all parasympa- more continuous and generalized than
thetic neuroeffector junctions is acetylcholine, those mediated by the somatic motor
which acts on the various subtypes of mus- system.
carinic receptors. Whereas the main conse- • Autonomic output involves a two-neuron
quences of denervation in the striated mus- pathway that has at least one synapse in an
cle are paralysis and atrophy, postganglionic autonomic ganglion.
efferent denervation produces an exaggerated • The preganglionic axons are small myeli-
response of the target when it is exposed nated (type B) cholinergic fibers that
emerge from neurons of the general vis- and pass through the white ramus commu-
ceral efferent column in the brain stem nicans of the corresponding spinal nerve to
or spinal cord and pass through peripheral reach the paravertebral sympathetic chain and
nerves to the autonomic ganglia. to synapse on neurons located in the paraver-
• Postganglionic autonomic neurons send tebral or prevertebral ganglia.16 Paravertebral
unmyelinated (type C) axons to innervate ganglia innervate all tissues and organs except
peripheral organs. those in the abdomen, pelvis, and perineum.
• At most sympathetic neuroeffector junc- Their postganglionic fibers destined for the
tions the primary postganglionic neuro- trunk and limbs follow the course of spinal
transmitter is norepinephrine. nerves or blood vessels or both. Spinal fibers
• In the sweat glands, sympathetic effects join the peripheral spinal (somatic) nerve
are mediated principally by acetylcholine. through the gray ramus communicans. These
• The primary postganglionic neurotrans- fibers provide vasomotor, sudomotor, and pilo-
mitter at all parasympathetic neuroeffec- motor input to the extremities and trunk. Sym-
tor junctions is acetylcholine. pathetic fibers are intermingled with somatic
motor and sensory fibers, and their distribution
is similar to that of the corresponding somatic
nerve. Most sympathetic fibers are destined for
SYMPATHETIC FUNCTION the hand and foot and are carried mainly by the
median, peroneal, and tibial nerves and, to a
lesser extent, the ulnar nerve.
Functional Anatomy of the
Sympathetic Outflow SYMPATHETIC REFLEXES
Preganglionic sympathetic neurons have a Unlike somatic motor neurons, sympathetic
slow, irregular, tonic activity that depends preganglionic neurons are not monosynapti-
mainly on multiple segmental afferent and cally driven by visceral, muscle, or cutaneous
descending inputs.13, 14 Preganglionic sympa- sensory input. These afferents converge on
thetic neurons are organized into specialized second-order neurons located in laminae I,
spinal sympathetic functional units that con- V, VII, and X of the spinal cord that project
trol specific target organs and are differen- to the sympathetic preganglionic neurons to
tially influenced by input from the hypotha- initiate segmental somatosympathetic and vis-
lamus and brain stem. Sympathetic functional cerosympathetic reflexes.13 Segmental reflex
units include skin vasomotor, muscle vasomo- pathways arising from somatic or visceral affer-
tor, visceromotor, pilomotor, and sudomotor ents activate ipsilateral local interneuronal net-
units.13–15 There are, for example, clear dif- works in laminae I, V, and VII, which may
ferences between the patterns of sympathetic directly activate or inhibit the sympathetic pre-
outflow to skin and to muscle, but sympathetic ganglionic neurons. With the exception of Ia
activity recorded simultaneously from differ- muscle spindle and Ib Golgi tendon organ
ent muscles at rest or from skin sympathetic afferents, activation of all other groups of pri-
nerves innervating the palms and the feet are mary sensory fibers arising from skin (Aβ,
remarkably similar.14, 16, 17 Aδ, and C), muscle (groups II, III, and IV),
Preganglionic sympathetic output has a seg- and viscera (groups Aδ and C) can modu-
mental organization, but the segmental distri- late preganglionic neuron activity through seg-
bution of preganglionic fibers does not follow mental pathways.13, 14 Segmental sympathetic
the dermatomal pattern of somatic nerves.13, 16 reflexes are segmentally biased, predominantly
For example, sudomotor functional units in uncrossed, and exhibit ipsilateral, function-
segments T1 and T2 innervate the head and specific, reciprocal, and nonreciprocal patterns
neck, units in T3–T6 innervate the upper of response. For example, nociceptive stim-
extremities and thoracic viscera, those in T7– uli reflexively activate segmental circuitry that
T11 innervate the abdominal viscera, and ones generates excitation of vasoconstrictor outflow
in T12–L2 innervate the lower extremities to skeletal muscle and inhibition of vasocon-
and pelvic and perineal organs.16 Preganglionic strictor outflow to the skin.14 Unlike somatic
sympathetic axons exit through ventral roots reflexes (e.g., the flexor reflex), segmental
sympathetic reflexes do not exhibit reciprocal those in the abdomen, pelvis, and per-
contralateral response patterns, but they may ineum.
exhibit crossed, nonreciprocal responses. For • Postganglionic sympathetic fibers travel
example, during execution of the cold pres- with spinal motor and sensory fibers or
sor test, decreases in cutaneous blood flow with arteries to provide vasomotor, sudo-
in an arm exposed to ice cold water are motor, and pilomotor innervation to the
accompanied by cutaneous vasoconstriction in trunk and limbs.
the contralateral forearm. All segmental spinal • Most sympathetic fibers supplying the
reflexes are subject to supraspinal modula- hand are carried by the median nerve.
tion through several parallel pathways aris- • Most sympathetic fibers supplying the foot
ing in the hypothalamus, pons, and medulla are carried by the peroneal and tibial
and innervating the sympathetic preganglionic nerves.
neurons.9–11 • Unlike somatic motor neurons, sym-
pathetic preganglionic neurons are not
monosynaptically driven by afferent input.
Assessment of Sympathetic • Sympathetic reflexes are segmentally bia-
Function in Humans sed, predominantly uncrossed, and exhibit
ipsilateral, function-specific, reciprocal
Sympathetic function in humans can be and nonreciprocal patterns of response.
• Sympathetic function can be assessed
assessed, directly or indirectly, with various
noninvasive or invasive techniques. Indirect noninvasively by indirect techniques, or
methods include noninvasive tests of sudo- invasively by direct microneurographic
motor and cardiovascular function described recording of postganglionic sympathetic
in Chapters 36–39; measurement of plasma nerve activity.
norepinephrine concentration in forearm veins
with the subject supine and standing;18, 19
assessment of splanchnic20, 21 and cerebral SYMPATHETIC INNERVATION
blood flow22, 23 using Doppler techniques; and
assessment of sympathetic innervation of the OF THE SKIN
heart using radioisotope methods.24, 25 In com-
parison, microneurographic technique allows Sympathetic vasomotor, pilomotor, and sudo-
direct recording of postganglionic sympathetic motor innervation of skin effectors has pri-
nerve activity in humans.16, 17 Nerve record- marily a thermoregulatory function.12, 26, 27 Skin
ings are made with tungsten microelectrodes sympathetic activity is a mixture of sudo-
inserted percutaneously into a nerve, espe- motor and vasoconstrictor impulses and may
cially the median, peroneal, or tibial nerve. sometimes include pilomotor and vasodila-
This technique allows multiunit recordings of tor impulses. The average conduction veloc-
two different types of outflow: skin sympathetic ity for skin sudomotor and vasomotor fibers
nerve activity and muscle sympathetic nerve is 1.3 m/second and 0.8 m/second, respec-
activity.16, 17 tively.16, 17 The intensity of the skin sympathetic
activity is determined mainly by environmen-
tal temperature and the emotional state of the
Key Points
subject. Decreased or increased environmental
• Sympathetic outflow is organized into spe- temperature can produce selective activation
cialized functional units which include of the vasoconstrictor or sudomotor system,
skin vasomotor, muscle, vasomotor, vis- respectively, with suppression of activity in the
ceromotor, pilomotor, and sudomotor other system. Emotional stimuli or inspiratory
units. gasp also increases spontaneous skin sympa-
• Preganglionic sympathetic outflow is orga- thetic activity, but in this case the bursts are
nized segmentally and not by dermatome. caused by simultaneous activation of sudo-
• Preganglionic sympathetic axons exit thr- motor and vasomotor impulses.16, 17 Cutaneous
ough ventral roots to reach the paraver- arteries and veins have a prominent noradren-
tebral sympathetic chain ganglia, which ergic innervation that regulates both nutritive
innervate all tissues and organs except and arteriovenous skin blood flow.12, 26 Nutritive
skin flow is carried by capillaries and is regu- • The sudomotor axon reflex is mediated by
lated by sympathetic (α- and β-adrenergic) and sympathetic sudomotor C fibers.
local nonadrenergic mediators. Arteriovenous • Sympathetic inputs to the sweat glands are
skin blood flow is carried by low-resistance mediated by acetylcholine, acting via M3
arteriovenous shunts, which receive abun- muscarinic receptors.
dant sympathetic vasoconstrictor input and
have a key role in thermoregulation.12, 26 Skin
vasoconstrictor neurons may coordinate their MUSCLE SYMPATHETIC
activity with vasomotor neurons in other vas- ACTIVITY
cular beds to maintain cardiac output, but they
are not sensitive to baroreflex input. Muscle sympathetic activity is composed of
Skin blood flow is also controlled by vasoconstrictor impulses that are strongly mod-
somatosympathetic reflexes13, 14 and three local ulated by arterial baroreceptors.14, 16, 17 Conduc-
axon reflexes: (1) the axon flare response, tion velocity of the postganglionic C fibers
(2) the sudomotor axon reflex, and (3) the has been estimated to be 0.7 m/second in
venoarteriolar reflex. The axon flare response the median nerve and 1.1 m/second in the
is mediated by nociceptive C-fiber terminals.28 peroneal nerve.16 At rest, there is a striking
Activation by noxious chemical or mechanical similarity between muscle sympathetic activ-
stimuli produces antidromic release of neu- ity recorded in different extremities. How-
ropeptides (substance P and others) that cause ever, this activity in the arm and leg can be
skin vasodilatation directly and through stimu- dissociated during mental stress and during
lation of histamine release by mast cells. The forearm ischemia after isometric exercise.29
sudomotor axon reflex is mediated by sympa- Muscle sympathetic activity is important for
thetic sudomotor C fibers. The venoarteriolar buffering acute changes of blood pressure and
reflex is mediated by sympathetic vasomotor decreases in response to moment-to-moment
axons innervating small veins and arterioles. baroreceptor influence.16, 17, 30 It has much less
Skin vasomotor activity has been studied clini- importance, however, for long-term control of
cally by using several noninvasive methods for blood pressure.17 At rest, muscle sympathetic
measuring skin blood flow, including plethys- activity correlates positively with antecubital
mography and laser Doppler flowmetry.28 venous plasma norepinephrine levels.31 Muscle
The eccrine sweat glands in humans have sympathetic activity is also inhibited by car-
a major role in thermoregulation. The seg- diopulmonary receptors. The respiratory cycle,
mental pattern of distribution of sudomotor changes of posture, or the Valsalva maneuver
fibers to the trunk and limbs is irregular and may modulate the muscle sympathetic activity
varies substantially among individuals and even caused by changes in arterial pressure. How-
between the right and the left sides of an ever, hypercapnia, hypoxia, isometric hand-
individual.12 Sympathetic inputs to the sweat grip, emotional stress, or the cold pressor test
glands are mediated by acetylcholine, acting via increase muscle sympathetic activity despite
M3 muscarinic receptors. unchanged or increased arterial pressure.16, 17, 31

• Sympathetic vasomotor, pilomotor, and • Muscle sympathetic activity is modulated
sudomotor innervation of skin effectors by arterial baroreflexes.
has primarily a thermoregulatory function. • Arterial baroreflexes are important for
• The intensity of skin sympathetic activ- buffering short-term changes in blood
ity is determined mainly by environmental pressure but are relatively unimportant in
temperature and the emotional state of regard to long-term blood pressure regu-
the subject. lation.
• Skin blood flow is controlled by somatosym- • Muscle sympathetic activity is modulated
pathetic reflexes and three local axon by the respiratory cycle, changes of pos-
reflexes: (1) the axon flare response, (2) ture or intrathoracic pressure, plasma CO2
the sudomotor axon reflex, and (3) the and O2 tension, muscle exertion, and emo-
venoarteriolar reflex. tional and thermal stress.
AUTONOMIC CONTROL OF • The vagal influence on heart rate is

HEART RATE strongly influenced by respiration (sinus
arrhythmia), is more marked during expi-
Heart rate depends on the effects of parasym- ration, and is absent during inspiration.
pathetic and sympathetic modulation of the • Methods for assessing autonomic control
intrinsic firing rate of the sinus node. Parasym- of heart rate include (1) power spectral
pathetic control is provided by cardiovagal analysis of R–R intervals, (2) heart rate
neurons in the nucleus ambiguus and dor- responses to periodic deep breathing, and
sal motor nucleus in the medulla.12, 32 Sym- (3) heart rate responses to the Valsalva
pathetic noradrenergic control derives mainly maneuver.
from the cervical and upper thoracic gan- • Heart rate variability declines with advanc-
glia.12 In humans at rest, parasympathetic tone ing age.
predominates over the excitatory sympathetic
β-adrenergic influence. Effects mediated by
the vagus nerve have a shorter latency and CARDIOVASCULAR REFLEXES
duration than those mediated by the sym-
pathetic nerves. Heart rate has spontaneous
fluctuations, which reflect changing levels Arterial Baroreflexes
of autonomic activity modulating sinus-node
discharge.33 In normal subjects, control of arterial pressure
Use of power spectral analysis of heart rate depends primarily on the sympathetic inner-
fluctuations allows a noninvasive quantitative vation of the blood vessels, particularly in the
assessment of beat-to-beat modulation of neu- splanchnic bed.36–40 The number of splanch-
ronal activity affecting the heart.34–36 Fluctua- nic sympathetic preganglionic neurons pro-
tions of heart rate at respiratory frequencies gressively decreases with age, and orthostatic
(approximately 0.15 Hz) are mediated almost hypotension occurs after more than 50% of
exclusively by the vagus nerve.33–35 Vagal influ- them have been lost.28 Changes in sympathetic
ences on the heart are associated directly with outflow are regulated by arterial baroreceptors
respiratory activity and are minimal during and chemoreceptors, cardiopulmonary recep-
inspiration and maximal at the end of inspira- tors, and receptors in skeletal muscles (i.e.,
tion and in early expiration. This is the basis of ergoreceptors).36–41 In addition, these reflexes
the respiratory sinus arrhythmia.28, 33, 34 Spon- are modulated by central commands, partic-
taneous and baroreflex-induced firing of cen- ularly from the amygdala, hypothalamus, and
tral cardiovagal neurons is inhibited during periaqueductal gray.9, 10, 42 In humans, arterial
inspiration and is maximal during early expi- pressure is regulated primarily by the carotid
ration.31 Increased tidal volume increases res- sinus and aortic baroreceptors, innervated by
piratory sinus arrhythmia, whereas increased branches of CNs IX and X, respectively, and
respiratory frequency decreases it. Heart rate by cardiopulmonary mechanoreceptors, inner-
variability is correlated inversely with age in vated by vagal and sympathetic afferents. The
normal subjects at rest.28 primary role of arterial baroreflexes is the
Spontaneous lower frequency fluctuations rapid adjustment of arterial pressure around
(those less than 0.15 Hz) of heart rate are the existing mean arterial pressure.36–40 The
mediated by both the vagus and the sympa- baroreceptor reflexes provide a negative feed-
thetic nerves and may be related to baroreflex back that buffers the magnitude of arterial
activity, temperature regulation, and other fac- pressure oscillations throughout the day.
tors. Upright posture in humans dramatically Baroreflexes induce short-term changes in
increases sympathetic nerve activity and pro- heart rate opposite in direction to the changes
duces a large increase in low-frequency heart in arterial pressure, thus increasing heart rate
rate power.33–35 variability. The carotid baroreflex has been
studied by applying negative and positive pres-
sures to the neck, which increase and decrease,
respectively, carotid sinus transmural pres-
Key Points
sure.43 The combined influence of carotid and
• Heart rate depends on parasympathetic aortic baroreceptors in the control of heart
(fast) and sympathetic (slow) influences. rate has been studied by measuring heart
rate responses to changes in arterial pressure function of this reflex is to increase total
induced by intravenous infusion of vasocon- peripheral resistance.28
strictor or vasodilator agents.43 Despite their
major influence on heart rate, the buffer-
ing effects of the carotid baroreflex depend Ergoreflexes
predominantly on changes in total peripheral
resistance.36–40 Static muscle contraction increases heart rate
Baroreflex control of regional circulation is and blood pressure. The mechanisms under-
heterogeneous and largely affects resistance lying these responses involve (1) an exer-
vessels in the splanchnic area.36–41 Sympa- cise pressor reflex initiated by the activation
thetic vasoconstriction in the skeletal muscle is of chemosensitive endings of small myeli-
also strongly modulated by the baroreflex;16, 17 nated and unmyelinated afferent fibers by
this control is dynamic and more suitable local metabolites in the contracting muscle and
for buffering short-term than long-term vari- mediated by group III and IV skeletal mus-
ations of arterial pressure.37, 38 During ortho- cle afferent fibers,45 (2) a pressor reflex initi-
static stress, baroreflex control over skeletal ated by muscle mechanosensitive receptors,46
muscle sympathetic activity declines prior to and (3) a central command that influences
orthostatic syncope.44 During exercise, carotid descending autonomic pathways.42 Cardiovas-
baroreceptor activity is rapidly adjusted to a cular responses to moderately intense static
higher level; this allows increased arterial pres- contraction may be produced primarily by the
sure to meet the metabolic demands of the motor command, which is solely responsible
contracting muscles.36 for increased heart rate. At higher intensity,
responses depend on both the motor command
and the muscle chemoreflexes.36–40
Cardiopulmonary Reflexes
Key Points
Cardiopulmonary receptors are innervated • Control of blood pressure depends pri-
by vagal and sympathetic myelinated and
marily on sympathetically-mediated splan-
unmyelinated afferent fibers. Atrial receptors
chnic vascular tone.
innervated by vagal myelinated fibers are acti- • Cardiovascular sympathetic outflow is reg-
vated by atrial distention or contraction and
ulated by (1) carotid sinus, atrial, aortic,
initiate reflex tachycardia caused by selec-
and cardiopulmonary baroreceptors, (2)
tive increase of sympathetic outflow to the
central command, and (3) skeletal muscle
sinus node. Cardiopulmonary receptors with
ergoreceptors.
unmyelinated vagal afferents, similar to arte- • The primary role of the arterial barore-
rial baroreceptors, tonically inhibit vasomotor
flexes is the rapid adjustment of arterial
activity. Unlike atrial baroreceptors, cardiopul-
pressure around the existing mean arterial
monary receptors provide sustained rather
pressure, predominantly via modulation
than phasic control in sympathetic activity and
of peripheral resistance in the splanchnic
vasomotor tone in the muscle and have no
and skeletal muscle vascular beds.
major effect in controlling heart rate.36–40 • The vagus and glossopharyngeal nerves
innervate carotid and aortic barorecep-
tors, whereas the vagus nerve and sym-
Venoarteriolar Reflexes pathetic myelinated and unmyelinated
afferents innervate cardiopulmonary mec-
The venoarteriolar reflex is a sympathetic post- hanoreceptors.
ganglionic C-fiber axon reflex, with receptors • Carotid sinus and cardiac atrial barore-
in small veins and effectors in muscle arte- ceptors produce phasic bursts, whereas
rioles. Venous pooling activates receptors in cardiopulmonary receptors produce tonic
small veins of the skin, muscle, and adipose changes, to regulate sympathetic vasomo-
tissue; the result is vasoconstriction in the tor activity.
arterioles supplying these tissues. During limb • Venous pooling activates the venoarteri-
dependency, this local reflex vasoconstriction olar reflex, which is mediated by sympa-
may decrease blood flow by 50%. The main thetic postganglionic C fibers, resulting in
vasoconstriction in the arterioles supply- to occur in the skeletal muscle, cutaneous,

ing skin, muscle, and adipose tissue. and splanchnic vascular beds of the dependent
• Static muscle contraction activates ergore- parts. The abdominal compartment (splanch-
flexes which increase heart rate and blood nic circulation) and perhaps skin vasculature
pressure via (1) chemosensitive endings are the most likely sites of venous pool-
of group III and IV skeletal muscle affer- ing. The counterregulatory mechanism that
ent fibers, (2) a pressor reflex mediated by reduces pooling in the lower extremities is acti-
muscle mechanosensitive receptors, and vation of the skeletal muscle pump. Active
(3) a central command. muscle contraction increases intramuscular
pressure, which opposes the hydrostatic forces
and reduces venous pooling in the legs. This
may explain why some patients with ortho-
MAINTENANCE OF POSTURAL static hypotension habitually fidget and con-
NORMOTENSION tract their leg muscles when upright.48 Such
physical countermaneuvers may be taught to
In healthy subjects, orthostatic stress such as patients to help them compensate for ortho-
active standing, head-up tilt, or application of static hypotension.49
lower body negative pressure results in pool-
ing of venous blood in the capacitance vessels
of the legs and abdomen. The bulk of venous Key Points
pooling occurs within the first 10 seconds. In • Orthostatic stress, such as active standing
addition, central blood volume decreases fol- or passive head-up tilting, in the healthy
lowing transcapillary filtration of fluid into the subject results in (1) pooling of venous
interstitial space in the dependent tissues.47 blood in capacitance vessels of the legs
These combined effects lead to a decrease in and abdomen within the first 10 seconds,
venous return to the heart, end-diastolic fill- (2) transcapillary fluid loss into extravas-
ing of the right ventricle, and stroke volume, cular tissue space, (3) decreased venous
resulting in approximately a 20% decrease in return to the heart, and (4) an approxi-
cardiac output. mately 20% decrease in cardiac output.
The adaptation to orthostatic stress con- • The adaptation to orthostatic stress con-
sists of several mechanisms. The decrease in sists of (1) baroreflex-mediated vasocon-
mean arterial pressure is prevented by a com- striction of resistance and capacitance
pensatory vasoconstriction of the resistance vessels of the splanchnic, renal, and
and capacitance vessels of the splanchnic, muscle vascular beds, (2) mechanical
renal, and muscle vascular beds. The initial compression of lower extremity vascu-
adjustments to orthostatic stress are mediated lar beds by contracting skeletal muscle,
by baroreflexes and cardiopulmonary reflexes. and (3) humoral mechanisms includ-
During prolonged orthostatic stress, additional ing the arginine–vasopressin and renin–
adjustments are mediated by humoral mecha- angiotensin–aldosterone systems.
nisms, including the arginine–vasopressin and • The maintenance of postural normoten-
renin–angiotensin–aldosterone systems.38, 47 sion depends on (1) reflex arterial vaso-
Carotid sinus baroreceptors are of primary constriction in the splanchnic, renal, and
importance during standing, whereas car- muscular beds, (2) cardiac output, (3) ade-
diopulmonary receptors have a small role. quate plasma volume, and (4) compres-
When venous filling in the dependent parts sion of capacitance veins by lower extrem-
increases intravenous pressure to 25 mm Hg, it ity skeletal muscles.
activates a local axon reflex, called the venoar-
teriolar reflex, which further contributes to
adaptation to orthostatic stress.
Patients with autonomic failure have dis- SUMMARY
turbed neural reflex arterial vasoconstriction,
and this is the primary mechanism of orthostatic The autonomic nervous system consists of
hypotension. The inability to increase vascular three divisions: the sympathetic (thoracolum-
resistance allows considerable venous pooling bar), parasympathetic (craniosacral), and enteric
nervous systems. The sympathetic and parasym- 5. Chambers, J. B., M. J. Sampson, D. C. Sprigings,
pathetic autonomic outflows involve a two- G. Jackson. 1990. QT prolongation on the electrocar-
neuron pathway with a synapse in an autonomic diogram in diabetic autonomic neuropathy. Diabetic
Medicine 7:105–10.
ganglion. Preganglionic sympathetic neurons 6. Knuttgen, D., D. Weidemann, M. Doehn. 1990. Dia-
are organized into various functional units betic autonomic neuropathy: Abnormal cardiovascular
that control specific targets and include skin reactions under general anesthesia. Klinische Wochen-
vasomotor, muscle vasomotor, visceromotor, schrift 68:1168–72.
7. Burgos, L. G., T. J. Ebert, C. Asiddao, et al. 1989.
pilomotor, and sudomotor units. Microneuro- Increased intraoperative cardiovascular morbidity in
graphic techniques allow recording of postgan- diabetics with autonomic neuropathy. Anesthesiology
glionic sympathetic nerve activity in humans. 70:591–7.
Skin sympathetic activity is a mixture of sudo- 8. Cervero, F., and R. D. Foreman. 1990. Sensory inner-
motor and vasoconstrictor impulses and is vation of the viscera. In Central Regulation of Auto-
nomic Functions, ed. A. D. Loewy, and K. M. Spyer,
regulated mainly by environmental tempera- 104–25. New York: Oxford University Press.
ture and emotional influences. Muscle sym- 9. Loewy, A. D. 1990. Anatomy of the autonomic nervous
pathetic activity is composed of vasoconstric- system: An overview. In Central Regulation of Auto-
tor impulses that are strongly modulated by nomic Functions, ed. A. D. Loewy, and K. M. Spyer,
3–16. New York: Oxford University Press.
arterial baroreceptors. Heart rate is controlled 10. Dampney, R. A. 1994. Functional organization of cen-
by vagal parasympathetic and thoracic sym- tral pathways regulating the cardiovascular system.
pathetic inputs. Vagal influence on the heart Physiological Reviews 74:323–64.
rate is strongly modulated by respiration; it is 11. Jänig, W., and H.-J. Häbler. 1999. Organization of the
more marked during expiration and is absent autonomic nervous system: Structure and function. In
Handbook of clinical neurology, Vol. 74; series 30: The
during inspiration. This is the basis for the autonomic nervous system. Part I. Normal functions,
so-called respiratory sinus arrhythmia, which ed. P. J. Vinken, and G. W. Bruyn, 1–52. Amsterdam:
is an important index of vagal innervation of Elsevier Science Publishers.
the heart. Power spectral analysis of heart 12. Gibbins, I. 1990. Peripheral autonomic nervous sys-
tem. In The human nervous system, ed. G. Paxinos,
rate fluctuations allows noninvasive assess- 93–123. San Diego: Academic Press.
ment of beat-to-beat modulation of neuronal 13. Cabot, J. B. 1996. Some principles of the spinal
activity affecting the heart. Arterial barore- organization of the sympathetic preganglionic outflow.
flex, cardiopulmonary reflexes, venoarteriolar Progress in Brain Research 107:29–42.
reflex, and ergoreflexes control sympathetic 14. Jänig, W. 1996. Spinal cord reflex organization of
sympathetic systems. Progress in Brain Research
and parasympathetic influences on cardiovas- 107:43–77.
cular effectors. The main regulatory mecha- 15. Cheshire W. P. 2000. Peripheral autonomic function
nism that prevents orthostatic hypotension is and dysfunction. In The autonomic nervous system,
reflex arterial vasoconstriction in the splanch- O. Appenzeller, section editor, Handbook of clinical
neurology, ed. P. J. Vinken, and G. W. Bruyn, Vol. 75,
nic, renal, and muscular beds triggered by a 105–42. Amsterdam: Elsevier.
decrease in transmural pressure at the level of 16. Wallin, B. G., and J. Fagius. 1988. Peripheral sym-
carotid sinus baroreceptors. pathetic neural activity in conscious humans. Annual
Review of Physiology 50:565–76.
17. Wallin, B. G., and M. Elam. 1994. Insights from
intraneural recordings of sympathetic nerve traffic in
REFERENCES humans. News in Physiological Sciences 9:203–7.
18. Garland, E. M., S. R. Raj, B. K. Black, et al. 2007. The
1. Sandroni, P., J. E. Ahlskog, R. D. Fealey, P. A. Low. hemodynamic and neurohumoral phenotype of postu-
1991. Autonomic involvement in extrapyramidal and ral tachycardia syndrome. Neurology 69(8):790–8.
cerebellar disorders. Clinical Autonomic Research 19. Thieben, M. J., P. Sandroni, D. M. Sletten, et al.
1:147–55. 2007. Postural orthostatic tachycardia syndrome: The
2. Vinik, A. I., R. A. Maser, B. D. Mitchell, and R. Free- Mayo Clinic experience. Mayo Clinic Proceedings
man. 2003. Diabetic autonomic neuropathy. Diabetes 82(3):308–13.
Care 26:1553–79. 20. Fujimura, J., M. Camilleri, P. A. Low, et al. 1997.
3. Ewing, D. J., I. W. Campbell, B. F. Clarke. 1980. Effect of perturbations and a meal on superior mesen-
Assessment of cardiovascular effects in diabetic auto- teric artery flow in patients with orthostatic hypoten-
nomic neuropathy and prognostic implications. Annals sion. Journal of the Autonomic Nervous System
of Internal Medicine 92:308–11. 67(1–2):15–23.
4. Bellavere, F., M. Ferri, C. Cardone, et al. 1984. Analy- 21. Purewal, T. S., D. E. Goss, M. M. Zanone, et al. 1995.
sis of QT versus RR ECG interval variations in diabetic The splanchnic circulation and postural hypotension
patients shows a longer QT period in subjects with in diabetic autonomic neuropathy. Diabetic Medicine
autonomic neuropathy. Diabetologia 27:255A. 12(6):513–22.
22. Zhang, R., C. G. Crandall, and B. D. Levine. 36. Mancia, G., G. Grassi, A. Ferrari, and A. Zanchetti.
2004. Cerebral hemodynamics during the Valsalva 1985. Reflex cardiovascular regulation in humans.
maneuver: Insights from ganglionic blockade. Stroke Journal of Cardiovascular Pharmacology 7(Suppl
35(4):843–7. 3):S152–9.
23. Horowitz, D. R., and H. Kaufmann. 2001. Autoreg- 37. Abboud, F. M. 1986. Interaction of cardiovascular
ulatory cerebral vasodilation occurs during ortho- reflexes in humans. In Neural mechanisms and car-
static hypotension in patients with primary auto- diovascular disease, ed. B. Lown, A. Malliani, and
nomic failure. Clinical Autonomic Research 11(6): M. Prosdocimi, 73–84. Padova, Italy: Liviana Press.
363–7. 38. Cowley, A. W. Jr. 1992. Long-term control of arterial
24. Yamashina, S., and J. Yamazaki. 2007. Neuronal blood pressure. Physiological Reviews 72:231–300.
imaging using SPECT. European Journal of Nuclear 39. Joyner, M. J., and J. T. Shepherd. 1997. Autonomic
Medicine and Molecular Imaging 34(6):939–50. regulation of circulation. In Clinical autonomic dis-
25. Goldstein, D. S. 2004. Functional neuroimaging of orders: Evaluation and management, ed. P. A. Low,
sympathetic innervation of the heart. Annals of the 2nd ed., 61–71. Philadelphia: Lippincott-Raven Pub-
New York Academy of Sciences 1018:231–43. lishers.
26. Johnson, J. M. 1986. Nonthermoregulatory control of 40. Wieling, W., and K. H. Wesseling. 1993. Importance of
human skin blood flow. Journal of Applied Physiology reflexes in circulatory adjustments to postural change.
61:1613–22. In Cardiovascular reflex control in health and disease,
27. Cheshire, W. P., and P. A. Low. 2007. Disorders of ed. R. Hainsworth, 35–65. London:WB Saunders.
sweating and thermoregulation. Continuum: Lifelong 41. Iellamo, F. 2001. Neural mechanisms of cardiovascular
Learning in Neurology 13(6):143–64. regulation during exercise. Autonomic Neuroscience
28. Low, P. A. 1997. Laboratory evaluation of autonomic 90:66–75.
function. In Clinical autonomic disorders: Evaluation 42. Green, A. L., S. Wang, S. Purvis, et al. 2007. Identify-
and management, ed. P. A. Low, 2nd ed., 179–208. ing cardiorespiratory neurocircuitry involved in central
Philadelphia: Lippincott-Raven Publishers. command during exercise in humans. The Journal of
29. Wallin, B. G., R. G. Victor, and A. L. Mark. 1989. Physiology 578:605–12.
Sympathetic outflow to resting muscles during static 43. Eckberg, D. L., and J. M. Fritsch. 1993. How should
handgrip and postcontraction muscle ischemia. The human baroreflexes be tested? News in Physiological
American Journal of Physiology 256:H105–10. Sciences 8:7–12.
30. Somers, V. K., M. E. Dyken, A. L. Mark, and 44. Ichinose, M., M. Saito, N. Fujii, et al. 2006. Mod-
F. M. Abboud. 1993. Sympathetic-nerve activity dur- ulation of the control of muscle sympathetic nerve
ing sleep in normal subjects. The New England Journal activity during severe orthostatic stress. The Journal of
of Medicine 328:303–7. Physiology 576:947–58.
31. Eckberg, D. L., C. Nerhed, and B. G. Wallin. 1985. 45. Leshnower, B. G., J. T. Potts, M. G. Garry, and
Respiratory modulation of muscle sympathetic and J. H. Mitchell. 2001. Reflex cardiovascular responses
vagal cardiac outflow in man. The Journal of Physiol- evoked by selective activation of skeletal mus-
ogy (London) 365:181–96. cle ergoreceptors. Journal of Applied Physiology
32. Spyer, K. M., P. A. Brooks, and P. N. Izzo. 1994. 90:308–16.
Vagal preganglionic neurons supplying the heart. In 46. Nakamoto, T., and K. Matsukawa. 2007. Muscle
Vagal control of the heart: Experimental basis and clin- mechanosensitive receptors close to the myotendinous
ical implications, ed. M. N. Levy, and P. J. Schwartz, junction of the Achilles tendon elicit a pressor reflex.
45–64. New York: Futura Publishing Company. Journal of Applied Physiology 102:2112–20.
33. van Ravenswaaij-Arts, C. M., L. A. Kollee, 47. Smit, A. A., J. R. Halliwill, P. A. Low, and W. Wieling.
J. C. Hopman, G. B. Stoelinga, and H. P. van 1999. Pathophysiological basis of orthostatic hypoten-
Geijn. 1993. Heart rate variability. Annals of Internal sion in autonomic failure. The Journal of Physiology
Medicine 118:436–47. (London) 519:1–10.
34. Baillard, C., P. Goncalves, L. Mangin, B. Swyn- 48. Cheshire, W. P. 2000. Hypotensive akathisia: Auto-
ghedauw, and P. Mansier. 2001. Use of time frequency nomic failure associated with leg fidgeting while sit-
analysis to follow transitory modulation of the car- ting. Neurology 55:1923–6.
diac autonomic system in clinical studies. Autonomic 49. Bouvette, C. M., B. R. McPhee, T. L. Opfer-Gehrking,
Neuroscience 90:24–8. and P. A. Low. 1996. Role of physical countermaneu-
35. Bernardi, L., C. Porta, A. Gabutti, L. Spicuzza, and vers in the management of orthostatic hypotension:
P. Sleight. 2001. Modulatory effects of respiration. Efficacy and biofeedback augmentation. Mayo Clinic
Autonomic Neuroscience 90:47–56. Proceedings 71:847–53.
Chapter 36
Quantitative Sudomotor Axon

Reflex and Related Tests
Phillip A. Low
INTRODUCTION Abnormal Patterns

LABORATORY EVALUATION OF Significance of the Test
AUTONOMIC FUNCTION Applications
QUANTITATIVE SUDOMOTOR
IMPRINT METHODS OF SWEAT
AXON REFLEX TEST
MEASUREMENT
Normal Response
SUMMARY
INTRODUCTION Because autonomic tests are affected substan-

tially by many confounding variables, stan-
Typically the distal ends of the longest dardization, the recognition of pitfalls,3 and
nerve fibers are first affected in the axonal patient preparation (Table 36–2) are critically
neuropathies.5, 12 Small nerve fibers are often important.
selectively involved and these are termed dis- The indications for autonomic testing are
tal small fiber neuropathy (DSFN).5, 12 One summarized in Table 36–3. The presence
method of recording C fiber function is to of generalized autonomic failure, a major
study the function of the postganglionic sympa- medical problem, adversely affects progno-
thetic sudomotor “C” fiber function.1, 2, 5, 12 We sis. It is important to quantify the deficits
describe below the use of methods to measure by system and by autonomic level. Cer-
sudomotor fiber function. tain benign disorders, such as chronic idio-
pathic anhidrosis4 and benign syncopes, need
to be differentiated from the more serious
dysautonomias. Distal small-fiber neuropathy
LABORATORY EVALUATION OF is diagnosable with autonomic studies, and
AUTONOMIC FUNCTION QSART is abnormal in 80% of cases,2, 5, 6 sug-
gesting that most patients who have distal
The quantitative sudomotor axon reflex test somatic C-fiber involvement also have auto-
(QSART) is a routine test of autonomic func- nomic C-fiber impairment. This relationship
tion and a component of the autonomic reflex is not invariable, and some patients with
screen (Table 36–1). The autonomic reflex distal small-fiber neuropathy and abnormal
screen noninvasively and quantitatively eval- epidermal skin fibers will have a normal
uates sudomotor, adrenergic, and cardiovagal QSART.7 Sympathetically maintained pain is
functions.1, 2 Sudomotor function is discussed associated with unilateral vasomotor and sudo-
in this chapter, and adrenergic and cardiovagal motor abnormalities8, 9 and is described in
functions are considered in Chapter 37. Chapter 40.10
629
Table 36–1 Routine Tests of Autonomic Function

The autonomic reflex screen comprises:
1. QSART distribution (4 sites)
2. Tests of cardiovagal function
a. Heart rate response to deep breathing
b. Valsalva ratio
3. Tests of adrenergic function
a. Blood pressure and heart rate response to head-up tilt
b. Beat-to-beat blood pressure response to Valsalva maneuver and deep breathing
QSART, Quantitative sudomotor axon reflex test.
Table 36–2 Patient Preparation for QSART

No food for 3 hours before testing. The antecedent meal should be a light breakfast or lunch
without coffee or tea.
Treatment with anticholinergic agents should be stopped 48 hours (preferably 5 half-lives of
the drug before the study).
The patient should be comfortable and discomfort-free (e.g., bladder recently emptied).
The room and personnel should be warm and quiet.
Table 36–3 Indications for Autonomic Laboratory Evaluation

Strong indications
1. Suspicion of generalized autonomic failure
2. Diagnosis of benign autonomic disorders that mimic more serious dysautonomia
3. Detection of distal small-fiber neuropathy
4. Diagnosis of autonomic involvement in peripheral neuropathy
5. Pain associated with sympathetic dysfunction
6. Orthostatic intolerance, such as the postural tachycardia syndrome
Desirable indications
1. Monitoring course of autonomic failure
2. Evaluation of response to therapy
3. Peripheral neuropathies
4. Syncope
5. Amyotrophic lateral sclerosis
6. Extrapyramidal and cerebellar degenerations
7. Research questions
The QSART is a routine test of autonomic • To assess patients in whom there is a

function (Table 36–1). The commercial unit suspicion of an autonomic neuropathy.
QSWEAT, sold by WR Medical, Minnesota, • To follow the progress of patients with
is modeled after the Mayo unit. A detailed autonomic neuropathies.
evaluation of autonomic function currently
comprises an evaluation of sudomotor, adren-
ergic, and cardiovagal function. This chapter
focuses on the evaluation of postganglionic QUANTITATIVE SUDOMOTOR
sudomotor function. AXON REFLEX TEST
Purpose and Role of QSART and QSWEAT Many tests of sudomotor function of histor-
ical and limited clinical and research inter-
• To evaluate the integrity of the postgan- est exist. For modern clinical neurophysiology
glionic sudomotor axon. laboratories, only the thermoregulatory sweat
Quantitative Sudomotor Axon Reflex and Related Tests 631
test, QSART, peripheral autonomic surface simultaneously. The commercial unit, although
potential, and sweat imprint test (SIT) require built very similarly, for currently inexplica-
consideration. The thermoregulatory sweat ble reasons generates a response that appears
test is described in Chapter 38. QSART and to be approximately 50% of the volume of
SIT are described in this chapter. the Mayo unit. A study is nearly complete
that will provide a national normative database
for the WR unit (WR Medical, Minneapolis,
Normal Response Minnesota).
The multicompartmental sweat cell
The neural pathway evaluated by QSART (Fig. 36–2) is attached to the skin and per-
consists of an axon reflex mediated by the mits iontophoresis of acetylcholine through
postganglionic sympathetic sudomotor axon the stimulus compartment (C) and a constant-
(Fig. 36–1). The axon terminal is activated by current generator. Axon-reflex-mediated sweat
iontophoresed acetylcholine. Iontophoresis is response is recorded from compartment A.
relatively painless, but is slow and takes min- QSART responses are recorded from standard
utes to occur. Alternatively, the C fiber can sites, which include the distal forearm, prox-
be directly stimulated electrically, but is very imal leg, distal leg, and proximal foot. The
painful.11 The impulse first travels antidromi- tests are sensitive and reproducible in con-
cally to a branch point, and then orthodromi- trol subjects and in patients with neuropathy.
cally to release acetylcholine from the nerve The coefficient of variation was checked in two
terminal. Acetylcholine traverses the neurog- ways. In a group of individuals checked on
landular junction and binds to muscarinic two occasions the value was 14.7%. For three
receptors on eccrine sweat glands to evoke the subjects (with low, moderate, and high sweat
sweat response. volumes recorded on multiple occasions) the
Equipment needed to perform QSART (or mean coefficient of variation was 8%.12 Exten-
QSWEAT) includes the sudorometer, a mul- sive control data are available from QSART
ticompartmental sweat cell, a constant cur- responses in 139 normal subjects (74 females
rent generator, and some method of dis- and 65 males) between 10 and 83 years
playing the sweat response. Mayo builds its old. Mean sweat output was 3.01 μL/cm2 and
own sudorometer consoles; each consists of 1.15 μL/cm2 for the forearm of males and
four sudorometers, thus permitting dynamic females, respectively. This difference was sig-
recording of sweat output from four sites nificant (p < 0.001). Values for the foot were
Figure 36–1. Left, Neural substrate of sudomotor axon reflex test (see text). Right, Representative response from a
36-year-old woman. Arrows indicate onset and cessation of iontophoresis. (From Low, P. A., T. L. Opfer-Gehrking, and
M. Kihara. 1992. In vivo studies on receptor pharmacology of the human eccrine sweat gland. Clinical Autonomic Research
2:29–34. By permission of Lippincott Williams & Willkins.)
C
B
Male 43 yrs
Left forearm
A Output = 3.81 uL
off
on
5 minutes
Figure 36–2. Quantitative sudomotor axon reflex test. Left, Sweat cell and, right, response. The sweat response in com-
partment A is evoked in response to iontophoresis of acetylcholine in compartment C. Compartment B is an air gap.
Sweating causes a change in thermal mass of the nitrogen stream (D) that is sensed by the sudorometer and displayed
(right). (From Low, P. A. 1992. Sudomotor function and dysfunction. In Diseases of the nervous system, ed. A. K. Asbury,
G. M. McKhann, and W. I. McDonald, Vol. 1, 2nd ed., 479–89. Philadelphia: WB Saunders Company. By permission of
Elsevier Science.)
2.65 μL/cm2 and 1.15 μL/cm2 for males and

Left forearm
females, respectively. Again, the difference was 100
statistically significant (p < 0.001). Thus, the
two groups were considered separately. For the 50
forearm, QSART response did not regress with
age. For the foot, however, there was a consis- 0
tent, negative slope of QSART responses with Left proximal leg
increasing age. 100
50
Abnormal Patterns
0
Several abnormal QSART patterns are rec- Left distal leg
100
ognized. The most reliable pattern is a
length-dependent pattern of QSART reduc-
tion (Fig. 36–3). Typically, the foot response is 50
absent first and then sweating over the distal
leg is lost. Sweat reduction is also accepted if 0
the distal site is less than one-third that of the Left foot
100
more proximal site. Sometimes the proximal
site may have an excessive volume. Sometimes 50
a response is associated with an ultrashort
latency (presumably response is due to electri- 0
cal and not chemical stimulation) and fails to 0 2 4 6 8 10 12 14 16 18
turn off when the stimulus ceases; it is often Minutes
seen in painful diabetic and other neuropathies Figure 36–3. Quantitative sudomotor axon reflex test
and in florid reflex sympathetic dystrophy. distribution in a patient with distal small-fiber neuropa-
Another pattern is that of focal postgan- thy. Note reduction of sweat response in the distal leg
glionic denervation, as might occur with a and anhidrosis over the foot. (From Low, P. A., and J. G.
peroneal or ulnar nerve lesion. Especially McLeod. 1997. Autonomic neuropathies. In Clinical auto-
nomic disorders: Evaluation and management, ed. P. A.
with focal abnormalities, the best approach Low, 2nd ed., 463–86. Philadelphia: Lippincott-Raven
is to combine QSART with thermoregulatory Publishers. By permission of Mayo Foundation for Medical
sweat test (see Chapter 38). (Preganglionic Education and Research.)
denervation can be recognized by an intact Applications

QSART response in an area that is anhidrotic
on thermoregulatory sweat test.) QSART has been used to detect postganglionic
sudomotor failure in neuropathies,12, 13 aging,14
Lambert–Eaton myasthenic syndrome,15 amy-
Significance of the Test otrophic lateral sclerosis,16 and preganglionic
neuropathies with presumed transsynaptic
Normal test results indicate integrity of the degeneration.17, 18 In patients with distal small-
postganglionic sympathetic sudomotor axon. fiber neuropathy, it is the most sensitive non-
The absence of a response indicates failure of invasive diagnostic test available.2, 5 It is an
this axon, providing that iontophoresis is suc- important part of the autonomic reflex screen
cessful and eccrine sweat glands are present. when QSART distribution is taken in conjunc-
Because the axonal segment mediating the tion with cardiovascular heart rate tests and
response is likely to be short, the test probably Finapres recordings of the Valsalva maneu-
evaluates distal axonal function. The interpre- ver and tilt.10 Taken together, it helps define
tation of persistent sweat activity, manifest as a the distribution and severity of autonomic
sweat response that is still at least two-thirds of failure. Done under standardized conditions,
maximal response by the end of the recording, the test is highly reproducible and can be
is less certain, but could indicate mild neuropa- utilized to follow the course of sudomotor
thy, especially if associated with an ultrashort deficit. This is useful in monitoring the course
latency. It occurs commonly in mild or painful of a neuropathy and in evaluating response
neuropathies. Because it is unlikely that con- to therapy. Recently, the test has been used
tinuous activity occurs at the level of the sweat to define the threshold and duration of the
gland, the mechanism of persistent sweat activ- local anhidrotic effect of botulinum toxin.19
ity is likely to be repetitive firing of the sym- In studies of sympathetically maintained pain,
pathetic axon. Damaged axons can fire spon- QSART recordings are performed bilaterally
taneously, and on stimulation, they may have and simultaneously for evidence of asymme-
more persistent activity than normal axons. try of latency, volume, and morphology of the
The latency is sometimes markedly reduced responses.8
in painful neuropathies with persistent sweat Another application of QSART is in the
activity. The mechanism of the very short detection of the site of the lesion. The QSART
latency (10–30 seconds) is likely augmented in combination with the thermoregulatory
somatosympathetic reflexes, with a reduced sweat test can be used to determine whether a
threshold of polymodal C nociceptors—it often lesion is preganglionic or postganglionic. A pre-
occurs with allodynia. ganglionic site is deduced when anhidrosis on
the thermoregulatory sweat test is associated
Key Points with normal results on the QSART, and the
lesion is postganglionic when both tests show
• The neural pathway of QSART is the post- anhidrosis.
ganglionic axon. The nerve terminal is For comparisons among patients of differ-
activated by the stimulus, acetylcholine. ent ages and sex, these confounding variables
• The afferent and efferent pathway is the can be managed by at least two approaches.
postganglionic axon (different branches) One, the data can be expressed as normal devi-
and the effector is the eccrine sweat gland. ates.8 Two, an autonomic scoring scale can be
• The test has a coefficient of variation generated that corrects for the effects of age
<20%. Four sites are studied and the and sex.20 On this scale, severity is scored from
pattern of sweat responses is obtained. 0 (no deficit) to 3 (maximal deficit).
• A normal QSART test result indicates
integrity of the postganglionic sympa-
Key Points
thetic sudomotor axon.
• The absence of a QSART response • QSART is useful in demonstrating a
indicates failure of the postganglionic length-dependent pattern of small-fiber
sympathetic sudomotor axon, providing neuropathy, where there is a progressive
that iontophoresis is successful and eccrine reduction in volume from the proximal to
sweat glands are present. distal sites.
• QSART performed with thermoregula- Key Points

tory sweat test determines the site of the
• Sweat imprint methods, such as those in
lesion. If there is anhidrosis on thermoreg-
ulatory sweat test and a normal QSART, which sweat droplets indent a soft plas-
the lesion is preganglionic in site. tic mold, can be used to study the sweat
process.
IMPRINT METHODS OF SWEAT SUMMARY

MEASUREMENT
The application of noninvasive, sensitive, quan-
Several imprint methods have been used, but titative, and dynamic tests of sudomotor func-
only the sweat imprint method is useful clini- tion enhances significantly our ability to quan-
cally. The subject is reviewed elsewhere.21 titate one aspect of the autonomic deficit. The
QSART has an important role in clinical appli-
cations to better definition of the course of
PLASTIC AND SILICONE IMPRINTS neuropathy, its response to treatment, and fur-
A colloidal graphite plastic method was ther exploration of sudomotor physiology.
introduced in 1952.22 The plastic is dis-
solved in a solvent, which dries in approxi-
mately 30 seconds. The sweat droplet forms REFERENCES
an imprint. Alternatives are silicone rub-
ber monomers23 and colloidal graphite plas- 1. Low, P. A. 1993. Autonomic nervous system function.
tic.24 Harris et al.24 compared bromophenol- Journal of Clinical Neurophysiology 10:14–27.
2. Low, V. A., P. Sandroni, R. D. Fealey, and P. A.
blue, colloidal graphite plastic, and silicone Low. 2006. Detection of small-fiber neuropathy by
monomer imprint methods and concluded that sudomotor testing. Muscle & Nerve 34:57–61.
the last was the most sensitive and reliable. 3. Low, P. A. 1993. Pitfalls in autonomic testing. In
Kennedy et al.25 used a combination of Clinical autonomic disorders: Evaluation and manage-
pilocarpine iontophoresis, the Silastic imprint ment, ed. P. A. Low, 355–65. Boston: Little, Brown and
Company.
method, and imaging techniques to systemati- 4. Low, P. A., R. D. Fealey, S. G. Sheps, W. P. Su, J. C.
cally study the sweat response in humans and Trautmann, and N. L. Kuntz. 1985. Chronic idiopathic
rodents. For human studies, a 1% solution of anhidrosis. Annals of Neurology 18:344–8.
pilocarpine is iontophoresed for 5 minutes over 5. Stewart, J. D., P. A. Low, and R. D. Fealey. 1992.
Distal small fiber neuropathy: Results of tests of sweat-
a 1 cm2 area of skin on the dorsum of the ing and autonomic cardiovascular reflexes. Muscle &
hand and foot. The imprint material (0.5 mL Nerve 15:661–5.
base and 4 droplets of accelerator [Syringe 6. Tobin, K., M. J. Giuliani, and D. Lacomis. 1999.
Elasticon, Kerr Co]) is spread thinly over Comparison of different modalities for detection
the skin. The material sets within 3 minutes. of small fiber neuropathy. Clinical Neurophysiology
110:1909–12.
Counts of sweat gland density are made under 7. Periquet, M. I., V. Novak, M. P. Collins, et al. 1999.
a dissecting microscope and counting grid. To Painful sensory neuropathy: Prospective evaluation
estimate volume, the size of the droplet can be using skin biopsy. Neurology 53:1641–7.
determined with computerized imaging anal- 8. Low, P. A., and P. G. McManis. 1985. Autonomic reflex
testing in sympathetic reflex dystrophy. Neurology
ysis techniques. The mean density is 311 and 35(Suppl 1):148A.
281 active glands/cm2 for the hand and foot, 9. Sandroni, P., P. A. Low, T. Ferrer, T. L. Opfer-
respectively, without differences for age or Gehrking, C. L. Willner, and P. R. Wilson. 1998. Com-
sex.26 plex regional pain syndrome I (CRPS I): Prospective
This technique has been applied systemati- study and laboratory evaluation. The Clinical Journal
of Pain 14:282–9.
cally to patients with diabetic neuropathy.25, 26 10. Low, P. A. 1993. Laboratory evaluation of autonomic
Counts of active sweat glands were abnormal failure. In Clinical autonomic disorders: Evaluation
in the hands of 24% of patients with diabetic and management, ed. P. A. Low, 169–95. Boston:
neuropathy and in the feet of 56% of them. Little, Brown and Company.
11. Namer, B., A. Bickel, H. Kramer, F. Birklein, and
Of interest is that approximately one-third of M. Schmelz. 2004. Chemically and electrically
patients with normal electrophysiologic tests induced sweating and flare reaction. Autonomic Neu-
have abnormal results on the SIT.21, 27 roscience 114:72–82.
12. Low, P. A., P. E. Caskey, R. R. Tuck, R. D. Fealey, of botulinum toxin injections: Measured by sudometry.
and P. J. Dyck. 1983. Quantitative sudomotor axon The British Journal of Dermatology 144:111–17.
reflex test in normal and neuropathic subjects. Annals 20. Low, P. A. 1993. Composite autonomic scoring scale
of Neurology 14:573–80. for laboratory quantification of generalized autonomic
13. Low, P. A., B. R. Zimmerman, and P. J. Dyck. failure. Mayo Clinic Proceedings 68:748–52.
1986. Comparison of distal sympathetic with vagal 21. Kennedy, W. R., and X, Navarro. 1993. Evaluation
function in diabetic neuropathy. Muscle & Nerve 9: of sudomotor function by sweat imprint methods. In
592–6. Clinical autonomic disorders: Evaluation and manage-
14. Low, P. A., T. L. Opfer-Gehrking, C. J. Proper, and ment, ed. P. A. Low, 253–61. Boston: Little, Brown and
I. Zimmerman. 1990. The effect of aging on cardiac Company.
autonomic and postganglionic sudomotor function. 22. Sutarman, and Thomson M. L. 1952. A new tech-
Muscle & Nerve 13:152–7. nique for enumerating active sweat glands in man. The
15. McEvoy, K. M., A. J. Windebank, J. R. Daube, and Journal of Physiology 117:51–2.
P. A. Low. 1989. 3,4-Diaminopyridine in the treat- 23. Sarkany, I., and P. Gaylarde. 1968. A method for
ment of Lambert–Eaton myasthenic syndrome. The demonstration of sweat gland activity. The British
New England Journal of Medicine 321:1567–71. Journal of Dermatology 80:601–5.
16. Litchy, W. J., P. A. Low, J. R. Daube, and A. J. Winde- 24. Harris, D. R., B. F. Polk, and I. Willis. 1972. Evaluat-
bank. 1987. Autonomic abnormalities in amyotrophic ing sweat gland activity with imprint techniques. The
lateral sclerosis. Neurology 37(Suppl 1):162. Journal of Investigative Dermatology 58:78–84.
17. Cohen, J., P. Low, R. Fealey, S. Sheps, and N. S. Jiang. 25. Kennedy, W. R., M. Sakuta, D. Sutherland, and F. C.
1987. Somatic and autonomic function in progressive Goetz. 1984. Quantitation of the sweating deficit in
autonomic failure and multiple system atrophy. Annals diabetes mellitus. Annals of Neurology 15:482–8.
of Neurology 22:692–9. 26. Kennedy, W. R., and X. Navarro. 1989. Sympathetic
18. Sandroni, P., J. E. Ahlskog, R. D. Fealey, and P. A. sudomotor function in diabetic neuropathy. Archives
Low. 1991. Autonomic involvement in extrapyramidal of Neurology 46:1182–6.
and cerebellar disorders. Clinical Autonomic Research 27. Navarro, X., and W. R. Kennedy. 1989. Sweat gland
1:147–55. reinnervation by sudomotor regeneration after differ-
19. Braune, C., F. Erbguth, and F. Birklein. 2001. Dose ent types of lesions and graft repairs. Experimental
thresholds and duration of the local anhidrotic effect Neurology 104:229–34.
Chapter 37
Evaluation of Adrenergic Function

Phillip A. Low
INTRODUCTION Denervation Supersensitivity

SKIN VASOMOTOR REFLEXES Plasma Norepinephrine
Beat-to-Beat BP Response to Sustained Handgrip
the VM Baroreflex Indices
Beat-to-Beat BP Response to Tilt-Up
Venoarteriolar Reflex SUMMARY
INTRODUCTION shortcomings. Microneurography is used occa-

sionally to evaluate adrenergic function, but it
Standing up causes a transient decrease in is invasive and is not described here.
pulse and blood pressure (BP). Barorecep-
tors in the carotid sinus and aortic arch are
unloaded. The efferent portion of the barore- Purpose and Role of Adrenergic Function
flex consists of a vagally mediated increase Testing
in heart rate and activation of sympathetic • To test the post-ganglionic sympathetic
efferents, which increases total peripheral nervous system.
resistance. Preganglionic sympathetic fibers • To assess for autonomic disorders that
are cholinergic and postganglionic fibers are involve maintenance of blood pressure,
adrenergic. The preganglionic fiber is short heart rate, and sudomotor function.
and ends in a paravertebral ganglion. In com-
parison, the postganglionic fiber traverses the
entire length of a peripheral nerve. Periph-
eral adrenergic function is important in the SKIN VASOMOTOR REFLEXES
maintenance of postural normotension. It may
be impaired in peripheral neuropathies and The integrity of postganglionic sympathetic
this may be manifested as alterations in acral adrenergic fibers can be evaluated by test-
temperature, color, or sweating. In spite of ing skin vasoconstrictor reflexes.1 Studies are
its importance, simple, accurate, and repro- performed usually on the toe or finger pads,
ducible tests of peripheral adrenergic func- because the sympathetic innervation in these
tion are generally unavailable. This chapter sites is purely vasoconstrictor, whereas other
describes methods used to determine periph- sites, such as the forearm, have both vasocon-
eral adrenergic function and their value and strictor and vasodilator fibers.2
637
Skin blood flow (SBF) is measured with a agents.10 These latter tests are generally insen-
laser Doppler flowmeter or with plethysmog- sitive (catecholamines), poorly reproducible in
raphy, and the vasoconstrictor response to an the clinical laboratory setting, or too invasive.
autonomic maneuver is determined. Vasocon- The best validated tests are the BP and heart
striction can be induced by maneuvers such rate responses to HUT and beat-to-beat BP
as inspiratory gasp, response to standing (for responses to the VM. These are the focus of
the finger), contralateral cold stimulus, or the this chapter.
Valsalva maneuver (VM). The response can be Beat-to-beat BP responses to the VM and
expressed as a percentage of vasoconstriction. tilt are recorded simultaneously with the heart
The pathways of these reflexes are complex. rate. Dynamic alterations during tilt and the
For instance, the response to standing is medi- VM are particularly important in detecting
ated by the venoarteriolar reflex,3 low-pressure adrenergic failure. The VM has four main
and high-pressure baroreceptors,4, 5 and, to a phases: I, II, III, and IV, with phase II subdi-
lesser extent, by increased levels in nore- vided into early (II-E) and late (II-L) phases
pinephrine,6 rennin,7 and vasopressin.8 The (Fig. 37–1 and Table 37–1). The mecha-
advantage of these reflexes is that they have nisms of the VM are summarized briefly in
different afferents but an identical final effer- Table 37–1.
ent pathway. Thus, the relevant afferent path- The phases of the VM can be used to eval-
way can be evaluated. However, this advantage uate adrenergic function. This method of eval-
is offset by a major shortcoming of tests of uating adrenergic function has been validated
skin adrenergic function, that is, the marked in two ways. First, pharmacologic dissection
sensitivity of skin sympathetic fibers to emo- studies have demonstrated that phase II-L is
tional and temperature changes,9 which means primarily under peripheral α-adrenergic con-
there is considerable ambient fluctuation of trol and is selectively blocked by phentolamine,
SBF. Skin vasomotor reflexes have a coeffi- whereas phase IV is completely blocked by
cient of variation greater than 20% and are best propranolol, indicating that it depends on
regarded as semiquantitative tests. However, β-adrenoreceptors.11 Second, the technique
the tests are useful in comparing vasoconstric- has also been validated by studies of con-
tor reflexes from identical sites simultaneously. trols and age-matched and sex-matched patient
groups with graded adrenergic failure.12–14 One
group had generalized autonomic failure, with
Key Points an orthostatic decrease in systolic BP during
• Skin peripheral resistance is determined tilt of 30 mm Hg or greater. A second group
by skin vasomotor reflexes. had less orthostatic decrease in BP (less than
• Vasoconstriction can be induced by mane- 30 mm Hg but greater than 10 mm Hg), and
uvers such as inspiratory gasp, response to a third group had well-documented periph-
standing (for the finger), contralateral cold eral autonomic failure (absence of response
stimulus, or the Valsalva maneuver. on the quantitative sudomotor axon reflex
• The pathways are complex and methods of test) but did not have orthostatic hypoten-
measurement are not very reproducible. sion. In contrast to controls, all the patient
groups, including group 3, exhibited a signif-
icant reduction in phase II-L. An excessive
decrease in BP in phase II and absence of a
Beat-to-Beat BP Response to phase-IV overshoot were observed in the group
the VM with florid orthostatic hypotension. Interme-
diate changes were seen in the group with
The widely used tests to evaluate adrener- borderline orthostatic hypotension. The beat-
gic function in an autonomic laboratory are to-beat BP changes during the VM, when cou-
beat-to-beat BP (and heart rate) responses pled with BP responses to tilt, provide a sig-
to the VM and head-up tilt (HUT). Other nificantly better evaluation of adrenergic fail-
tests include measurement of plasma levels ure than bedside recordings of BP. Patients
of catecholamines, especially norepinephrine with peripheral adrenergic failure, for example
(supine and standing), the sustained hand- those with neuropathy involving autonomic C
grip test, and the response to pharmacologic fibers, have an absence of phase II-L and a
Evaluation of Adrenergic Function 639
Figure 37–1. Beat-to-beat BP responses to the VM in a control subject (A) and in a patient with panautonomic neuropa-
thy (B). In the control subject reflex vasoconstriction, manifested as phases II-L and IV are well developed. In the patient
with panautonomic neuropathy, pulse pressure is much reduced during the maneuver, phases II-L and IV are absent, and
PRT is delayed.
Table 37–1 Major Mechanisms of Phases of the VM

Phase Main Mechanism(s)
I & III Compression (I) and release from compression (III) of thoracic aorta
II-E ↓ in venous return; partial compensation by vagus nerve
II-L ↑ in peripheral arteriolar tone
IV ↑ in cardiac sympathetic tone; sustained ↑ in arteriolar tone
E, early; L, late; ↑, increase; ↓, decrease.
slight increase in phase II-E. Patients with (PRT)13, 14 and absent phase IV (Fig. 37–1).
more severe and widespread autonomic fail- The gradations of alterations of the VM in dif-
ure, for example widespread involvement of ferent types and degrees of autonomic failure
limb and splanchnic adrenergic fibers, have a have been summarized on the basis of the
large phase II-E, an absence of phase II-L, experience in the Mayo Autonomic Laboratory
and typically have delayed BP recovery time (Table 37–2). The simplest approach is to
Table 37–2 Changes in Phases of VM with Types of Autonomic Failure

Condition II-E II-L IV Valsalva ratio
Normal 5–15 mm To baseline >baseline Normal

Vagal lesion <Normal <Normal Normal <Normal
Adrenergic failure
Peripheral <Normal ↓ or absent; ↑ PRT <Normal <Normal
Generalized ↑ Fall Absent; ↑ PRT Absent Reduced
↑, increase; ↓, decrease; PRT, BP recovery time.
measure PRT. A more complete evaluation sphygmomanometer cuff with the patient
related PRT to the antecedent fall in BP gen- supine and following tilt to 70◦ . It is impor-
erating adrenergic sensitivity.14 A normative tant to perform the upright tilt procedure at
database has been generated for these adren- a standard time after the patient lies down,
ergic indices.15 because the orthostatic reduction in BP is
greater after 20 minutes of preceding rest than
Key Points after 1 minute; we routinely perform HUT
after a minimum of 30 minutes of recumbency.
• Tests to evaluate adrenergic function in an
Beat-to-beat recordings of systolic, mean, and
autonomic laboratory are beat-to-beat BP diastolic BP are displayed continuously on the
(and heart rate) responses to the VM and computer console, as is heart rate, which is
HUT. derived from the electrocardiographic leads
• In the clinical laboratory, an accurate and
and monitor. Also, it is important to ensure that
reproducible way to evaluate total periph- the arm is at heart level, because arm position
eral resistance is to study beat-to-beat BP influences the measurement of BP.
responses to the VM. During upright tilt, normal subjects have
• The VM has four main phases: I, II, III,
a transient decrease in systolic, mean, and
and IV, with phase II subdivided into early diastolic BP, followed by recovery within 1
(II-E) and late (II-L) phases. minute.16 The decrement is modest (less than
• The maneuver results in a transient fall
10 mm Hg, mean BP). Patients with adren-
in BP. ergic failure have a marked and progres-
• The baroreflex response results in reflex
sive decrease in BP and pulse pressure. The
vasoconstriction due to an increase in total heart rate response typically is attenuated, but
peripheral resistance (late phase II). in patients whose cardiac adrenergic inner-
• An excessive decrease in BP in phase II
vation is spared, the response is intact and
and absence of a phase-IV overshoot are may be increased. Indices of mild adrener-
observed in patients with florid orthostatic gic impairment include excessive oscillations
hypotension. of BP, an excessive decrease (more than 50%)
• The beat-to-beat BP changes during the
in pulse pressure, a transient (first minute)
VM, when coupled with BP responses to decrease in systolic BP greater than 30 mm
tilt, provide a significantly better evalu- Hg, an excessive increment in heart rate (≥30
ation of adrenergic failure than bedside beats/minute), and a failure of total systemic
recordings of BP. resistance to increase. Premonitory signs of
• PRT increases with age. An increased PRT
syncope are a progressive decrease in BP
indicates an impairment of the adrenergic (especially diastolic BP), total peripheral resis-
component of the baroreflex. tance, pulse pressure, and loss of BP (and
heart rate) variability. Some of these indices are
expected abnormalities in a failure of arterio-
Beat-to-Beat BP Response to lar vasoconstriction (total peripheral resistance
Tilt-Up and diastolic BP). Some are signs of increased
vascular capacitance (reduction in pulse pres-
Orthostatic BP recordings to tilt are recorded sure and excessive increment in heart rate).
using beat-to-beat BP and verified with a Increased oscillations are indicative of intact
compensatory mechanisms (but are abnormal patients whose cardiac adrenergic inner-
because they indicate a system under stress), vation is spared, the response is intact and
whereas the gradual loss of variability indicates may be increased.
the failure of compensation. • Indices of mild adrenergic impairment
A large autonomic database is available. The include excessive oscillations of BP, an
normal response varies by age and sex. To facil- excessive decrease (more than 50%) in
itate comparison, a 10-point composite auto- pulse pressure, a transient (first minute)
nomic severity score (CASS) of autonomic decrease in systolic BP greater than
function has been developed.12 The scheme 30 mm Hg, an excessive increment in
allots 4 points for adrenergic and 3 points heart rate (≥30 beats/minute), and a fail-
each for sudomotor and cardiovagal failure. ure of total systemic resistance to increase.
Each score is normalized for the confound-
ing effects of age and sex. Patients with a
CASS score of 3 or less have only mild auto-
nomic failure, those with scores of 7–10 have Venoarteriolar Reflex
severe failure, and those with scores between
these two ranges have moderate autonomic When venous transmural pressure is increased
failure. The sensitivity and specificity of the by 25 mm Hg (e.g., by lowering the limb
method were assessed by evaluating CASS in by 40 cm), reflex arteriolar vasoconstriction
four groups of patients with known degrees of occurs, reducing blood flow by 50%.3 This
autonomic failure: 18 patients with multisys- reflex, called the venoarteriolar reflex, has
tem atrophy, 20 with autonomic neuropathy, 20 receptors in small veins, and its neural pathway
with Parkinson’s disease, and 20 with periph- appears to be that of the sympathetic C fiber
eral neuropathy but no autonomic symptoms. local axon reflex (Fig. 37–2).17 The function of
The composite scores (mean ± SD) for these the reflex has been suggested to be to increase
four groups were 8.5 ± 1.3, 8.6 ± 1.2, 1.5 ± total peripheral resistance, compensating by up
1.1, and 1.7 ± 1.3, respectively. Patients with to 45% of the orthostatic decrease in cardiac
symptomatic autonomic failure had scores of 5
or more, and those without symptomatic auto-
nomic failure had scores of 4 or less; no overlap
existed among these groups.
Key Points
• Beat-to-beat BP recordings in response to
HUT are a reliable method of studying
baroreflex function.
• The normal response is affected by age
and gender.
• It is important to perform the upright tilt
procedure at a standard time after the
patient lies down, because the orthostatic
reduction in BP is greater after 20 minutes
of preceding rest than after 1 minute.
• During upright tilt, normal subjects have
a transient decrease in systolic, mean, and
diastolic BP, followed by recovery within 1 Figure 37–2. Neural pathway for the venoarteriolar
minute. reflex. The postganglionic axon comprises both the afferent
• Patients with adrenergic failure have a and efferent limbs of the reflex. (From Moy, S., T. L. Opfer-
Gehrking, C. J. Proper, and P. A. Low. 1989. The venoarte-
marked and progressive decrease in BP riolar reflex in diabetic and other neuropathies. Neurology
and pulse pressure. The heart rate 39:1490–2. By permission of the American Academy of
response typically is attenuated, but in Neurology.)
output.17, 18 It may also reduce the orthostatic supersensitivity is an increase in receptor den-
increase in tissue fluid by adjusting the sity, affinity, efficacy of receptor–effector cou-
precapillary-to-postcapillary resistance ratio. pling, or other postreceptor events. These tests
The venoarteriolar reflex was measured in of adrenergic denervation supersensitivity are
the feet of patients with diabetes mellitus and too invasive for routine use and are insensitive.
was reported to be reduced in those with dia- It would be preferable to measure acral
betic neuropathy.19 The value of this test is adrenergically mediated vasoconstriction in
its theoretical ability to examine the status of response to the above-mentioned infusion.
sympathetic vasoconstrictor fibers at the post- However, recordings of vasoconstriction of the
ganglionic level. muscle bed are indirect. Plethysmography can
Moy et al.20 studied the venoarteriolar reflex, be performed, but recorded flow is contami-
quantitative sudomotor axon reflex test, and nated by SBF.
heart rate responses to deep breathing and the
VM in 40 control subjects, 49 diabetic sub-
jects, and 29 subjects with other neuropathies. Key Points
The mean vasoconstrictor response was greater • Denervation supersensitivity is an exag-
in the control group than the diabetic group gerated BP rise in response to a drug that
or other neuropathy group, but the overlap acts on postganglionic adrenergic recep-
among the groups was marked. The venoarte- tors.
riolar reflex appeared to have lower specificity • It is due to an increase in density of these
and much lower sensitivity than other tests of receptors that occurs secondary to post-
autonomic function. The sensitivity and speci- ganglionic denervation.
ficity were considered inadequate for it to be
used as a clinical test of autonomic function.
Plasma Norepinephrine
Key Points
Plasma norepinephrine results from a spillover
• The venoarteriolar reflex is an axon reflex of norepinephrine from adrenergic postgan-
that has receptors in small veins, and its glionic nerve terminals. The supine value is
neural pathway appears to be that of the an index of net sympathetic activity24–26 and is
sympathetic C fiber local axon reflex. affected by the rate of norepinephrine secre-
• The pathway is the afferent and efferent tion and clearance.27 Plasma norepinephrine
components of a postganglionic axon. has been used to differentiate postganglionic
• The normal response is an increase in total from preganglionic failure. In a disorder in
peripheral resistance. which the lesion is preganglionic, resting
• The function of the reflex has been sug- supine norepinephrine values are normal, but
gested to be to increase total peripheral the response to standing is absent because of
resistance. failure of activation. In a postganglionic lesion,
• The venoarteriolar reflex appeared to have the supine norepinephrine plasma values are
lower specificity and much lower sensitiv- decreased if the lesion is widespread. The
ity than other tests of autonomic function. disadvantage of the method is its low sensi-
tivity, largely because 80% of released nore-
pinephrine is taken up again by the neuron and
only 20% enters venous blood. One method
Denervation Supersensitivity of enhancing clinical utility is to measure the
intraneuronal metabolite dihydroxyphenylgly-
An exaggerated pressor response to the intra- col (DHPG).
arterial or intravenous application of directly
acting α-agonists (such as phenylephrine or Key Points
norepinephrine), indicating denervation super-
sensitivity, may occur when there is widespread • Plasma norepinephrine measures the neu-
denervation of postganglionic sympathetic rotransmitter that binds to postganglionic
fibers.21–23 The mechanism of denervation adrenergic receptors and increases BP.
• Plasma norepinephrine has been used to to changes in BP. Baroreflex gain to an increase
differentiate postganglionic from pregan- or decrease in BP primarily evaluates the vagal
glionic failure. responses to changes in BP. Recently, there has
• Its main limitation is that it is lacking in been considerable interest in baroreflex gain in
sensitivity. response to spontaneous or induced changes
in BP. Commonly used maneuvers include
spectral analysis, the VM, and the sequence
Sustained Handgrip method.32
All these methods evaluate the vagal com-
Sustained muscle contraction causes an increase ponent of the baroreflex and an efficient
in systolic and diastolic BP and heart rate. The approach is to relate the fall in BP during phase
stimulus derives from exercising muscle, the II-E to the resultant heart period. However,
reaction is reflexive in nature, and the increase the adrenergic component can be selectively
in BP is mediated by an increase in cardiac out- involved, as in certain neuropathies. Recently,
put and peripheral resistance.28 The test has we have described methodology to quantitate
been adapted as a simple test of sympathetic the adrenergic component of baroreflex sen-
autonomic failure.29 Ewing et al.29 recommend sitivity.13, 14 The adrenergic component of the
30% maximal contraction for up to 5 min- baroreflex is reflected in phase II-L and its con-
utes. Many patients have difficulty sustaining tinuation, the PRT,13 which provides a repro-
the test for 5 minutes, but 3 minutes is suf- ducible index of adrenergic baroreflex sensitiv-
ficient. The test evaluates generalized rather ity. A more complete index can be derived by
than peripheral adrenergic function. relating PRT to the preceding fall in BP.14
Key Points
Key Points
• Sustained handgrip results in a rise in BP
• Baroreflex indices evaluate the heart
due to an increase in cardiac output and
period (reciprocal of heart rate) responses
peripheral resistance.
• It is lacking in sensitivity and is difficult to to induced increases and decreases in BP.
• The baroreflex maintains a stable BP in
do and standardize.
spite of changes in position.
• There is a vagal component that changes
heart rate and an adrenergic component
Baroreflex Indices that changes total peripheral resistance.
• The vagal component is evaluated by
Baroreflex indices evaluate the heart period quantitating the change in heart period in
(reciprocal of heart rate) responses to induced response to a fall in BP.
increases and decreases in BP. Phenylephrine • The adrenergic component can be eval-
or norepinephrine can be used to increase uated by measuring PRT or, more com-
BP, and tilt or trinitroglycerin can be used pletely, by dividing the antecedent fall in
to decrease it. One approach, adapted from BP by PRT.
the method of Korner, is to determine the
range and mean gain of the heart period.21
These two indices express the range of heart
period in response to a moderate pressor- SUMMARY
hypotensive stress, and the mean rate of
change in heart period in response to sud- For noninvasive evaluation of autonomic func-
den changes in BP.21 Another related approach, tion, tests of peripheral adrenergic function
described by Pickering et al.,30 relates beat- have recently been developed so that it is
to-beat systolic BP to heart rate. An alter- possible to separately evaluate the vagal and
native approach to stimulating baroreceptors adrenergic components of baroreflex sensitiv-
is to use a neck chamber whose pressure ity. The vagal component is derived from the
can be increased or decreased.31 Finally, the heart period response to BP change and the
heart period responses to the decrease and adrenergic component by the PRT in response
increase in BP during the VM can be related to the preceding fall in BP, induced by the VM.
REFERENCES 17. Henriksen, O. 1977. Local sympathetic reflex mech-

anism in regulation of blood flow in human subcuta-
1. Low, P. A., C. Neumann, P. J. Dyck, R. D. Fealey, neous adipose tissue. Acta Physiologica Scandinavica
and R. R. Tuck. 1983. Evaluation of skin vasomo- 450(Suppl):1–48.
tor reflexes by using laser Doppler velocimetry. Mayo 18. Henriksen, O., K. Skagen, O. Haxholdt, and V. Dyr-
Clinic Proceedings 58:583–92. berg. 1983. Contribution of local blood flow regulation
2. Roddie, I. C., J. T. Shepherd, and R. F. Whelan. 1957. mechanisms to the maintenance of arterial pressure
A comparison of the heat elimination from the normal in upright position during epidural blockade. Acta
and nerve-blocked finger during body heating. The Physiologica Scandinavica 118:271–80.
Journal of Physiology (London) 138:445–8. 19. Rayman, G., S. A. Williams, P. D. Spencer, L. H.
3. Henriksen, O. 1991. Sympathetic reflex control of Smaje, P. H. Wise, and J. E. Tooke. 1986. Impaired
blood flow in human peripheral tissues. Acta Physio- microvascular hyperaemic response to minor skin
logica Scandinavica 603(Suppl):33–9. trauma in type I diabetes. British Medical Journal
4. Sundlof, G., and B. G. Wallin. 1978. Human mus- (Clinical Research Edition) 292:1295–8.
cle nerve sympathetic activity at rest. Relationship to 20. Moy, S., T. L. Opfer-Gehrking, C. J. Proper, and P. A.
blood pressure and age. The Journal of Physiology Low. 1989. The venoarteriolar reflex in diabetic and
274:621–37. other neuropathies. Neurology 39:1490–2.
5. Zoller, R. P., A. L. Mark, F. M. Abboud, P. G. Schmid, 21. Low, P. A., J. C. Walsh, C. Y. Huang, and J. G. McLeod.
and D. D. Heistad. 1972. The role of low pres- 1975. The sympathetic nervous system in diabetic
sure baroreceptors in reflex vasoconstrictor responses neuropathy. A clinical and pathological study. Brain
in man. The Journal of Clinical Investigation 51: 98:341–56.
2967–72. 22. Moorhouse, J. A., S. A. Carter, and J. Doupe. 1966.
6. Vendsalu, A. 1960. Studies on adrenaline and nora- Vascular responses in diabetic peripheral neuropathy.
drenaline in human plasma. Acta Physiologica Scan- British Medical Journal 1:883–8.
dinavica 49(Suppl 173):1–123. 23. Smith, A. A., and J. Dancis. 1964. Exaggerated
7. Oparil, S., C. Vassaux, C. A. Sanders, and E. Haber. response to infused norepinephrine in familial dysau-
1970. Role of renin in acute postural homeostasis. tonomia. The New England Journal of Medicine
Circulation 41:89–95. 270:704–7.
8. Williams, T. D., D. Da Costa, C. J. Mathias, R. Ban- 24. Polinsky, R. J., I. J. Kopin, M. H. Ebert, and V. Weise.
nister, and S. L. Lightman. 1986. Pressor effect of 1981. Pharmacologic distinction of different ortho-
arginine vasopressin in progressive autonomic failure. static hypotension syndromes. Neurology 31:1–7.
Clinical Science 71:173–8. 25. Wallin, B. G., G. Sundlof, B. M. Eriksson,
9. Burton, A. C., and R. M. Taylor. 1940. Study of adjust- P. Dominiak, H. Grobecker, and L. E. Lindblad.
ment of peripheral vascular tone to requirements of 1981. Plasma noradrenaline correlates to sympathetic
regulation of body temperature. The American Journal muscle nerve activity in normotensive man. Acta
of Physiology 129:565–77. Physiologica Scandinavica 111:69–73.
10. Low, P. A. 1997. Laboratory evaluation of autonomic 26. Ziegler, M. Q., C. R. Lake, and I. J. Kopin. 1977.
function. In Clinical autonomic disorders: Evaluation The sympathetic nervous system defect in primary
and management, ed. P. A. Low, 179–208. Philadel- orthostatic hypotension. The New England Journal of
phia: Lippincott-Raven. Medicine 296:293–7.
11. Sandroni, P., E. E. Benarroch, and P. A. Low. 1991. 27. Polinsky, R. J., D. S. Goldstein, R. T. Brown, H. R.
Pharmacological dissection of components of the Val- Keiser, and I. J. Kopin. 1985. Decreased sympathetic
salva maneuver in adrenergic failure. Journal of neuronal uptake in idiopathic orthostatic hypotension.
Applied Physiology 71:1563–7. Annals of Neurology 18:48–53.
12. Low, P. A. 1993. Composite autonomic scoring scale 28. Coote, J. H., S. M. Hilton, and J. F. Perez-Gonzalez.
for laboratory quantification of generalized autonomic 1971. The reflex nature of the pressor response to mus-
failure. Mayo Clinic Proceedings 68:748–52. cular exercise. The Journal of Physiology 215:789–804.
13. Vogel, E. R., P. Sandroni, and P. A. Low. 2005. 29. Ewing, D. J., J. B. Irving, F. Kerr, J. A. W. Wildsmith,
Blood pressure recovery from Valsalva maneuver and B. F. Clarke. 1974. Cardiovascular responses to
in patients with autonomic failure. Neurology 65: sustained handgrip in normal subjects and in patients
1533–7. with diabetes mellitus: A test of autonomic function.
14. Schrezenmaier, C., W. Singer, N. M. Swift, D. Sletten, Clinical Science and Molecular Medicine 46:295–306.
J. Tanabe, and P. A. Low. 2007. Adrenergic and vagal 30. Pickering, T. G., B. Gribbin, E. S. Petersen, D. J. Cun-
baroreflex sensitivity in autonomic failure. Archives of ningham, and P. Sleight. 1972. Effects of autonomic
Neurology 64:381–6. blockade on the baroreflex in man at rest and during
15. Huang, C. C., P. Sandroni, D. M. Sletten, S. D. exercise. Circulation Research 30:177–85.
Weigand, and P. A. Low. 2007. Effect of age on 31. Eckberg, D. L., M. S. Cavanaugh, A. L. Mark, and
adrenergic and vagal baroreflex sensitivity in normal F. M. Abboud. 1975. A simplified neck suction device
subjects. Muscle & Nerve 36(5):637–42. for activation of carotid baroreceptors. The Journal of
16. Low, P. A., J. C. Denq, T. L. Opfer-Gehrking, P. J. Laboratory and Clinical Medicine 85:167–73.
Dyck, P. C. O’Brien, and J. M. Slezak. 1997. Effect of 32. Dawson, S. L., T. G. Robinson, J. H. Youde, et al.
age and gender on sudomotor and cardiovagal function 1997. The reproducibility of cardiac baroreceptor
and blood pressure response to tilt in normal subjects. activity assessed non-invasively by spectral sequence
Muscle & Nerve 20:1561–8. techniques. Clinical Autonomic Research 7:279–84.
Chapter 38
Thermoregulatory Sweat Test

Robert D. Fealey
INTRODUCTION METHOD
ROLE OF THERMOREGULATORY THERMOREGULATORY SWEAT
SWEAT TESTING: CLINICAL DISTRIBUTION
SYNDROMES AND PROBLEMS REPORTING RESULTS
EVALUATED DIFFICULTIES AND PITFALLS IN
Small Fiber Neuropathy INTERPRETATION
Diabetic Peripheral Neuropathies SUMMARY
Severity of Autonomic Failure APPENDIX: SWEAT TEST
Central Disorders PROCEDURE
Hyperhidrotic Disorders
INTRODUCTION the M3 muscarinic cholinergic synapse, and

the sweat glands. Because the entire anterior
From a physiologic viewpoint, the thermoreg- body surface is tested for both preganglionic
ulatory sweat test (TST) consists of giving and postganglionic lesions, the TST is well
a controlled heat stimulus to the body in suited for screening patients with certain clin-
a tolerable fashion to produce a generalized ical symptoms or for demonstrating autonomic
sweat response (i.e., recruiting all areas of involvement in many disorders.1
skin capable of sweating). The TST assesses
the integrity of efferent (central and periph- Purpose and Role of TST
eral) sympathetic sudomotor pathways. Central
(preganglionic) structures tested include the • To assess the integrity of efferent sympa-
hypothalamus, bulbospinal projections, inter- thetic sudomotor pathways.
mediolateral cell column in the thoracic and • To screen patients for small fiber periph-
upper lumbar spinal cord, and white rami. eral neuropathy.
Peripheral (postganglionic) autonomic struc- • To assess for autonomic involvement in
tures tested include the sympathetic chain patients with many neurological and med-
ganglia and postganglionic sudomotor axons, ical disorders.
645
ROLE OF THERMOREGULATORY Small Fiber Neuropathy

SWEAT TESTING: CLINICAL
SYNDROMES AND PROBLEMS Peripheral neuropathies affecting smaller diam-
eter nerve fibers often cause the “burning feet”
EVALUATED syndrome. Nerve conduction studies and elec-
tromyography and clinical neurologic exam are
A comprehensive listing of clinical syndromes often normal in these patients, whereas tests of
that can be evaluated with TST is given in skin autonomic innervation are usually abnor-
Table 38–1.2 Some common neurologic appli- mal. The TST has been investigated in this
cations of the TST evaluation are described in syndrome and found to show impaired sweat-
the following sections. ing of the distal limbs in 72%–74% of patients
Table 38–1 Diseases Evaluated with TST
I. Generalized Autonomic Failure Syndromes

Pure autonomic failure (PAF)
Progressive autonomic failure as part of multiple system atrophy (MSA)
II. Primary disorders with isolated acquired idiopathic anhidrosis
Progressive isolated segmental anhidrosis
Idiopathic pure sudomotor failure
Chronic idiopathic anhidrosis
Ross syndrome
III. Primary autonomic neuropathies
Panautonomic (acute pandysautonomia) neuropathy
Autoimmune autonomic neuropathy (or ganglionopathy)
Postural tachycardia syndrome (POTS) (some cases)
IV. Anhidrosis associated with other neurologic disorders
Central nervous system lesions (stroke, tumor, infection, infiltration, trauma)
Hypothalamic lesions (heat stroke)
Brain stem lesions (pontine and lateral medullary stroke)
Spinal cord lesions (Traumatic spinal cord injury [SCI], multiple sclerosis)
Cold-induced sweating syndrome
Degenerative disorders
Diffuse lewy-body disease, Parkinson’s disease—autonomic failure
Peripheral nerve lesions causing anhidrosis
Hereditary sensory and autonomic neuropathy types I, II, IV
Guillain–Barré syndrome (acute inflammatory demyelinating polyneuropathy [AIDP])
Diabetic autonomic neuropathy
Hereditary and primary systemic amyloidosis
Lepromatous neuropathy
Myasthenic syndrome
Alcoholic neuropathy
Fabry’s disease
Idiopathic small fiber neuropathy
Erythromelalgia
Sympathectomy and other surgical lesions
Harlequin syndrome
Postural tachycardia syndrome
Anhidrosis due to toxins and pharmacologic agents
Botulism, botulinum toxin injections
Ganglionic blockers, anticholinergics, carbonic anhydrase inhibitors
Opioids
(Continued)
III. Anhidrosis associated with disorders of skin and sweat glands

Anhidrosis due to physical agents damaging skin
Trauma, burns, pressure, scar formation, radiation therapy
Anhidrosis due to congenital and acquired skin diseases
Fabry and other congenital metabolic diseases
Anhidrotic ectodermal dysplasia
Ichthyosis
Neutrophilic eccrine hidradenitis
Sjogren’s syndrome
Systemic sclerosis (scleroderma)
Incontinentia pigmenti
Dermatomal vitiligo
Bazex–Dupre–Christol syndrome
Disorders affecting the sweat duct
Miliaria
Palmoplantar pustulosis
Psoriasis
Lichen planus
Atopic dermatitis
IV. Primary (essential) focal hyperhidrosis
Palmoplantar, axillary, craniofacial, generalized hyperhidrosis
V. Secondary causes of localized hyperhidrosis
Due to cerebral infarction
Frontal opercular infarct
Brain stem stroke
Associated with SCI
Autonomic dysreflexia
Posttraumatic Syringomyelia
Orthostatic hypotension triggered
Associated with other central nervous system disorders
Arnold–Chiari type 1 malformation
Myelopathies due to infarction, syringomyelia, tumor
Cold-induced sweating syndrome
Olfactory hyperhidrosis
Associated with peripheral nervous system disorders
Peripheral motor neuropathy with autonomic dysfunction
Dermatomal or focal hyperhidrosis due to nerve trunk irritation
Compensatory segmental hyperhidrosis (postsympathectomy, Ross syndrome, pure autonomic failure)
Gustatory sweating
Physiologic
Idiopathic
Postherpetic
Post nerve injury (post surgical, diabetic autonomic neuropathy, post infectious, tumor invasion)
Lacrimal sweating
Harlequin syndrome
Idiopathic, localized hyperhidrosis
Idiopathic unilateral circumscribed hyperhidrosis
Postmenopausal localized hyperhidrosis
Associated with local skin disorders
Blue rubber bleb nevi
Glomus tumor
Burning feet syndrome
Granulosis rubra nasi
(Continued)
647
Pretibial myxedema
POEMS syndrome
VI. Secondary causes of generalized hyperhidrosis
Associated with central nervous system disorders
Episodic hypothermia with hyperhidrosis (Hines–Bannick or Shapiro’s syndrome)
Posttraumatic or posthemorrhagic “diencephalic epilepsy”
Fatal familial insomnia and Parkinson’s disease
Associated with fever and chronic infection
Tuberculosis, malaria, brucellosis, endocarditis
Associated with metabolic and systemic medical diseases
Hyperthyroidism, diabetes mellitus, hypoglycemia, hypercortisolism, acromegaly
Associated with malignancy
Leukemia, lymphoma, pheochromocytoma, Castleman’s disease, carcinoids, renal cell cancer
Medication induced
Neuroleptic malignant syndrome
Serotonin syndrome, other medications
Toxic syndromes
Alcohol, opioid withdrawal, delirium tremens
Associated with central and peripheral nervous system disorders
Familial dysautonomia (Riley–Day), Morvan’s fibrillary chorea
SCI, spinal cord injury.
clinically diagnosed.3, 4 The value of the TST in the section on Abnormal TST Distributions are
characterizing the distribution of neuropathy from patients with diabetic neuropathy.
has been emphasized in a published algo- Peripheral neuropathy first produces dis-
rithm on the evaluation of peripheral neu- tal sweat loss in the lower extremities and as
ropathy.5 Patients can be serially evaluated the neuropathy advances the fingertips and
via TST to document progression or recovery the lower anterior abdomen become affected.
(Fig. 38–1, for color image, see color plates). Painful truncal radiculopathy has a distinct
We have recently shown that neurogenic forms clinical presentation of agonizing, and at times,
of erythromelalgia have characteristic TST lancinating pain associated with cutaneous
abnormalities in most instances.6 dyesthesia and a characteristic TST pattern
of patchy, asymmetric anhidrosis primarily in
Key Points the anterior distribution of one or several
• TST identifies abnormality of small fiber adjacent thoracic dermatomes in the distri-
function when nerve conduction studies bution of the pain. Patterns relating to sym-
and electromyography are normal. pathetic chain or ganglionic involvement pro-
• TST can be used to follow small fiber duce an “autosympathectomy.” Uncommonly,
nerve involvement over time. severe diabetic autonomic neuropathy pro-
duces global anhidrosis. The percentage of
body surface anhidrosis (TST%) correlates
Diabetic Peripheral Neuropathies highly with the degree of autonomic neuropa-
thy symptoms and signs.9 Other neuropathies
Diabetes mellitus produces distinct periph- (primary systemic amyloid,10 subacute auto-
eral neurologic disorders including length- nomic,11, 12 paraneoplastic, and leprosy2 ) with
dependent axonal neuropathy, painful truncal widespread, multifocal involvement are best
radiculopathy, and segmental and diffuse auto- evaluated via the TST.
nomic neuropathy.7, 8 The ability to examine
the whole anterior body surface in detail and Key Points
the common involvement of sudomotor axons
in diabetes makes the TST particularly well • Proximal, distal, truncal, radiculoplexus,
suited to the evaluation of this disorder. Most and autonomic ganglionic types of involve-
of the abnormal TST distributions described in ment can be evaluated with TST.
Thermoregulatory Sweat Test 649
3/21/01 4/12/01 11/5/01 1/2/03 1/13/06
70 64 41 22 6
Figure 38–1. Recovering small fiber neuropathy. Serial TSTs in a patient with small fiber neuropathy. Extensive anhidrosis
(yellow) correlated with impaired temperature perception, the latter is the only abnormality on neurological exam. Patient
underwent intravenous immunoglobulin (IVIG) therapy between 3/21/01 and 11/5/01 and began to improve. Subsequent
TSTs document near complete recovery of sudomotor function in a length-dependent fashion. The percentage of anhidrosis
of the anterior body surface (TST%) appears below each figure providing a quantitative measure of the improvement
(sweating in purple-shaded regions).
Severity of Autonomic Failure • Restricted sudomotor failure syndromes

like chronic idiopathic anhidrosis and Ross
The TST is helpful in identifying autonomic syndrome are ideally evaluated via TST.
involvement in many neurologic disorders.
An important use involves the evaluation of
patients presenting with an extrapyramidal Central Disorders
syndrome. The differential diagnosis often
involves separating multiple system atrophy Central disorders such as traumatic and inflam-
(MSA) with its severe autonomic failure, matory myelopathy23, 44 cause preganglionic
widespread anhidrosis, and poor prognosis TST abnormalities often showing the level
from Parkinson’s disease with mild or no auto- of spinal cord involvement in great detail
nomic involvement on TST13–15 and better (Fig. 38–2, for color image, see color plates).
prognosis and response to treatment. Often The TST is often coupled with a test of
the patient with early MSA will exhibit a postganglionic function to confirm the central
preganglionic, segmental sweat deficit com- involvement.
patible with a central lesion affecting the Key Points
intermediolateral cell columns of the spinal
cord.1, 16 Syndromes with selective sympathetic • Traumatic and inflammatory myelopathies,
sudomotor failure include chronic idiopathic hypothalamic neuroendocrine disorders,
anhidrosis1, 17–19 and Ross syndrome,20–22 which and disorders of thermoregulation (heat
are other clinical disorders where TST can stroke, Shapiro syndrome may show
provide diagnostic information. abnormalities on TST).
Hyperhidrotic Disorders
Key Points
Focal hyperhidrosis syndromes are routinely
• Both Parkinson’s disease with autonomic evaluated via TST. Many times the area of
failure and MSA have highly characteristic excessive sweating is unilateral and occurs in
and prevalent abnormalities on TST. compensation to widespread loss of sweating
• TST is routinely used to distinguish MSA elsewhere (e.g., in pure autonomic failure
from Parkinson’s disease. [PAF] or Ross syndrome). When excessive
Figure 38–2. Traumatic myelopathy. A patient with segmental anhidrosis (yellow) with compensatory left-sided hemihy-
perhidrosis (purple) due to a right greater than left-sided upper thoracic spinal cord injury (alizarin red indicator powder).
(From Fealey, R. D., and K. Sato. 2008. Disorders of the eccrine sweat glands and sweating. In Fitzpatrick’s Dermatology
in general medicine, ed. K. Wolff, L. A. Goldsmith, S. I. Katz, B. A. Gilchrest, A. S. Paller, and D. J. Leffell, Vol. 1, 7th ed.,
720–30. New York City, NY: McGraw-Hill Companies, Figure 82–2.).
sweating is confined to the palmar and plan- Therefore, proper technique includes control-
tar areas and there is normal thermoregulatory ling the ambient air temperature and humid-
sweating elsewhere, the TST provides confir- ity as well as the patient’s core and skin
mation of a diagnosis of primary focal (essen- temperature.
tial) hyperhidrosis. More recently, we have Several techniques, including hot baths and
used TST to delineate the effects of sympatho- infrared (IR) and incandescent heat lamps,
tomy, a less invasive, endoscopic technique to have been used for the last 50 years to produce
treat hyperhidrosis of the hands24 and to eval- sweating, but the most satisfactory method is
uate any patients having compensatory hyper- to use a cabinet in which the environment is
hidrosis after such procedures. controlled and the entire body (including the
head) is heated. Guttmann28 described such a
Key Points cabinet and demonstrated the usefulness of the
TST in the diagnosis and monitoring of spinal
• Primary focal hyperhidrosis, postsympa- cord and peripheral nerve lesions. The TST
thectomy, and perilesional compensatory conducted in the Mayo Clinic Thermoregula-
hyperhidrosis are best evaluated via TST. tory Laboratory is a modification of Guttmann’s
quinizarin sweat test1 and is described below.
The patient, clad only in a towel(s) to main-
METHOD tain modesty, is placed in a supine position
on a gurney and enclosed in the cabinet
Thermoregulatory sweating is influenced by (Fig. 38–3A).
the mean and local skin temperature as well The “head end” of the cabinet is a clear vinyl
as by the central (blood/core) temperature.25 curtain (tented over the gurney); this arrange-
A maximal sweat response occurs when both ment allows the head to be in the heated envi-
central (oral) and mean skin temperatures are ronment yet provides the patient a clear view of
increased in a moderately humid (about 35% the technician and outside world to minimize
relative humidity) environment in which some claustrophobia. Windowed access doors allow
degree of sweat evaporation can occur.1, 25–27 access and patient visibility. The rear end of the
Figure 38–3. Top two images: Mayo Thermoregulatory Sweat Cabinet (outside front view on left, rear view on right).
Bottom two images: Inside view of Mayo Thermoregulatory Sweat Cabinet (ceiling shown on right). IR, infrared heaters;
DC, digital camera; Sk/OrPr, skin and oral temperature probes; FL, fluorescent light; Spk, speakers.
cabinet contains the electrical and plumbing of 45◦ C–50◦ C and a relative humidity of
components that regulate the environmental 35%–40% while maintaining skin tempera-
temperature, humidity, and slow airflow rate. ture between 39◦ C and 40◦ C. Components
Suitable parameters for achieving a gen- providing accurate measurement and control
eralized, maximal sweat response within 60 of skin temperature are contained in the
minutes include an ambient temperature control box.
Recently, a study using a similar magnitude skin and turns dark purple on sweating skin.
of air and skin temperature and humidity and Other indicators currently used include iod-
exposure duration has clearly shown that this inated cornstarch30 and starch and iodine in
thermal stress produces maximal thermoregu- solution.31
latory sweating.24, 29 Achieving maximal sweat- The average response of the oral tempera-
ing is of great importance in evaluating sympa- ture in 35 healthy control subjects (20–75 years
thetic sudomotor function and in test reliability old) who achieved full-body, maximal sweat-
and reproducibility. ing during the TST was an increase of 1.2◦ C
The major inside components of the TST in 35–40 minutes (Fig. 38–4). Because all sub-
cabinet are shown below in Fig. 38–3B. Within jects had sweat profusely at an oral tempera-
the insulated walls are three overhead IR ture of less than or equal to 38◦ C, we use this
heaters that heat the skin and are carefully reg- temperature as a test end point.
ulated by skin temperature feedback control. During 2006, the mean oral temperature
The wide gurney, music played over rear- increase in 1412 patients was 1.5◦ C, with 38◦ C
mounted speakers (Spk), and bright fluores- or a 1◦ C increase above baseline (whichever
cent lighting provide comfort for the patients yielded the higher temperature) as an end
as they rest supine with palms down. Ceiling- point. These observations indicate that the
mounted, remotely operated digital cameras often-quoted 1◦ C temperature increase cri-
(DC) photograph the developing sweat distri- terion for an adequate TST32 is inadequate
bution during the test and provide real-time in patients with lower (i.e., <37◦ C) initial
patient monitoring. temperatures.
Thermistor probes (skin and oral tempera- If generalized sweating occurs at a lower
ture probes [Sk/OrPr]) continuously measure body temperature, the test can be ended
skin and oral temperatures during the test. before 38◦ C is reached.
Sweating on the skin surface is best visual- For reasons of patient comfort and safety,
ized by an indicator powder that is applied we do not increase oral temperature above
to the body before heating. A mixture of 38.5◦ C or extend the heating period beyond
alizarin red, cornstarch, and sodium carbon- 65 minutes. The current procedure followed
ate in a 50:100:50 gram ratio, respectively, is by the TST lab technician is specified in the
suitable.23 It is light orange on nonsweating appendix.
39.0
38.0
°C
37.0
36.0
0 5 10 15 20 25 30
Heating time (min)
Skin temp Sweat rate (arbitrary units)

Oral (core) temp
Figure 38–4. Mean skin and oral temperatures and sweat rates of 35 healthy controls during the TST. By 38◦ C oral
temperature, all had reached a maximum sweat rate. (From Fealey, R. D. 1997. Thermoregulatory sweat test. In Clin-
ical autonomic disorders: Evaluation and management, ed. P. A. Low, 2nd ed., 245–57. Philadelphia: Lippincott-Raven,
Figure 2.)
Key Points peptides and proteins making up the structure

• An enclosure (TST cabinet) with a regu- and innervation of the sweat gland.
QSART measures postganglionic axon reflex
lated temperature and humidity is most
sweating. The SSR measures hand and foot
suitable. The entire body is placed in the spinal and postganglionic sympathetic activ-
heated cabinet. The cabinet must accom-
ity. Silastic mold, iodine-impregnated paper
modate a standard hospital gurney and imprint, and Macroduct measure direct sweat
have visibility for patient and technician gland activation and postganglionic innervation
communication.
• Computer-controlled heating of the skin via pilocarpine iontophoresis. Punch biopsies
can examine sweat gland and intraepidermal
surface via overhead IR heaters maximizes nerve fiber densities. These techniques are
sweating and minimizes test time.
• Maintain an air temperature of 45◦ C– sometimes done at multiple sites to provide
information on the distribution of sweating
50◦ C, a relative humidity of 35%–40%, abnormalities.
and skin temperature between 39◦ C and It is desirable to combine several methods
40◦ C. Monitor core (oral or tympanic for determining the integrity of the eccrine
membrane) and skin temperature contin- sweat response. For example, a TST can be
uously. combined with tests of the sweat gland and/or
• Test end point is whole body sweating or,
its peripheral nerve innervation to localize a
if not fully sweating, raising core temper- sweating disorder to the peripheral or central
ature to 38◦ C or a 1◦ C increase above nervous system, or a volumetric technique can
baseline (whichever yields the higher tem- be combined with a sweat droplet distribution
perature). imprint to estimate the sweat volume per active
• An indicator powder providing high color
gland. The TST done first can provide direction
contrast between sweating and nonsweat- to the optimum testing site for more focused
ing skin applied over most of the anterior techniques.
body surface beforehand gives the best
result.
• Overhead photographs of the body sur- Key Points
face at test end document the result and
provide a digital image that can be ana- • Other techniques exist that measure sweat-
lyzed for percent anhidrosis (TST%). ing or sweat gland innervation at one or
more focal areas of the skin. These include
QSART, SSR, silastic mold and iodine-
impregnated paper sweat imprint, and
OTHER METHODS OF Macroduct sweat collection after pilo-
MEASURING SWEAT carpine stimulation, punch skin biopsy
DISTRIBUTION with PGP (protein gene product) 9.5 axo-
nal staining, and membrane microdialysis.
Sweat gland activity can be studied quantita-
tively by a number of other techniques includ-
ing the sympathetic skin response (SSR),33, 34
filter paper collection, weighing and analyzing THERMOREGULATORY SWEAT
of sweat composition, the quantitative sudomo- DISTRIBUTION
tor axon reflex test (QSART),35 silastic mold
or iodine-impregnated paper imprint follow- Normal distributions. The normal variants in
ing pilocarpine stimulation,36 microcannulation sweat distribution observed in the Mayo TST
of the sweat duct or coil,37 collection into clinic laboratory in healthy control subjects
a Macroduct® coil,38 and humidity sensors from 20 to 75 is shown in Figure 38–5, for
within ventilated capsules. Other more sophis- color image, see color plates.7 Areas of “nor-
ticated techniques utilize microdialysis mem- mal” anhidrosis may occur over bony promi-
branes delivering minute quantities of trans- nences (e.g., the patella and clavicle) and in
mitter substances to the dermis39 or employ the inner thighs. The proximal extremities fre-
confocal electronmicroscopy and immunohis- quently show less sweating than the distal,
tochemical analysis of biopsied skin stained for but left–right symmetry is always the rule.
Figure 38–5. Normal thermoregulatory sweat distributions. Sweating areas in purple shading. From Low, P. A., and
R. D. Fealey. Sudomotor neuropathy.9
Males tend to show the type 1 (heavy, general- characterize the segmental anhidrosis as
ized sweating) pattern, whereas females usually central or peripheral (preganglionic vs.
show the type 2 (heavy, generalized sweating postganglionic).
but less in proximal extremities) or type 3 (gen- 3. Regional anhidrosis refers to large anhi-
eralized sweating but less in proximal extremi- drotic areas (but <80%) that blend grad-
ties and lower abdomen) pattern. Elderly men ually into sweating areas and that may or
and women tend to have types 2 and 3, and may not be contiguous; anhidrosis of the
the lighter sweating areas may have a higher trunk alone and anhidrosis of the proximal
threshold of activation.24 parts of all four extremities are examples
Abnormal distributions. Six types of abnor- of this pattern (Fig. 38–6C).
mal thermoregulatory sweat patterns or distri- 4. Mixed patterns are combinations of any of
butions have been described.7 An example of the above in the same patient, for exam-
each pattern is shown in Figure 38–6, for color ple, focal and distal patterns of anhidrosis
image, see color plates. (Fig. 38–6B, 38–6D, and 38–6E).
5. Focal sweat loss is confined to isolated
1. Distal anhidrosis is characterized by dermatomes, peripheral nerve territo-
sweat loss in the fingers, the distal legs, ries, or small localized areas of the skin
feet and toes, and the lower anterior (Fig. 38–6E). This pattern is usually diag-
abdomen (Fig. 38–6A). When this pattern nostic of a peripheral nerve branch or a
is noted, a peripheral autonomic neuropa- focal skin disorder.
thy is highly likely. 6. Global anhidrosis, by definition, occurs
2. Segmental anhidrosis involves large con- when more than 80% of the body surface
tiguous zones of the body surface bor- is involved (Fig. 38–6F).
dered by areas of normal sweating; these
usually respect sympathetic dermatomal Key Points
borders or spinal cord levels(Fig. 38–6B
and 38–6D). Testing the TST area of • There are six basic abnormal TST distri-
anhidrosis with a peripheral sudomotor butions—distal, segmental, focal (radicu-
test such as QSART is often needed to lar), regional, global, and mixed.
Figure 38–6. Abnormal TST Distributions compared with Type 1 normal. Examples of the most commonly encoun-
tered abnormal sweat distribution patterns: Distal (1), Segmental (2 and 4), Regional (3), Global (6), Focal (5) and Mixed
(2, 4 and 5); normal image (7). Sweating areas in purple shading. (From Low, P. A., and R. D. Fealey. Sudomotor
neuropathy.8 )
• Several abnormal TST patterns provide and peripheral (preganglionic vs. postgan-
near diagnostic information regarding glionic) lesions.
the anatomical location of neuropath- • Normal thermoregulatory sweat patterns
ology. can include focal loss over bony promi-
• The use of a peripheral sudomotor test nences and striae and show reduced
such as QSART with the TST allows sweating over the medial thighs and lateral
the TST to distinguish between central calves.
REPORTING RESULTS reviewed by the reporting physician to ensure

accuracy and give an impression of what the
Data about the age, sex, identity number, clin- results are and then their clinical significance
ical problem of the patient, and the date of is made.
the TST are indicated on the report (Fig. 38–7, A customized computer program is used
for color image, see color plates). to count the number of anhidrotic pixels in
The body of the report includes a results the body image. This amount is divided by
section showing the initial and final core the total number of pixels constituting the
temperature, the percentage of body surface anterior body surface drawing and multiplied
anhidrosis, the type of sweat pattern, and an by 100 to produce the percent anhidrosis.
anatomical figure showing the distribution of Another accurate method for deriving the
the sweating or anhidrosis. This anatomical TST% is to use planimetry (LASICO, Model
figure is generated by modifying a computer 1252 M; resolution = 0.005 cm2 ). The drawn
graphics body image to look like the digital body image is scrutinized and the region(s)
camera images taken of the patient at the end of anhidrosis are outlined and measured; indi-
of the test. This figure is used to calculate the vidual areas are summated to produce a total
percentage of anterior body surface anhidro- anterior body surface estimate. The TST% pro-
sis (TST%)9, 23 as described below. The figure vides for quantitation of the test result and can
also provides a permanent record of the sweat be useful in clinical studies9, 14 and in following
distribution when printed on a color ink-jet the clinical course of patients with autonomic
printer. The figure and camera images are disorders40 (Fig. 38–1).
Figure 38–7. TST Report form: Example from a patient with Ross’s Syndrome.
Key Points Whether this represents a variant of chronic

idiopathic anhidrosis18 or a loss of functional
• The TST report should contain patient
sweat glands or reflects the loss of pregan-
demographics, testing conditions, initial
glionic autonomic neurons known to occur
and final core temperature, the sweat dis- with aging42 is unclear. Aging is associated with
tribution pattern (ideally with the per-
declining postganglionic sympathetic sudomo-
centage of anhidrosis estimated), and a
tor responses in the distal leg and foot as well.43
physician’s interpretation of the results.
• A color graphic of the sweat distribution in In our controls aged 20–75 years (19 men and
16 women), there was no significant regres-
the report is desirable.
• Construct an electronic copy (for the med- sion of percent anhidrosis with age, although
the mean age was a relatively young 52 years.9
ical record) and paper copy (for labora-
Senile atrophy of the skin and differences in
tory) of the report.
sweat gland training between young and old
may also be a factor. Our current view is to
interpret the anhidrosis of the elderly as abnor-
DIFFICULTIES AND PITFALLS IN mal and deserving of additional workup to find
INTERPRETATION an etiology.
Potential problems with the TST include
The interpreter must be aware of the normal the untidiness and duration of the test, the
patterns of thermoregulatory sweating, includ- possibility of skin heat injury, or skin irrita-
ing areas where anhidrosis may be seen nor- tion caused by the indicator powder alizarin.
mally, such as over bony prominences or the Contact dermatitis occurs rarely (observed fre-
lateral calves (Fig. 38–5). quency, 3:1000 subjects) and is readily treated
Patients who have worn pressure wraps (i.e., with oral and topical agents. A more common
ace bandages and abdominal binders) within but not harmful problem is the persistence of
12 hours before the test may show anhidrosis in purple discoloration of small skin areas; it may
the areas that had been covered. This is usually take several days for the color to wash off.
recognized by the straight edges of the deficit. Because of repeated exposure to the indica-
Severely dehydrated patients may sweat tor powder, technicians in our laboratory wear
less overall,41 but they generally do not have masks, gloves, and goggles when applying the
focal defects. Anticholinergic drugs, including powder to minimize inhalation, oral ingestion,
most tricyclic antidepressants, may inhibit ther- or contact with the eyes. Patients are also given
moregulatory sweating and should be stopped goggles and masks while the powder is applied
48 hours before the TST is performed. Mu under a ventilated hood that traps airborne
receptor opioid agonists (narcotic analgesics) indicator dust protecting personnel, patients,
can elevate the set point temperature for sweat and equipment.
onset and so may produce a false-positive result Technicians and patients are made aware of
if the usual endpoint temperature is used to possible skin reactions before application and
end the test. Retesting 24–48 hours off short- extremely sensitive individuals are not tested.
acting opioids is required for an accurate test. Our legal department is aware of the non-FDA
The application of skin lotions such as mois- approved use; we accept the use of alizarin as
turizing creams may produce a discoloration a reasonable medical risk where no good sub-
of alizarin-covered skin, making it difficult to stitute (except perhaps iodinated starch) exists
discern areas of anhidrosis as well as normal and the incidence of skin reactions is very low.
sweating. Consequently, the use of lotions is Patients who are extremely claustrophobic,
prohibited on the day of the TST. who indicate a history of severe contact der-
Anhidrosis in elderly patients may present an matitis, or who are less than 12 years old are
interpretative challenge, because the effect of usually not tested.
aging on the autonomic nervous system may be It is anticipated that TST can be used for
responsible for the regional anhidrosis (most controlled trials to monitor progression of dis-
often affecting the lower abdomen and prox- ease or response to treatment as long as one
imal extremities) that is seen in some older adheres rigorously to thermoregulatory stimu-
women who have normal neurologic examina- lus guidelines and has a way to quantitate the
tions but report having dry skin for many years. results (such as the percentage of body surface
anhidrosis). As such the TST should produce 2. Meet the patient and briefly describe the
reproducible data and be complementary to test; weigh the patient and then have
anatomically focused techniques of autonomic them undress.
testing. 3. Place towel(s) on the patient to cover
as little skin as feasible to maintain
Key Points modesty. Place respirator mask on the
• Inadequate heating and low endpoint patient. Remove after powdering. Put
on your respirator mask, goggles, and
temperature are the commonest errors.
• Also common is testing with patient taking gloves.
4. Start the ventilator hood and apply the
anticholinergic drugs; 48 hours off such
indicator powder (alizarin red, corn-
medication is recommended.
• Do not do test if patient is febrile or has starch, and sodium carbonate).
5. Place elastic straps to hold the ante-
taken opioid analgesics within 4 hours of
rior abdomen, proximal leg, and fore-
the test.
• Skin lotions and compression stockings head thermistor probes. Take the sterile
oral probe and its sponge holder and
can cause artifacts and should not be worn
place it in the patient’s mouth between
the day of test.
the gum and the cheek. Turn down the
overhead IR heater control setting from
48◦ C (step 1) to 36◦ C.
SUMMARY 6. As quickly as possible, open the vinyl
curtain and place the patient and gurney
The TST assesses the integrity of central in the cabinet; close the curtain. Open
and peripheral efferent sympathetic sudomo- side doors and place the skin tempera-
tor neural pathways. A controlled heat and ture probes under the elastic straps.
humidity stimulus is given to produce a gen- 7. Connect probes to thermometer input
eralized sweating response in all skin areas plugs. Record baseline oral and skin
capable of sweating. Sweating is visualized by temperatures. Check stability of oral
placing an indicator powder on the skin before- temperature during breathing and talk-
hand. The entire anterior body surface can ing; allow 3 minutes for stabilization;
be examined and abnormalities can usually be begin timing the sweat test.
detected at a glance. 8. Check the cabinet temperature and
Clinical disorders effectively evaluated by humidity every 5 minutes; operating
the TST, the characteristic normal and abnor- conditions have to be as follows to
mal sweat distributions, the methods to reliably ensure an adequate and safe heat
perform the TST, preparation of a report of the stimulus:
test results, and the techniques to quantify the
response including the “percentage of anhidro- Air temperature 44◦ C–48◦ C
sis” are described herein. Important parame- Relative humidity 35%–40%
ters of the heat stimulus, the patient’s oral and Skin temperature 38.5◦ C–39.5◦ C
skin temperature response, and pitfalls in the 9. Try to make all observations through the
interpretation of the sweat test results are also closed side doors; open the curtain if
described. necessary to wipe the patient’s forehead.
10. If the patient sweats completely (fully
saturating the powder on all skin areas)
APPENDIX: SWEAT TEST before the oral temperature reaches
PROCEDURE 38.0◦ C, advise the physician and end
the test. Otherwise, continue the test
1. Preheat and humidify the sweat cabinet until the oral temperature is 38.0◦ C or
by turning it on 45–60 minutes before 1.0◦ C above the initial oral temperature
the test. The set points are 48◦ C air tem- (whichever is greater). Do not exceed
perature and 37%–39% relative humid- oral temperature of 38.5◦ C or total heat-
ity. The overhead IR heaters can also be ing time of 65 minutes. Using the over-
set to 48◦ C to accelerate the preparation. head digital cameras, photograph the
developing sweat distribution, obtaining 5. Dyck, P. J., I. A. Grant, and R. D. Fealey. 1996. Ten
additional photos for abnormalities. steps in characterizing and diagnosing patients with
11. At the end of the test, turn off heat, peripheral neuropathy. Neurology 47(1):10–17.
6. Davis, M. D., J. Genebriera, P. Sandroni, and
remove the probes, and take the patient R. D. Fealey. 2006. Thermoregulatory sweat testing
out of the cabinet. Obtain additional in patients with erythromelalgia. Archives of Derma-
digital photos of abnormal areas, chart tology 142(12):1583–8.
the pattern of sweating on the report 7. Dyck, J. B., and P. J. Dyck. 1999. Diabetic polyneu-
ropathy. In Diabetic neuropathy, ed. P. J. Dyck,
form, and call the physician to inspect and P. K. Thomas, 2nd ed., 255–78. Philadelphia:
the sweat distribution and prepare the W. B. Saunders Company.
report. Help the patient to shower. Keep 8. Low, P. A., and R. D. Fealey. 1999. Sudomotor neu-
the laboratory neat by making sure the ropathy. In Diabetic neuropathy, ed. P. J. Dyck,
shower curtain is inside the stall, by vac- and P. K. Thomas, 2nd ed., 191–9. Philadelphia:
W. B. Saunders Company.
uuming any loose powder and briefly 9. Fealey, R. D., P. A. Low, and J. E. Thomas. 1989.
mopping floor. Wash the skin probes Thermoregulatory sweating abnormalities in diabetes
with soap and warm water; wash off the mellitus. Mayo Clinic Proceedings 64(6):617–28.
oral probe and put it in its pouch for 10. Wang, A. K., R. D. Fealey, T. L. Gehrking, and
P. A. Low. 1999. Autonomic failure in amyloidosis.
sterilization. Neurology 52(Suppl. 2):A388.
12. Put fresh linen on gurney for next test. 11. Suarez, G. A., R. D. Fealey, M. Camilleri, and
13. Check to see that the patient has P. A. Low. 1994. Idiopathic autonomic neuropathy:
showered off most of the powder and Clinical, neurophysiologic, and follow-up studies on 27
help him or her to dress if necessary. patients. Neurology 44(9):1675–82.
12. Vernino, S. J. Adamski, T. J. Kryzer, R. D. Fealey,
Weigh the patient again. Offer elec- and V. A. Lennon. 1998. Neuronal nicotinic ACh
trolyte replacement drink. Place all used receptor antibody in subacute autonomic neuropa-
towels and linen in the laundry bag. thy and cancer-related syndromes. Neurology 50(6):
14. Enter the test data into the laboratory 1806–13.
13. Cohen, J., P. Low, R. Fealey, S. Sheps, and N. S. Jiang.
computer; draw the final body sweat- 1987. Somatic and autonomic function in progressive
distribution image and calculate or mea- autonomic failure and multiple system atrophy. Annals
sure the percentage of anhidrosis. Pre- of Neurology 22(6):692–9.
pare the electronic report for patient’s 14. Sandroni, P., J. E. Ahlskog, R. D. Fealey, and P. A. Low.
medical record and print a paper copy 1991. Autonomic involvement in extrapyramidal and
cerebellar disorders. Clinical Autonomic Research
for the laboratory report file. 1(2):147–55.
15. When performing a second sweat test on 15. Fealey, R. D. 2001. Use of the thermoregulatory
the same patient, ensure the same end- sweat test in the evaluation of patients with autoim-
point temperature and increase in oral mune autonomic neuropathy and early multiple sys-
tem atrophy syndromes. Japanese Journal of Perspira-
temperature are achieved to allow com- tion Research 8:37–40.
parison of resulting sweat distributions. 16. Kihara, M., J. Sugenoya, and A. Takahashi. 1991.
The assessment of sudomotor dysfunction in mul-
tiple system atrophy. Clinical Autonomic Research
1(4):297–302.
17. Faden, A. I., P. Chan, and E. Mendoza. 1982. Pro-
REFERENCES gressive isolated segmental anhidrosis. Archives of
Neurology 39(3):172–5.
1. Fealey, R. D. 1997. Thermoregulatory sweat test. In 18. Low, P. A., R. D. Fealey, S. G. Sheps, W. P. Su,
Clinical autonomic disorders, ed. P. A. Low, 2nd ed., J. C. Trautmann, and N. L. Kuntz. 1985. Chronic
245–57. Philadelphia: Lippincott-Raven. idiopathic anhidrosis. Annals of Neurology 18(3):
2. Fealey, R. D., and K. Sato. 2008. Disorders of the 344–8.
eccrine sweat glands and sweating. In Fitzpatrick’s 19. Ando, Y., S. Fujii, N. Sakashita, M. Uchino, and
dermatology in general medicine, ed. L. A. Goldsmith, M. Ando. 1995. Acquired idiopathic generalized
K. Wolff, S. I. Katz, B. Gilchrest, A. S. Paller, and anhidrosis: clinical manifestations and histochemical
D. J. Leffell, 7th ed., 789–800. New York City, NY: studies. Journal of Neurological Sciences 132(1):80–3.
McGraw-Hill Companies. 20. Chemmanam, T., J. D. Pandian, R. S. Kadyan, and
3. Low, V. A., P. Sandroni, R. D. Fealey, P. A. Low. 2006. S. M. Bhatti. 2007. Anhidrosis: A clue to an underlying
Detection of small-fiber neuropathy by sudomotor autonomic disorder. Journal of Clinical Neuroscience
testing. Muscle & Nerve 34(1):57–61. 14(1):94–6.
4. Stewart, J. D., P. A. Low, and R. D. Fealey. 1992. 21. Nolano, M., V. Provitera, and A. Perretti, et al. 2006.
Distal small fiber neuropathy: Results of tests of sweat- Ross syndrome: A rare or a misknown disorder of ther-
ing and autonomic cardiovascular reflexes. Muscle & moregulation? A skin innervation study on 12 subjects.
Nerve 15(6):661–5. Brain 129(Pt 8):2119–31.
22. Ross, A. T. 1958. Progressive selective sudomotor 34. Hilz, M. J., and M. Dutsch. 2006. Quantitative stud-
denervation: A case with coexisting Adie’s syndrome. ies of autonomic function. Muscle & Nerve 33(1):
Neurology 8(11):809. 6–20.
23. Fealey, R. D. 1997. Autonomic dysfunction in spinal 35. Attia, M., and P. Engel. 1983. Thermoregulatory set
cord disease. In Spinal cord disease basic science, diag- point in patients with spinal cord injuries (spinal man).
nosis and management, ed. E. Critchley, and A. Eisen, Paraplegia 21(4):233–48.
219–27. London, Great Britain: Springer-Verlag. 36. Kennedy, W. R., G. Wendelschafer-Crabb, and
24. Atkinson, J. L., and R. D. Fealey. 2003. Sympathotomy T. C. Brelje. 1994. Innervation and vasculature of
instead of sympathectomy for palmar hyperhidrosis: human sweat glands: An immunohistochemistry-
Minimizing postoperative compensatory hyperhidro- laser scanning confocal fluorescence microscopy
sis. Mayo Clinic Proceedings 78(2):167–72. study. The Journal of Neuroscience 14(11 Pt 2):
25. McCaffrey, T. V., R. D. Wurster, H. K. Jacobs, 6825–33.
D. E. Euler, and G. S. Geis. 1979. Role of skin tem- 37. Sato, K., and F. Sato. 1983. Individual variations in
perature in the control of sweating. Journal of Applied structure and function of human eccrine sweat gland.
Physiology 47(3):591–7. American Journal of Physiology 245(2): R203–8.
26. Sugenoya, J., and T. Ogawa. 1985. Characteristics of 38. Cole, D. E., and M. J. Boucher. 1986. Use of a new
central sudomotor mechanism estimated by frequency sample-collection device (Macroduct) in anion analysis
of sweat expulsions. Japanese Journal of Physiology of human sweat. Clinical Chemistry 32(7):1375–8.
35(5):783–94. 39. Morgan, C. J., P. S. Friedmann, M. K. Church, and
27. Ogawa, T., and M. Asayama. 1986. Quantitative analy- G. F. Clough. 2006. Cutaneous microdialysis as a
sis of the local effect of skin temperature on sweating. novel means of continuously stimulating eccrine sweat
Japanese Journal of Physiology 36(2):417–22. glands in vivo. The Journal of Investigative Dermatol-
28. Guttmann, L. 1947. The management of the ogy 126(6):1220–5.
quinizarin sweat test (Q.S.T.). Postgraduate Medical 40. Vernino, S., P. A. Low, R. D. Fealey, J. D. Stewart,
Journal 23:353–66. G. Farrugia, and V. A. Lennon. 2000. Autoantibodies
29. Hsieh, C., K. McNeeley, and T. C. Chelimsky. 2001. to ganglionic acetylcholine receptors in autoimmune
The clinical thermoregulatory sweat test induces autonomic neuropathies. The New England Journal of
maximal sweating. Clinical Autonomy Research Medicine 343(12):847–55.
11(4):227–34. 41. Fortney, S. M., E. R. Nadel, C. B. Wenger, and
30. Sato, K. T., A. Richardson, D. E. Timm, and K. Sato. J. R. Bove. 1981. Effect of blood volume on sweating
1988. One-step iodine starch method for direct visu- rate and body fluids in exercising humans. Journal of
alization of sweating. The American Journal of the Applied Physiology 51:1594–600.
Medical Sciences 295(6):528–31. 42. Low, P. A., H. Okazaki, and P. J. Dyck. 1977. Splanch-
31. Khurana, R. K. 1995. Oculocephalic sympathetic nic preganglionic neurons in man. 1. Morphometry of
dysfunction in posttraumatic headaches. Headache preganglionic cytons. Acta Neuropathologica (Berlin)
35(10):614–20. 40:55–61.
32. Bannister, R., L. Ardill, and P. Fentem. 1967. Defec- 43. Low, P. A. 1997. The effects of ageing on the auto-
tive autonomic control of blood vessels in idiopathic nomic nervous system. In Clinical autonomic disor-
orthostatic hypotension. Brain 90:725–46. ders, ed. P. A. Low, 2nd ed., 161–75. Philadelphia:
33. Kucera, P., Z. Goldenberg, and E. Kurca. 2004. Lippincott-Raven Publishers.
Sympathetic skin response: Review of the method 44. Attia, M., and P. Engel. 1983. Thermoregulatory set
and its clinical use. Bratislavske Lekarske Listy point in patients with spinal cord injuries (spinal man).
105(3):108–16. Paraplegia 21(4):233–48.
Chapter 39
Cardiovagal Reflexes
INTRODUCTION Technique
HEART RATE RESPONSE TO DEEP The Valsalva Ratio
BREATHING Vagal Baroreflex Sensitivity
Physiologic Basis Factors Affecting the Valsalva Response
Technique Pitfalls of the Valsalva Maneuver
Methods of Analysis
CARDIOVAGAL SCORING
Reproducibility
POWER SPECTRUM ANALYSIS
Factors Affecting the Heart Rate
HEART RATE RESPONSE TO
Response to Deep Breathing
STANDING
Problems and Controversies
OTHER TESTS OF CARDIOVAGAL
THE VALSALVA MANEUVER FUNCTION
Normal Response and Physiologic Basis SUMMARY
INTRODUCTION Divergent views exist regarding what types

of clinical studies constitute an adequate bat-
Vagal nerve traffic cannot be recorded directly tery of tests of autonomic function and what
in humans. Reliable, reproducible measure- are the optimal conditions for testing.2, 6 Nev-
ment of vagal function is possible indirectly, ertheless, a consensus has formed in regard
however, through noninvasive tests of heart to reasonable parameters for a number of
rate variability. The clinical relevance of heart basic methodologies for the evaluation of car-
rate variability in evaluating autonomic func- diovagal function that provide sensitive and
tion was first recognized in obstetrics as a reproducible results. The following description
marker of fetal distress.1 Although the phys- focuses on the most useful tests of cardio-
iologic basis of heart rate variability has vagal function. Their underlying physiology,
been known for a long time, it has been methodology, clinical utility, and shortcomings
mainly in the last two decades that cardio- are described.
vascular heart rate tests have been applied
clinically, thanks to the enthusiastic pro- Purpose and Role of Cardiovascular
motion by Ewing and associates.2 Clinical Heart Rate Testing
tests of cardiovagal function are now widely
used, although the full range of their appli- • These methodologies are used to assess
cation as well as their limitations remain the integrity of the vagal control of heart
underappreciated.3–5 rate.
661
• The heart rate response to deep breath- that the patient is relaxed and comfortable,
ing and to the Valsalva maneuver are because sympathetic activation diminishes the
easily assessed markers of cardiovagal response. Medications known to inhibit respi-
autonomic function. ratory sinus arrhythmia should be withheld if
• Cardiovagal failure is relevant to cardio- practical and safe to do so. The patient is tested
vascular status and is frequently a marker in the supine position to maximize vagal tone.
of autonomic disease. Either of two methods of training the res-
piratory cycle is customarily used. The more
common method is to have the subject visu-
HEART RATE RESPONSE TO DEEP ally follow and breathe in timing with a pattern,
BREATHING usually a sinusoidal oscillating bar generated by
a computer. The alternative is to instruct the
subject to breathe in and out as the technician
Physiologic Basis gestures. The difference of the two approaches
has not been studied systematically but is likely
The heart rate response to deep breathing to be minor. What is most important is that the
is probably the most reliable of the cardio- patient breathe slowly and continuously, rather
vascular heart rate tests, because the major than gasping or pausing, with full breaths, and
afferent and efferent pathways are both vagal.7 at a consistent rate of 5–6 cycles per minute.
The vagus nerve provides a beat-to-beat con- A standardized deep-breathing protocol can be
trol of the sinus node.8 This is mediated by used without the need to factor in variations in
M2 type muscarinic cholinergic receptors cou- the depth of respiration. This is because depth
pled to G-protein-activated inward rectifying of breathing above a tidal volume of approxi-
potassium channels (GIRKs). In humans, mus- mately 1.2 L causes insignificant changes in the
carinic receptor blockade with atropine com- heart rate response to deep breathing.13 Ben-
pletely abolishes respiratory sinus arrhythmia. nett and associates14 and Eckberg15 found little
The primary basis of respiratory sinus or no difference in the response for different
arrhythmia appears to be the interactions depths of respiration.
between the respiratory centers and the car- The electrocardiogram (ECG) is continu-
dioinhibitory centers in the medulla, partic- ously monitored. A computer program cap-
ularly the nucleus ambiguus.9 Evidence for tures each QRS complex over time and dis-
this is cessation of vagal efferent activity dur- plays beat-to-beat heart rate. Most laborato-
ing the inspiratory phase of natural—but not ries record the mean difference in R–R inter-
artificial—ventilation, and the loss of respira- vals over a series of six respiratory cycles.
tory sinus arrhythmia in some patients with Some investigators have advocated measur-
brain stem lesions.10 Respiratory sinus arrhyth- ing the heart rate response to one breath on
mia is modulated by input from the lungs, the basis that a single deep breath is a more
heart, and baroreceptors.11 Pulmonary stretch potent stimulus for heart rate change than
receptors that mediate the Hering–Breuer repeated deep breaths.16, 17 Direct comparisons
respiratory reflex modulate respiratory sinus have shown no advantage of a single deep
arrhythmia, although their role may be less breath over repeated breaths.18 Rather, assess-
important in humans than in experimental ani- ments of heart rate responses to a series of
mals. Receptors from the right atrium ini- breaths are more likely to be reproducible.
tiate the vagally mediated Bainbridge reflex
and a venoatrial mechanoreceptor sympathetic
reflex.9 Baroreflex sensitivity changes through- Methods of Analysis
out deep breathing, thus modulating respira-
tory sinus arrhythmia.12 The three methods of analysis generally used
are the heart rate range, the heart period
range, and the E:I ratio.10 The E:I ratio is
Technique the ratio of the shortest R–R interval during
inspiration to the longest R–R interval dur-
For testing the heart rate response to deep ing expiration. Our laboratory utilizes the heart
breathing, care should be taken to ensure rate range, because the effect of resting heart
Cardiovagal Reflexes 663
Figure 39–1. Heart rate response to deep breathing in a normal subject (top) and in a patient with diabetic autonomic
neuropathy (bottom). In each panel, the three upper tracings correspond to systolic blood pressure (SBP), mean blood
pressure (MBP), and diastolic blood pressure (DBP). The bottom tracing corresponds to heart rate (HR).
rate on the range is smaller than its effect on Reproducibility

heart period (Fig. 39–1).
Weinberg and Pfeifer19, 20 recommended cal- The tests have been reproducible. Typical
culation of the circular mean resultant, a coefficients of variation have been 11%14 and
method based on vector analysis, to eliminate 8.9%.21 When combined with a second, inde-
the effects of trends in heart rate over time pendent test of the parasympathetic control of
and to attenuate the effect of basal heart rate heart rate, the results are highly reliable in the
and ectopic beats in the calculated variability diagnosis of cardiovagal dysfunction.22
of heart rate. In this method, R–R intervals
over a 5-minute interval are plotted on a circu-
lar graph in which one revolution corresponds
to a single respiratory cycle. Normal heart rate Factors Affecting the Heart Rate
variability results in a clustering of time events, Response to Deep Breathing
whereas reduced heart rate variability results in
less periodically distributed time events on the Numerous factors affect the heart rate response
circle. to deep breathing. From the standpoint of
Table 39–1 Normative Data for Heart Rate Variability to Deep

Breathing
Percentile 20 years 40 years 60 years 80 years
2.5 13 9 7 7
5.0 14 10 7 7
95 41 33 27 27
97.5 43 36 29 29
Normative data were obtained from 376 subjects aged 10–83 and were derived by plotting the
regression of heart rate variation in beats/minute against age in years. A significant (p < 0.001) was
found in which y = 37.5448 ∗ log 10 0.9832x. Breakdown by gender was approximately equal.30
clinical autonomic testing, the most impor- number of studies have correlated cardiovagal
tant of these are the effects of age and rate control to depression, but the overall variance
of forced respiration. A progressive decrease is only about 2%.33
in the response with increasing age has been Prolonged hyperventilation and the reduc-
reported in all large studies.2, 16, 23–30 An inves- tion of Pco2 , however, result in a depression
tigation of 557 normal subjects, evenly dis- in respiratory sinus arrhythmia. There is a sym-
tributed by age and gender from 10 to 83 pathetic modulation of the heart rate response
years old, found that the decline in heart rate to deep breathing that is inhibited by stress
responses to deep breathing with age did not and enhanced by β-adrenergic blockade.34–36
regress to zero but leveled off in the older Also, the response is impaired during severe
subjects, indicating that it is possible to diag- tachycardia, in heart failure, and in deeply
nose cardiovagal failure in elderly patients in unconscious patients.10, 32, 34 Medications that
comparison to normative data30 (Table 39–1). have been shown to reduce the heart rate
Maximal heart rate response to deep breath- response to deep breathing include muscarinic
ing occurs at a breathing frequency of antagonists,37 nicotine,38 opioids,39 and nore-
5–6 respirations per minute in normal sub- pinephrine reuptake and serotonin-selective
jects.11, 14, 31, 32 This observation defines the basis reuptake inhibitors.40, 41
for the standard test of deep breathing.6 In
patients with vagal neuropathy, the maximal
heart rate response to deep breathing occurs Problems and Controversies
at lower respiratory frequencies, which is of
little pragmatic concern. Of greater practical The heart rate response to deep breathing is
importance is the selection for each subject of an indirect measure of cardiovagal function.
a respiratory rate in which the increase and A reduced response indicates a lesion any-
decrease are additive instead of subtractive.31 where in the intricate autonomic nervous sys-
A clue to waveform cancelation is the observa- tem; that is, in the afferent, central processing
tion of a large first or second response, followed unit, efferent, synapse, or effector apparatus.
by smaller responses. To further complicate interpretation, a
The position of the subject has some effect reduced response does not unequivocally indi-
on respiratory sinus arrhythmia. The response cate cardiovagal failure. The general observa-
is larger when the subject is supine than when tion that heart rate usually increases during
sitting or erect.14 Unlike some other tests of inspiration and decreases during expiration is
autonomic function, the heart rate response to an approximation of a composite set of phe-
deep breathing is not greatly affected by the sex nomena. Both inspiration and expiration are
of the subject.29, 30 The duration of antecedent followed by an increase, then a decrease, in
rest is not important in relaxed subjects. After 5 heart rate but at a different rate of change,
minutes of rest, another 25 minutes of supine amplitude, time of appearance, and duration.
rest will not alter the response. No significant Mehlsen and colleagues42 suggested that the
differences have been found in the response reason the maximal heart rate range in many
whether the test is performed in the morning subjects is 6 beats per minute is because
or in the afternoon in the same subjects.14 A they have well-defined heart rate maxima with
positive interference of phases. The reason sub- of arterial pressure with a photoplethysmo-
jects have a decreased heart rate range less graphic technique (Finapres) or tonometry
than 7 beats per minute is because of negative (Colin Pilot or Colin 7000) have provided
interference. important information about the hemodynamic
changes during the Valsalva maneuver in nor-
Key Points mal and pathologic conditions.47, 48
• Cardiovascular heart rate tests are useful, The responses to the Valsalva maneu-
ver have been divided into four phases43–49
sensitive, and reproducible tests of cardio-
(Fig. 39–2). Phase I consists of a brisk increase
vagal nerve function.
• The heart rate response to deep breath- in systolic and diastolic arterial pressure and
a decrease in heart rate immediately after the
ing is the most reliable test for evaluating
onset of the Valsalva strain and lasts approxi-
vagal afferent and efferent cardiac path-
mately 4 seconds. The increase in arterial pres-
ways.
• Vagal beat-to-beat control of the sinus sure during phase I reflects mechanical factors
and is not associated with an increase in sym-
mode (respiratory sinus arrhythmia) is
pathetic activity. It persists in patients with
mediated centrally by interactions between
transections of the high cervical spinal cord
the respiratory and the cardioinhibitory
and in normal subjects after administration of
centers in the medulla, and peripherally
α1 -adrenergic blocking drugs.49 The slowing of
by M2 muscarinic cholinergic receptors
the heart rate is reflexive and mediated by
coupled to GIRKs.
• Vagal efferent slowing of heart rate ceases increased parasympathetic efferent activity.50
Phase II consists of a decrease (early phase
during the inspiratory phase of natural,
II, IIe ) and subsequent partial recovery (late
but not artificial, respiration.
• The heart rate response to deep breath- phase II, IIl ) of arterial pressure and contin-
uous increase of heart rate during straining.
ing is best assessed by having the subject
Continuous straining impedes venous return
lie supine and take slow, continuous, full
to the heart and results in the displacement
breaths at a consistent rate of 5–6 breaths
of large amounts of blood from the thorax
per minute.
• A number of factors decrease the heart and abdomen to the limbs. The decrease in
venous return produces a reduction in left
rate response to deep breathing. These
atrial and left ventricular dimensions, left ven-
include (1) advancing age, (2) anxiety, (3)
tricular stroke volume, and cardiac output.43, 44
tachycardia, (4) heart failure, (5) uncon-
This triggers reflex compensatory tachycar-
sciousness, and (6) medications such
dia and vasoconstriction. The tachycardia dur-
as muscarinic antagonists, nicotine, opi-
ing phase II results from a prominent, early
oids, and norepinephrine reuptake and
component of inhibition of cardiovagal out-
serotonin-selective reuptake inhibitors.
put and is abolished with muscarinic blockade
with atropine.51 There is also a late contribu-
tion of increased sympathetic cardioaccelera-
THE VALSALVA MANEUVER tory output that is blocked with propranolol.
The progressive recovery of arterial pressure
The Valsalva maneuver is a global test of reflex during phase II reflects a similarly progressive
cardiovascular responses. It consists of an increase in total peripheral resistance45 caused
abrupt transient increase in intrathoracic and by increased sympathetic vasoconstrictor activ-
intra-abdominal pressures induced by blowing ity.52, 53 Increased arterial pressure during late
against pneumatic resistance while maintaining phase II is abolished by α1 -adrenergic blockade
a predetermined pressure (straining).43–46 with phentolamine.54 The fall in blood pressure
during phase II is more pronounced if com-
pensatory tachycardia is prevented by atropine
Normal Response and Physiologic and propranolol and if vasoconstriction is pre-
Basis vented by α1 -adrenergic blockade.54
Phase III consists of a sudden, brief (1–2
Intra-arterial recordings of arterial pressure seconds) further decrease in arterial pressure
and, more recently, noninvasive monitoring and increase in heart rate immediately after
Figure 39–2. Changes in systolic blood pressure (SBP) and heart rate (HR) during the Valsalva maneuver in a normal
subject (top) and in a patient with diabetic autonomic failure (bottom). The normal beat-to-beat SBP recording shows the
typical four phases (I, II, III, IV) of the Valsalva maneuver. The abnormal Valsalva maneuver is characterized by a profound
decrease in SBP in early phase II, absence of recovery in late phase II, and absence of SBP overshoot in phase IV. Note the
attenuated HR responses during phases I and IV, resulting in a reduced Valsalva ratio. EP, expiratory pressure.
the release of the straining. It is essentially in sympathetic nerve activity. The increase in
mechanical in nature and is the inverse of arterial pressure during phase IV can be pre-
phase I. vented by β-adrenergic blockade.54 Increases
Phase IV is characterized by increased sys- in both cardiac output and total peripheral
tolic and diastolic arterial pressure above con- resistance are important in producing the
trol levels, termed overshoot, accompanied by increase in arterial pressure in phase IV.
bradycardia relative to the control level of heart Recent pharmacologic evidence indicates
rate. In phase IV, venous return to the heart, that an increase in cardiac output-mediated
left ventricular stroke volume, and cardiac out- cardioacceleration is more important than
put return nearly to baseline, whereas the vasoconstriction in producing arterial pressure
arteriolar bed remains vasoconstricted because overshoot in phase IV. This overshoot is abol-
of the long time constant of sympathetic ished by β-blockade with propranolol but
responses.49 This combination results in an is maintained or even exaggerated during
overshoot of arterial pressure above baseline α-adrenergic blockade with phentolamine.44
values. Poststraining arterial pressure increases The increase in arterial pressure during phase
are proportional to the preceding increases IV stimulates the baroreceptors and results
Figure 39–3. Valsalva maneuvers in three different subjects showing beat-to-beat blood pressure (BP) and heart rate
(HR) in (A) a normal response, (B) impaired late phase II response and decreased Valsalva ratio in mild autonomic failure,
and (C) absent late phase II, absent phase IV, and absent HR response in severe autonomic failure.
in reflex bradycardia caused by increased continuously. Our laboratory47, 54, 57 and many
parasympathetic activity, which is abolished others58, 59 also monitor beat-to-beat arterial
with atropine.50, 55 Sympathetic inhibition after pressure with a noninvasive photoplethys-
straining persists much longer than the mographic technique. The “normal” Valsalva
increase in arterial pressure. response should be defined according to the
The responses during the four phases of the technique used in each laboratory, because sev-
Valsalva maneuver depend on the variable rela- eral technical variables affect the magnitude of
tionships between carotid and aortic barore- the response.
ceptor inputs. Pressure transients lasting only
seconds may reset the relationships between
the arterial pressure and the sympathetic or
vagal responses.56 Examples of the Valsalva The Valsalva Ratio
maneuvers in a normal subject and subjects
with mild and severe autonomic failure are The Valsalva ratio is defined as the longest
shown in Figure 39–3. R–R interval (pulse interval, in milliseconds)
after the maneuver (in phase IV) divided by
the shortest R–R interval during the maneu-
Technique ver (in phase II). In clinical settings, the
Valsalva maneuver commonly has been used
For testing the responses to the Valsalva to calculate the Valsalva ratio. The best of
maneuver, care should be taken, as in any other three responses is accepted. In more than
autonomic test, to ensure that the patient is 96% of control subjects, this ratio exceeds
well hydrated and is not taking medications 1.5.14, 32, 50, 51, 55 The Valsalva ratio decreases sig-
known to affect blood volume, cholinergic nificantly with age.29
function, or vasoreactivity. At our institution,
subjects are tested in the supine position and
asked to maintain a column of mercury at
40 mm Hg for 15 seconds through a bugle Vagal Baroreflex Sensitivity
with an air leak (to ensure an open glot-
tis). The responses are obtained in triplicate, The relationships between arterial pressure
and the largest response is accepted.47, 54, 57 In and heart rate during phase II and phase IV
some laboratories, only heart rate is monitored of the maneuver have been used to assess
Table 39–2 Normative Data for the Valsalva Ratio

Percentile 20 years 40 years 60 years 80 years
M F M F M F M F
2.5 1.50 1.41 1.36 1.47 1.21 1.36 1.21 1.36

5.0 1.59 1.46 1.44 1.51 1.29 1.39 1.29 1.39
95 2.87 2.73 2.52 2.64 2.18 2.41 2.18 2.41
97.5 2.97 2.97 2.60 2.88 2.23 2.65 2.23 2.65
Normative data were obtained by plotting the regression of Valsalva ratio against age in 425 subjects of 10–83 years.30
baroreflex sensitivity.60 Vagal baroreflex sensi- after release of the straining. Normative data
tivity, which is the change in heart rate result- on the phases of the Valsalva maneuver have
ing from a change in blood pressure, can been published30, 57, 64 (Table 39–2).
be quantified by plotting each heart interval
against the preceding systolic blood pressure
value. The slope of the simple linear regres-
sion is a measure of baroreflex sensitivity.61 EFFECTS OF POSTURE
One limitation of this method is that there is
In subjects in the supine position, changes
an unknown reflex latency between the pres-
in arterial pressure during phases II and IV
sure change and the change in heart period.
may be modest, because the large intrathoracic
Thus, the best of two regression slopes, one
blood volume may buffer the reduced venous
correlating blood pressure to its correspond-
return during phase II. In the supine position,
ing heart period and the other to its subse-
some normal subjects may show a square-wave
quent pulse interval, should be accepted. Vagal
response similar to that of patients with conges-
baroreflex sensitivity declines with age.62 The
tive heart failure. Changing from a supine to an
adrenergic component of baroreflex sensitivity,
inclined or upright posture increases the mag-
which relates the blood pressure recovery time
nitudes of the arterial pressure decrease during
to the preceding decrease in blood pressure,
phase II, the systolic blood pressure overshoot
correlates more closely than vagal baroreflex
during phase IV, and the Valsalva ratio.59
sensitivity to adrenergic failure.63
Factors Affecting the Valsalva EFFECTS OF TEST DURATION

Response The duration of straining during the Valsalva
maneuver differentially effects the vagally and
The cardiovascular changes during the Val- sympathetically mediated responses, because
salva maneuver are determined mainly by the of their different latencies and time constants.
magnitude of hemodynamic change during the When the Valsalva maneuver is performed at
forced expiratory effort, the time course and low expiratory pressures (20 mm Hg), the mag-
efficiency of reflex cardiovagal and sympa- nitude of the tachycardia in phase II is inde-
thetic vasomotor and cardiomotor responses, pendent of the duration of the test, consistent
and the modification of these responses by with the short latency of vagal responses. The
interactions with respiratory mechanisms at maximal increase in arterial pressure in late
both central and peripheral levels. Accordingly, phase II and phase IV correlates with the dura-
the responses to the Valsalva maneuver may tion of the Valsalva maneuver. This may reflect
be affected by (1) the position of the sub- the longer latency of sympathetic vasoconstric-
ject during the maneuver, (2) the magnitude tor and cardioacceleratory responses.49 A test
and duration of the straining, (3) the breathing duration of 10 seconds is effective, while 15
pattern before and after the maneuver, includ- seconds is a practical optimum and may be
ing depth and phase of respiration preceding sufficient to assess sympathetically mediated
the straining, and (4) the control of respiration responses in a clinical setting.47
EFFECTS OF EXPIRATORY PITFALLS IN INTERPRETING THE

PRESSURE VALSALVA RATIO
The magnitude of most heart rate and arterial There is evidence that the Valsalva ratio in nor-
pressure responses during phase II and phase mal subjects depends mainly on cardiovagal
IV correlates with the magnitude of expira- function.32, 50, 51, 55 However, the interpretation
tory pressure used during the Valsalva maneu- of this ratio as a test of cardiovagal func-
ver. Maximal arterial pressure and heart rate tion without simultaneous recordings of arte-
responses are obtained with expiratory pres- rial pressure may be misleading for several
sures of 40–50 mm Hg.46, 50, 51, 55 reasons:
PHASE OF RESPIRATION 1. The Valsalva ratio correlates better with

the heart rate response in phase II than
In normal subjects, the magnitude of heart with the response in phase IV.47 There-
rate responses during the Valsalva maneuver is fore, if sufficient tachycardia is present in
significantly lower if the expiratory strain is pre- phase II, the Valsalva ratio may be “nor-
ceded by maximal inspiration instead of tidal mal” even in the absence of significant
inspiration.50 bradycardia in phase IV. This may occur
in patients with cardiovagal impairment
but intact sympathetic innervation.65
Pitfalls of the Valsalva Maneuver 2. The Valsalva ratio also correlates with the
magnitude of arterial pressure overshoot
PITFALLS IN PERFORMING THE in phase IV.47 The absence of bradycar-
VALSALVA MANEUVER dia in phase IV, and thus an abnormal
Valsalva ratio, may be caused not only by
There are several precautions to keep in mind vagal dysfunction but also by the inability
when asking patients to perform the Valsalva to increase arterial pressure in phase IV.
maneuver:46 3. Both the heart rate increase in phase II
and the heart rate decrease in phase IV
1. The test requires patient cooperation and, are affected critically by the magnitude
thus, cannot be performed in patients of the decrease in venous return during
who are seriously ill or who have weak res- the Valsalva maneuver, which depends on
piratory, facial, or oropharyngeal muscles. the position of the subject and the pooling
2. The maneuver should be avoided in and buffering effect of thoracic vessels.59
patients with proliferative retinopathy, 4. Assessing the integrity of the total barore-
because of the risk of intraocular hemor- flex arc during the Valsalva maneuver by
rhage, and in patients with known cere- only testing the Valsalva ratio is unre-
bral aneurysms, because of the risk of liable, because the magnitude and time
subarachnoid hemorrhage. course of the heart rate response may
3. Theoretically, the Valsalva maneuver can be normal despite a response of arterial
precipitate arrhythmias and angina and pressure typical of sympathetic failure.65
may cause syncope, particularly in elderly
patients with impaired reflex mechanisms The integrity of reflex sympathetic responses
that respond to the decrease in venous cannot be inferred on the basis of the Val-
return. salva ratio. Both the magnitude of the decrease
4. Patients with congestive heart failure, in mean arterial pressure in phase II and
mitral stenosis, aortic stenosis, constric- the overshoot of arterial pressure in phase IV
tive pericarditis, or atrial septal defect have been considered indices of vasomotor
may have an abnormal square-wave function.49, 54 However, the magnitude of the
response of arterial pressure to the Val- decrease in arterial pressure in early phase II
salva maneuver because of the increase in is also affected by the heart rate responses,
pulmonary blood volume, which is capa- and the overshoot of arterial pressure in phase
ble of maintaining ventricular filling dur- IV may be more dependent on cardiac out-
ing the Valsalva strain. put.54 In our laboratory, the changes in arterial
pressure during early and late phase II and undergo an increase in intrathoracic pres-
phase IV are used to assess sympathetic vaso- sure, such as patients with large cerebral
motor function.54, 57 Late phase II is impaired aneurysms or some patients with prolifer-
in patients with α-adrenergic failure caused, ative retinopathy.
for example, by dopamine-β-hydroxylase defi- • Interpretation of the Valsalva ratio as a
ciency.66, 67 test of cardiovagal function may be mis-
leading without simultaneous beat-to-beat
recording of arterial pressure to ensure an
Key Points
adequate early phase II profile.
• The Valsalva maneuver consists of forced
voluntary expiratory effort against resis-
tance, which displaces blood from the
thorax and abdomen into the limbs and
CARDIOVAGAL SCORING
results in a complex set of cardiovascular
Low and associates have developed and vali-
autonomic responses.
• The autonomic responses to the Valsalva dated a 10-point composite autonomic severity
score (CASS), which grades autonomic fail-
maneuver are divided into four phases:
◦ Phases I and III reflect transient ure according to the results of clinical auto-
nomic testing while correcting for the con-
mechanical effects on blood pressure of
founding effects of age and gender. In addition
the onset and cessation of straining.
◦ Phase II consists of a decrease in arte- to adrenergic (maximum 4 points) and sudo-
motor (maximum 3 points) subscores, a cardio-
rial pressure (IIe ) and a subsequent
vagal subscore (maximum 3 points) is assigned.
partial recovery (IIl ) caused by (1)
A subscore of 1 is assigned if heart rate vari-
reflex compensatory tachycardia due to
ability to deep breathing is mildly reduced but
inhibition of cardiovagal output and
above 50% of the minimum value. A subscore
increased sympathetic cardioaccelera-
of 2 is assigned if heart rate variability is below
tory output and (2) increased sympa-
50% of the minimum value. A subscore of
thetic vasoconstrictor activity with pro-
3 indicates that both the heart rate variabil-
gressive increase in total peripheral
ity to deep breathing and the Valsalva ratio
resistance.
◦ Phase IV consists of (1) reflex parasym- are reduced to below 50% of their minimum
normative values.68
pathetically mediated bradycardia and
(2) increased cardiac output-mediated
cardioacceleration while the arteriolar Key Points
bed remains vasoconstricted, resulting • The CASS is a validated index of auto-
in an overshoot of blood pressure above
nomic failure that combines the results of
baseline.
• The subject, tested supine, is asked to clinical autonomic testing correcting for
differences of age and gender.
maintain a column of mercury at 40 mm
for 15 seconds through a paper bugle
with a small air leak (to ensure an open
glottis), while beat-to-beat blood pressure POWER SPECTRUM ANALYSIS
and heart rate responses are continuously
monitored. Autonomic data generally are evaluated in the
• The Valsalva ratio is an index of cardiova- time domain, with a focus on the changes
gal function and is defined as the longest over time in the amplitude of a response to
R–R interval during phase IV divided by a standardized stimulus. Spontaneous oscilla-
the shortest R–R interval during phase II. tions also contain key information. Frequency
• The Valsalva ratio may be decreased (1) in analysis focuses on the changes in amplitude
older subjects, (2) if the force and duration as a function of frequency.69 In recordings of
of respiratory straining are insufficient, or the heart period (the reciprocal of heart rate),
(3) in sitting or standing postures. oscillations at the respiratory frequency (typ-
• The Valsalva maneuver should not be ically approximately 0.25 Hz) are determined
performed in subjects who cannot safely by parasympathetic function; hence, its power
(amplitude) reflects the proportion of frequen- • When sampling heart rate for power
cies due to parasympathetic activity. A slower spectrum analysis, respiration should be
frequency, approximately 0.07–0.1 Hz, reflects recorded or paced, because respiration
the periodicity of the baroreflex loop. Power can entrain heart rate oscillations.
at this frequency reflects both sympathetic and
parasympathetic functions. Similar oscillations
occur in blood pressure recordings.
Several methods are available for evaluat- HEART RATE RESPONSE TO
ing autonomic signals in the frequency domain. STANDING
Fast Fourier transform and autoregressive
models are commonly used. Both of these The immediate heart rate response to stand-
require stationarity, a condition that is diffi- ing can be recorded using an ECG machine.
cult to satisfy with changing autonomic sig- In normal subjects, tachycardia is maximal at
nals. An alternative approach is time-frequency about the 15th beat, and relative bradycardia
analysis, a method that resolves signals in occurs around the 30th beat.72 The 30:15 ratio
both the time and the frequency domains (R–R interval at beat 30/R–R interval at beat
simultaneously. 15) has been recommended as an index of car-
Head-up tilt results in attenuation of the res- diovagal function. Reflex tachycardia is thought
piratory frequency and augmentation of the to be mediated by the vagus nerve, because
lower frequency. An advantage of frequency the response is abolished by atropine but not
analysis is its ability to evaluate sympathovagal by propranolol.72 The heart rate response to
balance. It can be expressed as the power in standing attenuates with age.73
the low frequency in blood pressure (reflecting The hemodynamic response to active stand-
pure sympathetic function) over the respira- ing is trimodal.74 There is an abrupt increase
tory frequency in heart period (pure parasym- in heart rate that peaks at 3 seconds, a fur-
pathetic function).70 ther more gradual tachycardia that peaks at
For clinical recordings of autonomic signals, 12 seconds, and a bradycardia at 20 seconds.
it is essential that respiration be recorded or The initial cardioacceleration is an exercise
paced, because respiration can entrain heart reflex mediated by muscle contraction, result-
rate oscillations over a wide range of frequen- ing in the sudden inhibition of vagal tone. The
cies (from 0.01 to 0.5 Hz), and if the sub- second tachycardia is caused by further vagal
ject breathes slowly, respiratory rhythms will inhibition and by reduced baroreflex activity,
eclipse the lower frequency signals.71 Either which is due to transient hypotension caused
an increased tidal volume at lower respiratory by a fall in peripheral vascular resistance. This
rates or breath-holding can interact with spec- transient hypotension appears to be the result
tral power at lower frequencies and greatly bias of local, nonautonomically mediated vasodi-
the estimation of power. Slowing of respiration latation in contracting muscles, central com-
can thus lead to a falsely positive increase in mand, and cholinergic-mediated vasodilata-
low-frequency power.71 A minimal duration of tion.75–77 An overshoot in blood pressure then
recording for valid analysis is at least 5 minutes evokes baroreflex-mediated bradycardia. The
of good quality recording. series of autonomic responses to standing thus
not only reflects the integrity of the cardiovagal
system, but also depends on an intact sympa-
Key Points thetic nervous system, local muscle reflexes,
baroreflexes, and central command.
• Power spectral analysis of heart rate vari-
ability is used to evaluate cardiac auto-
nomic function. Analysis is performed in Key Points
the time domain or the frequency domain.
• High-frequency spectral power (near • The 30:15 ratio is a test of cardiovagal
0.25 Hz) reflects parasympathetic cardio- function in which, upon active standing,
vagal tone, whereas low-frequency spec- the R–R interval at beat 30 is divided by
tral power (0.07–0.10 Hz) reflects baro- the R–R interval at beat 15 as recorded by
reflex modulation of autonomic outflow. an ECG machine.
OTHER TESTS OF CARDIOVAGAL evoked by vasoactive drugs.88–90 Alternatively,

FUNCTION baroreflex sensitivity has been estimated by
baroslopes generated by negative pressure
Available tests of cardiovagal function are applied to the carotid sinus by a finely regu-
numerous. In a reversal of testing the heart rate lated neck suction device.91, 92
response to standing, a related test measures
the heart rate response to lying down. Recum- Key Points
bency evokes an immediate, vagally mediated
decrease in the R–R interval that is maximal at • Additional, infrequently used tests of
the third or fourth beat. This is followed by a cardiovagal function include heart rate
sympathetically mediated increase in the R–R responses to (1) lying down, (2) squatting,
interval at the 25th–30th beat.77–79 (3) coughing, (4) facial immersion or cool-
The heart rate response to squatting has ing, (5) infusion of vasoactive drugs, or (6)
been defined in normal subjects in compari- carotid sinus external suction.
son to its decline in diabetic autonomic neu-
ropathy.80 In this test, the subject stands still
for 3 minutes, squats down for 1 minute, and SUMMARY
then stands up during inspiration. The normal
response is vagally mediated lengthening of Noninvasive cardiovascular tests are reliable
the R–R interval during squatting, followed by and reproducible and are widely used to eval-
sympathetically mediated shortening of the R– uate autonomic function in human subjects.
R interval at standing. The response declines The heart rate response to deep breathing is
with age.80 probably the most reliable test for assessing
Coughing, which involves inspiration and an the integrity of the vagal afferent and efferent
expiratory effort against a closed glottis, fol- pathways to the heart. This is because respira-
lowed by explosive expiration as the glottis tory sinus arrhythmia is a relatively pure test
opens, generates a transient intrathoracic pres- of cardiovagal function, whereas many other
sure gradient of 25–250 mm Hg. This evokes conditions, such as plasma volume, antecedent
a decrease in R–R interval, which is maximal rest, and cardiac and peripheral sympathetic
at 2–3 seconds and returns to resting value in functions, factor into the Valsalva response.
approximately 12–14 seconds.81, 82 The cough- Heart rate variability to deep breathing is usu-
induced cardioacceleration is primarily under ally tested at a breathing frequency of 5 or
cholinergic control, although it is partly related 6 respirations per minute and decreases lin-
to the intense contraction of abdominal and early with age. The Valsalva maneuver consists
expiratory muscles.82 The cough test has been of a forced expiratory effort against resistance
applied to the evaluation of cardiac parasympa- and produces mechanical (phases I and III)
thetic integrity.83 Comparison to conventional and reflex (phases II and IV) changes in arte-
tests of cardiovagal function such as the stand- rial pressure and heart rate. When performed
ing test and the Valsalva maneuver has shown under continuous arterial pressure monitor-
no advantage to the cough test.84 ing with a noninvasive technique, the Valsalva
The diving reflex is provoked by facial cool- maneuver provides valuable information about
ing and consists of reflex bradycardia, apnea, the integrity of the cardiac parasympathetic,
and vasoconstriction. The bradycardia depends cardiac sympathetic, and sympathetic vasomo-
on a trigeminal–cardiovagal reflex in which tor outputs. The responses to the Valsalva
stimulation of the sensory branches of the maneuver are affected by the position of the
trigeminal nerve evokes an efferent response subject and the magnitude and duration of the
in the motor nucleus of the vagus nerve.85, 86 expiratory effort. In general, it is performed
A practical alternative to facial immersion in at an expiratory pressure of 40 mm Hg sus-
cold water (the diving test) is the application of tained for 15 seconds. The Valsalva ratio, the
cold compresses to the face (the cold face test), relationship between the maximal heart rate
which avoids the potential for aspiration.87 response during phase II (straining) and phase
Baroreflex sensitivity has also been assessed IV (after release of straining), has been consid-
by regressing the change in heart period to a ered a test of cardiac parasympathetic function.
titrated decrease or increase in blood pressure However, without simultaneous recording of
arterial pressure, this may be misleading. An pressure. Scandinavian Journal of Clinical and Labo-
exaggerated decrease in arterial pressure dur- ratory Investigation 35:487–96.
ing phase II suggests sympathetic vasomotor 14. Bennett, T., P. H. Fentem, D. Fitton, J. R. Hamptom,
D. J. Hosking, and P. A. Riggott. 1977. Assessment of
failure, whereas an absence of overshoot dur- vagal control of the heart in diabetes. Measures of R–R
ing phase IV indicates the inability to increase interval variation under different conditions. British
cardiac output and cardiac adrenergic failure. Heart Journal 39:25–8.
15. Eckberg, D. L. 1983. Human sinus arrhythmia as
an index of vagal cardiac outflow. Journal of Applied
Physiology 54:961–6.
REFERENCES 16. O’Brien, I. A., P. O’Hare, and R. J. Corrall. 1986. Heart
rate variability in healthy subjects: Effect of age and
1. Hon, E. H., and S. T. Lee. 1965. Electronic evaluation the derivation of normal ranges for tests of autonomic
of the fetal heart rate patterns preceding fetal death: function. British Heart Journal 55:348–54.
Further observations. American Journal of Obstetrics 17. Bennett, T., I. K. Farquhar, D. J. Hosking, and J. R.
and Gynecology 87:814–26. Hampton. 1978. Assessment of methods for estimating
2. Ewing, D. J., C. N. Martyn, R. J. Young, and autonomic nervous control of the heart in patients with
B. F. Clarke. 1985. The value of cardiovascular auto- diabetes mellitus. Diabetes 27:1167–74.
nomic function tests: 10 years experience in diabetes. 18. Espi, F., D. J. Ewing, and B. F. Clarke. 1982. Testing
Diabetes Care 8:491–8. for heart rate variation in diabetes: Single or repeated
3. van Lieshout, J. J., W. Wieling, K. H. Wesseling, deep breaths? Acta Diabetologica Latina 19:177–81.
and J. M. Karemaker. 1989. Pitfalls in the assess- 19. Weinberg, C. R., and M. A. Pfeifer. 1984. An
ment of cardiovascular reflexes in patients with sympa- improved method for measuring heart-rate variability:
thetic failure but intact vagal control. Clinical Science Assessment of cardiac autonomic function. Biometrics
76:523–8. 40:855–61.
4. Wieling, W., and J. J. van Lieshout. 1990. The assess- 20. Genovely, H., and M. A. Pfeifer. 1988. RR-variation:
ment of cardiovascular reflex activity: Standardization The autonomic test of choice in diabetes. Dia-
is needed. Diabetologia 33:182–3. betes/Metabolism Reviews 4:255–71.
5. Low, P. A. 1997. Pitfalls in autonomic testing. In 21. Smith, S. A. 1982. Reduced sinus arrhythmia in dia-
Clinical autonomic disorders: Evaluation and manage- betic autonomic neuropathy: Diagnostic value of an
ment, ed. P. A. Low, 2nd ed., 391–401. Philadelphia: age-related normal range. British Medical Journal
Lippincott-Raven Publishers. 285:1599–601.
6. Low, P. A., and M. A. Pfeifer. 1997. Standardization of 22. Wieling, W. 1983. Reduced sinus arrhythmia in dia-
autonomic function. In Clinical autonomic disorders: betic autonomic neuropathy. British Medical Journal
Evaluation and management, ed. P. A. Low, 2nd ed., 286:1285.
287–95. Philadelphia: Lippincott-Raven Publishers. 23. Bergstrom, B., B. Lilja, K. Rosberg, and G. Sund-
7. Katona, P. G., and F. Jih. 1975. Respiratory kvist. 1986. Autonomic nerve function tests. Refer-
sinus arrhythmia: Noninvasive measure of parasympa- ence values in healthy subjects. Clinical Physiology
thetic cardiac control. Journal of Applied Physiology 6:523–8.
39:801–5. 24. Ingall, T. J. 1986. Autonomic nervous system function
8. Jalife, J., and D. C. Michaels. 1994. Neural control of in ageing and in diseases of the peripheral nervous
sinoatrial pacemaker activity. In Vagal control of the system. Thesis, University of Sydney.
heart: Experimental basis and clinical implications, ed. 25. Kaijser, L., and C. Sachs. 1985. Autonomic cardio-
M. N. Levy, and P. J. Schwartz, 173–205. New York: vascular responses in old age. Clinical Physiology
Futura Publishing Company. 5:347–57.
9. Spyer, K. M. 1999. Central nervous control of the car- 26. Masaoka, S., A. Lev-Ran, L. R. Hill, G. Vakil, and
diovascular system. In Autonomic failure: A textbook of E. H. Hon. 1985. Heart rate variability in diabetes:
clinical disorders of the autonomic nervous system, ed. Relationship to age and duration of the disease. Dia-
C. J. Mathias, and R. Bannister, 4th ed., 45–55. New betes Care 8:64–8.
York: Oxford University Press. 27. Wieling, W., J. F. van Brederode, L. G. de Rijk,
10. Vallbona, C., D. Cardus, W. A. Spencer, and C. Borst, and A. J. Dunning. 1982. Reflex control of
H. E. Hoff. 1965. Patterns of sinus arrhythmia in heart rate in normal subjects in relation to age: A data
patients with lesions of the central nervous system. The base for cardiac vagal neuropathy. Diabetologia 22:
American Journal of Cardiology 16:379–89. 163–6.
11. Saul, J. P., and R. J. Cohen. 1994. Respiratory sinus 28. Clark, C. V., and R. Mapstone. 1986. Age-adjusted
arrhythmia. In Vagal control of the heart: Experimen- normal tolerance limits for cardiovascular autonomic
tal basis and clinical implications, ed. M. N. Levy, and function assessment in the elderly. Age and Ageing
P. J. Schwartz, 511–36. New York: Futura Publishing 15:221–9.
Company. 29. Low, P. A., T. L. Opfer-Gehrking, C. J. Proper, and
12. Eckberg, D. L., Y. T. Kifle, and V. L. Roberts. 1980. I. Zimmerman. 1990. The effect of aging on cardiac
Phase relationship between normal human respira- autonomic and postganglionic sudomotor function.
tion and baroreflex responsiveness. The Journal of Muscle & Nerve 13:152–7.
Physiology (London) 304:489–502. 30. Low, P. A., J. C. Denq, T. L. Opfer-Gehrking, P. J.
13. Freyschuss, U., and A. Melcher. 1975. Sinus arrhyth- Dyck, P. C. O’Brien, and J. M. Slezak. 1997. Effect of
mia in man: Influence of tidal volume and oesophageal age and gender on sudomotor and cardiovagal function
and blood pressure response to tilt in normal subjects. 47. Benarroch, E. E., T. L. Opfer-Gehrking, and P. A. Low.
Muscle & Nerve 20:1561–8. 1991. Use of the photoplethysmographic technique to
31. Angelone, A., and N. A. Coulter, Jr. 1964. Respira- analyze the Valsalva maneuver in normal man. Muscle
tory sinus arrhythmia: A frequency dependent phe- & Nerve 14:1165–72.
nomenon. Journal of Applied Physiology 19:479–82. 48. Imholz, B. P., G. A. van Montfrans, J. J. Settels,
32. Pfeifer, M. A., D. Cook, J. Brodsky, et al. 1982. Quan- G. M. van der Hoeven, J. M. Karemaker, and
titative evaluation of cardiac parasympathetic activity W. Wieling. 1988. Continuous non-invasive blood
in normal and diabetic man. Diabetes 31:339–45. pressure monitoring: Reliability of Finapres device
33. Rottenberg, J. 2007. Cardiac vagal control in during the Valsalva manoeuvre. Cardiovascular
depression: A critical analysis. Biological Psychology Research 22: 390–7.
74:200–11. 49. Korner, P. I., A. M. Tonkin, and J. B. Uther. 1976.
34. Coker, R., A. Koziell, C. Oliver, and S. E. Smith. 1984. Reflex and mechanical circulatory effects of graded
Does the sympathetic nervous system influence sinus Valsalva maneuvers in normal man. Journal of Applied
arrhythmia in man? Evidence from combined auto- Physiology 40:434–40.
nomic blockade. The Journal of Physiology (London) 50. Eckberg, D. L. 1980. Parasympathetic cardiovascular
356:459–64. control in human disease: A critical review of meth-
35. Pitzalis, M. V., F. Mastropasqua, F. Massari, et al. 1998. ods and results. The American Journal of Physiology
Beta-blocker effects on respiratory sinus arrhyth- 239:H581–93.
mia and baroreflex gain in normal subjects. Chest 51. Leon, D. F., J. A. Shaver, and J. J. Leonard. 1970.
114:185–91. Reflex heart rate control in man. American Heart
36. Elghozi, J. L., and C. Julien. 2007. Sympathetic control Journal 80:729–39.
of short-term heart rate variability and its pharmaco- 52. Delius, W., K. E. Hagbarth, A. Hongell, and B. G.
logic modulation. Fundamental & Clinical Pharmacol- Wallin. 1972. Manoeuvres affecting sympathetic out-
ogy 21:337–47. flow in human muscle nerves. Acta Physiologica Scan-
37. Lacroix, D., R. Logier, S. Kacet, et al. 1992. Effects dinavica 84:82–94.
of consecutive administration of central and peripheral 53. Wallin, B. G., and D. L. Eckberg. 1982. Sympa-
anticholinergic agents on respiratory sinus arrhythmia thetic transients caused by abrupt alterations of carotid
in normal subjects. Journal of the Autonomic Nervous baroreceptor activity in humans. The American Jour-
System 39:211–17. nal of Physiology 242:H185–90.
38. Wang, J., X. Wang, M. Irnaten, et al. 2003. Endoge- 54. Sandroni, P., E. E. Benarroch, and P. A. Low. 1991.
nous acetylcholine and nicotine activation enhances Pharmacological dissection of components of the Val-
GABAergic and glycinergic inputs to cardiac vagal salva maneuver in adrenergic failure. Journal of
neurons. Journal of Neurophysiology 89:2473–81. Applied Physiology 71:1563–7.
39. Loula, P., V. Jäntti, and A. Yli-Hankala. 1997. Respira- 55. Levin, A. B. 1966. A simple test of cardiac function
tory sinus arrhythmia during anaesthesia: Assessment based upon the heart rate changes induced by the Val-
of respiration related beat-to-beat heart rate variabil- salva maneuver. The American Journal of Cardiology
ity analysis methods. International Journal of Clinical 18:90–9.
Monitoring and Computing 14:241–9. 56. Smith, M. L., L. A. Beightol, J. M. Fritsch-Yelle,
40. Davidson, J., L. Watkins, M. Owens, et al. 2005. K. A. Ellenbogen, T. R. Porter, and D. L. Eckberg.
Effects of paroxetine and venlafaxine XR on heart 1996. Valsalva’s maneuver revisited: A quantita-
rate variability in depression. Journal of Clinical Psy- tive method yielding insights into human auto-
chopharmacology 25:480–4. nomic control. The American Journal of Physiology
41. Low, P. A., and T. L. Opfer-Gehrking. 1992. Differen- 271:H1240–9.
tial effects of amitriptyline on sudomotor, cardiovagal, 57. Denq, J. C., P. C. O’Brien, and P. A. Low.
and adrenergic function in human subjects. Muscle & 1998. Normative data on phases of the Valsalva
Nerve 15:1340–4. maneuver. Journal of Clinical Neurophysiology 15:
42. Mehlsen, J., K. Pagh, J. S. Nielsen, L. Sestoft, and 535–40.
S. L. Nielsen. 1987. Heart rate response to breath- 58. Parati, G., R. Casadei, A. Groppelli, M. Di Rienzo,
ing: Dependency upon breathing pattern. Clinical and G. Mancia. 1989. Comparison of finger and
Physiology 7:115–24. intra-arterial blood pressure monitoring at rest
43. Booth, R. W., J. M. Ryan, H. C. Mellet, E. Swiss, and during laboratory testing. Hypertension 13:
and E. Neth. 1962. Hemodynamic changes asso- 647–55.
ciated with the Valsalva maneuver in normal men 59. Ten Harkel, A. D., J. J. Van Lieshout, E. J. Van
and women. The Journal of Laboratory and Clinical Lieshout, and W. Wieling. 1990. Assessment of car-
Medicine 59:275–85. diovascular reflexes: Influence of posture and period
44. Brooker, J. Z., E. L. Alderman, and D. C. Harrison. of preceding rest. Journal of Applied Physiology
1974. Alterations in left ventricular volumes induced 68:147–53.
by Valsalva manoeuvre. British Heart Journal 60. Kautzner, J., J. E. Hartikainen, A. J. Camm, and
36:713–18. M. Malik. 1996. Arterial baroreflex sensitivity assessed
45. Sharpey-Schafer, E. P. 1955. Effects of Valsalva’s from phase IV of the Valsalva maneuver. The American
manoeuvre on the normal and failing circulation. Journal of Cardiology 78:575–9.
British Medical Journal 1:693–5. 61. Trimarco, B., M. Volpe, B. Ricciardelli, et al. 1983.
46. Nishimura, R. A., and A. J. Tajik. 1986. The Val- Valsalva maneuver in the assessment of baroreflex
salva maneuver and response revisited. Mayo Clinic responsiveness in borderline hypertensives. Cardiol-
Proceedings 61:211–17. ogy 70:6–14.
62. Huang, C. C., P. Sandroni, D. M. Sletten, et al. 2007. at the beginning of isometric exercise in humans.
Effect of age on adrenergic and vagal baroreflex sensi- Circulation 79:815–24.
tivity in normal subjects. Muscle & Nerve 36:637–42. 78. Bellavere, F., and D. J. Ewing. 1982. Autonomic con-
63. Schrezenmaier, C., W. Singer, N. M. Swift, et al. trol of the immediate heart rate response to lying
2007. Adrenergic and vagal baroreflex sensitivity in down. Clinical Science 62:57–64.
autonomic failure. Archives of Neurology 64:381–6. 79. Bellavere, F., C. Cardone, M. Ferri, et al. 1987. Stand-
64. Faulkner, M. S., D. Hathaway, and B. Tolley. 2003. ing to lying heart rate variation. A new simple test
Cardiovascular autonomic function in healthy adoles- in the diagnosis of diabetic autonomic neuropathy.
cents. Heart & Lung 32:10–22. Diabetic Medicine 4:41–3.
65. Opfer-Gehrking, T. L., and P. A. Low. 1993. Impaired 80. Marfella, R., D. Giugliano, G. di Maro, R. Acampora,
respiratory sinus arrhythmia with paradoxically nor- R. Giunta, and F. D’Onofrio. 1994. The squatting
mal Valsalva ratio indicates combined cardiovagal test. A useful tool to assess both parasympathetic
and peripheral adrenergic failure. Clinical Autonomic and sympathetic involvement of the cardiovascu-
Research 3:169–73. lar autonomic neuropathy in diabetes. Diabetes 43:
66. Biaggioni, I., D. S. Goldstein, T. Atkinson, and 607–12.
D. Robertson. 1990. Dopamine-beta-hydroxylase defi- 81. Wei, J. Y., J. W. Rowe, A. D. Kestenbaum, and
ciency in humans. Neurology 40:370–3. S. Ben-Haim. 1983. Post-cough heart rate response:
67. Cheshire, W. P., D. W. Dickson, K. F. Nahm, et al. Influence of age, sex, and basal blood pressure. The
2006. Dopamine beta-hydroxylase deficiency involves American Journal of Physiology 245:R18–24.
the central autonomic network. Acta Neuropatholog- 82. Cardone, C., F. Bellavere, M. Ferri, and D. Fedele.
ica (Berlin) 112:227–9. 1987. Autonomic mechanisms in the heart rate
68. Low, P. A. 1997. Laboratory evaluation of autonomic response to coughing. Clinical Science 72:55–60.
function. In Clinical autonomic disorders: Evalua- 83. Cardone, C., P. Paiusco, G. Marchetti, et al. 1990.
tion and management, ed. P. A. Low, 2nd ed., 191–2. Cough test to assess cardiovascular autonomic reflexes
Philadelphia: Lippincott-Raven Publishers. in diabetes. Diabetes Care 13:719–24.
69. Novak, V., P. Novak, and P. A. Low. 1997. Time- 84. van Lieshout, E. J., J. J. van Lieshout, A. D. ten Harkel,
frequency analysis of cardiovascular function and its and W. Wieling. 1989. Cardiovascular responses
clinical applications. In Clinical autonomic disorders: to coughing: Its value in the assessment of auto-
Evaluation and management, ed. P. A. Low, 2nd ed., nomic nervous control. Clinical Science (London) 77:
323–48. Philadelphia: Lippincott-Raven Publishers. 305–10.
70. Novak, V., P. Novak, T. L. Opfer-Gehrking, and 85. Khurana, R. K., S. Watabiki, J. R. Hebel, R. Toro,
P. A. Low. 1996. Postural tachycardia syndrome: Time and E. Nelson. 1980. Cold face test in the assessment
frequency mapping. Journal of Autonomic Nervous of trigeminal-brain stem-vagal function in humans.
System 61:313–20. Annals of Neurology 7:144–9.
71. Novak, V., P. Novak, J. de Champlain, et al. 1993. 86. Schaller, B. 2004. Trigeminocardiac reflex: A clinical
Influence of respiration on heart rate and blood pres- phenomenon or a new physiological entity? Journal of
sure fluctuations. Journal of Applied Physiology 74: Neurology 251:658–65.
617–26. 87. Khurana, R. K., and R. Wu. 2006. The cold face
72. Ewing, D. J., I. W. Campbell, A. Murray, J. M. Neilson, test: A non-baroreflex mediated test of cardiac vagal
and B. F. Clarke. 1978. Immediate heart-rate response function. Clinical Autonomic Research 16:202–7.
to standing: Simple test for autonomic neuropathy in 88. Kingwell, B. A., G. A. McPherson, and P. I. Korner.
diabetes. British Medical Journal 1:145–7. 1991. Assessment of gain of tachycardia and brady-
73. Cybulski, G., and W. Niewiadomski. 2003. Influence of cardia responses of cardiac baroreflex. The American
age on the immediate heart rate response to the active Journal of Physiology 260:H1254–63.
orthostatic test. Journal of Physiology and Pharmacol- 89. Korner, P. I., M. J. West, J. Shaw, and J. B. Uther.
ogy 54:65–80. 1974. “Steady-state” properties of the baroreceptor-
74. Borst, C., W. Wieling, J. F. van Brederode, A. Hond, heart rate reflex in essential hypertension in man. Clin-
L. G. de Rijk, and A. J. Dunning. 1982. Mechanisms ical and Experimental Pharmacology & Physiology
of initial heart rate response to postural change. The 1:65–76.
American Journal of Physiology 243:H676–81. 90. Korner, P. I., J. Shaw, M. J. West, and J. R. Oliver.
75. Wieling, W., M. P. Harms, A. D. ten Harkel, et al. 1972. Central nervous system control of barorecep-
1996. Circulatory response evoked by a 3 s bout of tor reflexes in the rabbit. Circulation Research 31:
dynamic leg exercise in humans. Journal of Physiology 637–52.
494:601–11. 91. Ebert, T. J., J. J. Hayes, J. Ceschi, et al. 1984. Repeti-
76. Goodwin, G. M., D. I. McCloskey, and J. H. Mitchell. tive ramped neck suction: A quantitative test of human
1972. Cardiovascular and respiratory responses to baroreceptor function. The American Journal of Phys-
changes in central command during isometric exer- iology 247:H1013–17.
cise at constant muscle tension. Journal of Physiology 92. Kardos, A., H. Rau, M. W. Greenlee, et al. 1995.
226:173–90. Comparison of two mechanical carotid baroreceptor
77. Sanders, J. S., A. L. Mark, and D. W. Ferguson. 1989. stimulation techniques. Acta Physiologica Hungarica
Evidence for cholinergically mediated vasodilatation 83:121–33.
Chapter 40
Electrophysiology of Pain
Rose M. Dotson and Paola Sandroni
INTRODUCTION LASER EVOKED POTENTIALS

QUANTITATIVE SENSORY TEST CONTACT HEAT EVOKED
AUTONOMIC TESTS POTENTIALS
MICRONEUROGRAPHY SUMMARY
INTRODUCTION The tests are objective or semiobjective phys-

iologic measures that correlate with abnormal
The diagnosis and treatment of neuropathic function in the nociceptive system and docu-
pain are a challenge for physicians caring for ment a lesion that potentially could result in
patients who have medical conditions causing pain. They allow clinicians to assess neuro-
this type of pain. The incompletely understood pathic pain by quantifying responses to neu-
complex mechanisms of neuropathic pain con- rophysiologic tests and using the results to
tribute to the difficulty. Another major obstacle develop a better understanding of the underly-
has been finding a test to objectively demon- ing pathophysiologic mechanisms of the pain.
strate and quantify a physiologic disruption in This chapter reviews various neurophysi-
sensory function that may be the cause of the ologic techniques used to study the func-
pain. For many years, those involved in the tion and dysfunction of the nociceptive sys-
study of pain have attempted to quantify the tem in humans, including quantitative sensory
pain experience solely by bedside psychophys- tests (QSTs), autonomic tests, microneurogra-
ical evaluations and the subjective reports of phy (MCNG), laser evoked potentials (LEPs)
patients. Tests of large-fiber function, includ- and contact heat evoked potential stimulator
ing vibrometry, sensory nerve conduction stud- (CHEPS). The review outlines the physiologic
ies, and routine somatosensory evoked poten- basis of and methods for performing these
tials, have been used to document large myeli- techniques, some applications of the tests in
nated fiber sensory abnormalities that might patients with pain, and some of the potential
also involve small diameter axons of the noci- pitfalls in the use of these tests. Details of the
ceptive system. The first step was to develop techniques for performing autonomic tests are
neurophysiologic techniques to assess the func- given in Chapters 35–39.
tion of the peripheral and central compo-
nents of pain pathways in normal subjects. The Purpose and Role of Available Techniques
next step was to refine the techniques and to to Evaluate Pain Pathways
gather normal data. These techniques can be
applied to patients with altered sensory func- • QST is a neuropsychophysiologic tech-
tion, specifically those with neuropathic pain. nique: it complements the neurological
677
examination and neurophysiologic testing stimulation of a primary afferent can produce

by providing a quantification of the sub- sensitization. The method of limits exposes the
jective sensory experience. patient to a stimulus of changing intensity after
• Autonomic testing assesses the function of he or she has been instructed to signal the first
small fibers: although the small fibers con- perception of pain or temperature sensation,
veying pain are somatic, not autonomic, called the appearance threshold. Thresholds
they are often similarly affected by the may also be approached from above threshold
pathologic process, and autonomic func- level, with the patient reporting the cessation
tion can be used as a surrogate measure. of sensation, called the disappearance thresh-
• MCNG is a time-consuming, albeit accu- old. This algorithm is as sensitive and repro-
rate, direct way to assess single fiber func- ducible as other less time-efficient methods,
tion. It has no clinical role at this point, but for example, forced-choice testing, which may
is widely used in research. give lower absolute values but do not provide
• LEPs and CHEPS are the equivalent of more accurate data.2–4
somatosensory evoked potentials for small Magnitude estimation, an algorithm that
fibers. LEPs have limited clinical use. uses a visual analog scale to rate sensation
CHEPS are promising as a routine test, perceived when a suprathreshold stimulus is
but their role still needs to be proven. applied, is useful in evaluating pain.5–7 Dyck
et al.8 developed a method called the 4, 2, and
1 stepping algorithm that incorporates mag-
nitude estimation with the addition of null
QUANTITATIVE SENSORY TEST stimuli randomly interposed between pyrami-
dal and flat-topped pyramidal stimuli. The test-
The QST is useful for assessing myelinated and ing starts at an intermediate level of stimulus
unmyelinated sensory fibers in the periphery intensity, uses 25 steps to reach maximal inten-
and in nociceptive pathways of the central ner- sity, and uses the visual analogue scale that
vous system in persons with neurogenic pain.1 allows patients to rate the pain intensity. This
The test uses controlled thermal stimulation method of testing produces results that are
of the skin to allow examiners to reproducibly well correlated with the standard forced-choice
quantify sensory nerve function of nocicep- algorithm.8, 9
tive pathways. It relies on the subjective report The examiner can observe abnormal pat-
of the patient that he or she perceives the terns that correspond to pathophysiologic
stimulus. The QST quantifies sensory nerve changes in the sensory nervous system, central
function by allowing the examiner to deter- or peripheral. There are also clues to nonphys-
mine the intensity of stimulus required for iologic or psychogenic abnormal patterns with
perception and, at times, to correlate this with erratic results that are not reproducible.
the patient’s report of the type and intensity QST systems provide a temperature stim-
of sensation. The test must use one stimulus ulus by means of a contact Peltier-type ther-
type at a time, with precisely defined physi- mode, with a surface area of 3–13 cm,2 that is
cal characteristics and intensity. The apparatus applied to the skin to warm or cool the area
must have a program that allows the exam- under the thermode. A thermocouple is placed
iner to deliver stimuli to the skin of the patient at the skin–thermode interface to monitor the
using appropriate algorithms. This ensures temperature of the stimulus. The thermode is
that the test provides an accurate, consis- set initially at the adaptation or holding tem-
tent, and semiobjective assessment of sensory perature of 30◦ C–34◦ C, within the range of
function. temperature at which only a transient thermal
The most useful and efficient algorithm sensation is caused by placing the probe on the
for determining pain thresholds—that is, the skin.10 The examiner operates the apparatus
intensity at which at least 50% of the stimuli are to cause temperature changes within certain
perceived as painful—is the method of limits. limits, usually 0◦ C–50◦ C, that are set to pre-
In addition, this algorithm allows the exam- vent burning the patient’s skin. The rate of
iner to obtain reliable results with delivery temperature change should be standardized to
of the least number of potentially unpleasant ensure the validity of normal data. The findings
stimuli. This is important because repetitive from a study using psychophysical measures
Electrophysiology of Pain 679
and MCNG to record C nociceptor activity temperature ramp for gathering normal data
in response to noxious heat stimuli delivered and testing patients.
with a QST apparatus to the skin of normal The patient indicates when the first sensation
human volunteers are shown in Figure 40–1. occurs in response to cooling or warming the
This study showed that both the magnitude stimulus probe attached to the skin. The sensa-
of pain perceived by the subjects and the fre- tion (stimulus perception or pain) that occurs
quency of C nociceptor discharges in response in response to a cold or hot stimulus is indi-
to the stimulus increased with faster rates of cated by the patient, who either tells the exam-
temperature increase.11 Therefore, the examiner or presses a button to halt the stimulus
iner should conduct the QST using a consistent and to reset the thermode temperature to the
Figure 40–1. A–C, Discharge of a single

C nociceptor in response to heat stimulus
ramps between 32◦ C and 47◦ C, at three
rates of temperature rise. Note different
time bases for each of the three graphs. Dis-
charge rate recorded and peak magnitude
estimate of evoked pain are given. (From
Yarnitsky, D., D. A. Simone, R. M. Dotson,
M. A. Cline, and J. L. Ochoa. 1992. Single
C nociceptor responses and psychophysical
parameters of evoked pain: Effect of rate of
rise of heat stimuli in humans. The Journal of
Physiology 450:581–92. By permission of the
Physiological Society.)
holding level. The QST apparatus records the

stimulus temperature at which the patient indi-
cated perception of a change in temperature
or pain. The test apparatus includes standard-
ized thermode size, stimulation sites, rate of
change in stimulus temperature, interstimulus
intervals, and pretest skin temperature.
Tests are performed in normal volunteers
and in patients who have pain, to provide
standardized normal values and valid compar-
isons. In addition, comparison of the results
obtained on painful areas of skin with that
obtained from the same site on the uninvolved
side allows the examiner to determine whether
unilateral cold or heat hyperalgesia or allody-
nia is present. The interpretation of results
should consider that cold hypalgesia may occur
in normal subjects as well as in patients with
pain. Heat hypalgesia alone is rare and diffi-
cult to document because of the setting of an
upper temperature limit of 50◦ C to prevent Figure 40–2. Pure cold pain and/or heat pain hyperalge-
sia. A, Normal pattern; B, Pure cold pain hyperalgesia (30
injury. The normal ranges of pain thresholds patients); C, Pure heat hyperalgesia (26 patients); D, Com-
obtained using the method of limits are 44◦ C– bined cold pain and heat pain hyperalgesia (5 patients).
47◦ C for a hot stimulus and 9◦ C–12◦ C for a CP, cold pain; CS, cold sensation; HP, heat pain; WS,
cold stimulus.12 warm sensation. (From Verdugo, R., and J. L. Ochoa. 1992.
Quantitative somatosensory thermotest. A key method for
The QST evaluates the entire peripheral and functional evaluation of small calibre afferent channels.
central portions of the sensory system that par- Brain 115:893–913. By permission of Oxford University
ticipate in the transmission and perception of Press.)
painful hot or cold stimuli. Thermal stimula-
tion directly activates the receptor in the skin,
and the receptor in turn activates the axon
innervating it. In the peripheral nervous sys-
tem, the primary afferent fibers that convey
these messages to the central nervous system
are C and Aδ nociceptors. The central nocicep-
tive system consists of the spinothalamic tract,
cerebral cortex, antinociceptive areas of the
brain stem, and probably the hypothalamus,
amygdala, and limbic cortex.13
Decreased (hyperalgesia or allodynia) or
increased (hypalgesia) thresholds for the per-
ception of a cold or hot stimulus may occur in
various combinations, as demonstrated by Ver-
dugo and Ochoa12 in patients with somatosen- Figure 40–3. Thermal hypesthesia combined with ther-
sory disorders (Figs. 40–2 to 40–4). In some mal pain hyperalgesia. A, Normal pattern; B, Warm hypes-
patients with pain, the somatosensory abnor- thesia associated with cold pain hyperalgesia (6 patients);
C, Warm hypesthesia associated with heat pain hyper-
mality is hyperalgesia selectively in response algesia (1 patient); D, Cold hypesthesia associated with
to a cold or a hot stimulus. Ochoa and Yarnit- cold pain hyperalgesia (4 patients); E, Cold hypesthesia
sky14 described patients with neuropathy who associated with heat pain hyperalgesia (17 patients). CP,
had cold skin, reduced ability to perceive cool cold pain; CS, cold sensation; HP, heat pain; WS, warm
stimuli, and cold hyperalgesia. These patients sensation. (From Verdugo, R., and J. L. Ochoa. 1992.
Quantitative somatosensory thermotest. A key method for
had evidence of small myelinated Aδ-fiber functional evaluation of small calibre afferent channels.
neuropathy with sparing of unmyelinated C Brain 115:893–913. By permission of Oxford University
fibers. Press.)
actual area of injury in patients or in the area

where capsaicin is applied directly to the skin
of normal subjects.16 A study in human subjects
that used QST and MCNG to identify sen-
sitized C nociceptors in the peripheral nerve
innervating an area of skin where the patient
experienced spontaneous burning pain and
heat hyperalgesia showed that the neural dis-
charge of the abnormal, sensitized C nocicep-
tors correlated with the perceived magnitude
of pain as measured by the visual analog scale17
(Figs. 40–5 and 40–6). The QST is useful in
determining the pattern and degree of abnor-
mality, in following the clinical course, and in
documenting the response to treatment.15, 18
QST measurements can document and
quantify hypesthesia, hyperesthesia, hypalge-
sia, and hyperalgesia. Thus, the clinician has a
reproducible measurement of nerve dysfunc-
Figure 40–4. Thermal hypesthesia combined with ther- tion that can be attributed to a particular type
mal pain hyperalgesia (continuation of Fig. 40–3). A, Nor-
mal pattern; B, Cold hypesthesia associated with cold of primary afferent nerve fiber that may be
and heat pain hyperalgesia (5 patients); C, Cold and involved in pain production if the lesion is
warm hypesthesia associated with cold pain hyperalgesia in the periphery. However, the QST does not
(1 patient); D, Cold and warm hypesthesia associated with differentiate central from peripheral dysfunc-
cold and heat pain hyperalgesia (1 patient). CP, cold pain;
CS, cold sensation; HP, heat pain; WS, warm sensation.
tion, because the abnormalities found are not
(From Verdugo, R., and J. L. Ochoa. 1992. Quantita- specifically localized and may be located in the
tive somatosensory thermotest. A key method for func- nociceptive pathway at any level of the neu-
tional evaluation of small calibre afferent channels. Brain raxis. This test can be used to make the initial
115:893–913. By permission of Oxford University Press.) assessment, to follow the clinical course, and
to determine the response to medications and
other forms of intervention.19, 20
Many clinical reports indicate that patients
with complex regional pain syndrome (CRPS), Key Points
or reflex sympathetic dystrophy (RSD), fre-
quently report cold hyperalgesia on the QST or • QST is a neuropsychophysiologic test.
bedside examination. In fact, a comprehensive • All sensory fibers are tested; thus, it can
QST evaluation of patients with this clinical detect the degree of large vs. small fiber
diagnosis by means of the QST showed that dysfunction.
they may have cold hyperalgesia, heat hyper- • Patient’s cooperation is necessary for an
algesia, or both. These abnormalities can be accurate assessment: malingerers can usu-
quantified and followed clinically with the QST ally be detected by the inconsistencies,
to show response to treatment.12, 15 thanks to the paradigm used.
Heat hyperalgesia with spontaneous burning • Elevated thresholds (i.e., hypoesthesia) as
pain occurs in erythromelalgia or capsaicin- well as allodynia can be detected.
treated skin. The QST pattern in these
instances is one of decreased threshold for pain
to hot, but not to cold, stimuli. The spon-
taneous burning pain and mechanical hyper- AUTONOMIC TESTS
algesia in these patients can be lessened by
decreasing skin temperature, but the pain Patients with neurogenic pain and features
remains during compression-ischemia A-fiber indicating involvement of the sympathetic ner-
nerve block. This experimental finding indi- vous system, either sudomotor or vasomotor,
cates that C fibers mediate these sensory have a poorly understood pain symptom
changes of primary hyperalgesia within the complex known as complex regional pain
Maximum pain
Pain rating
No pain
2 seconds
Figure 40–5. Correlation between neural discharge of an identified sensitized C polymodal nociceptor with receptive
field in symptomatic skin (upper trace) and simultaneous temporal profile of pain magnitude (lower trace) in response to
gentle stroking of the receptive field. (From Cline, M. A., J. Ochoa, and H. E. Torebjörk. 1989. Chronic hyperalgesia and
skin warming caused by sensitized C nociceptors. Brain 112:621–47. By permission of Oxford University Press.)
syndrome (CRPS) or reflex sympathetic dystro- regarding the exact role of the sympathetic
phy (RSD).21, 22 Visual inspection and patients’ efferents in this context, and this may vary
reports of alterations in sweating and skin from patient to patient or even within the same
temperature of involved body areas implicate patient. These efferents may be the passive or
the sympathetic nervous system. The mul- the reactive arc of a somatosympathetic reflex
tiple pathophysiologic mechanisms that may response to noxious input. Also, denervated
result in this clinical pattern are not com- sympathetic end organs in case of nerve injury
pletely known. There are several possibilities may cause the clinical symptoms or signs of
autonomic nervous system involvement in neu-
ropathic pain. Roberts23 proposed that sympa-
thetic efferents normally have an active role in
helping to maintain low-threshold mechanore-
ceptor input to sensitized central nocicep-
tors. Another hypothesis is that nociceptor
activity is maintained by sympathetic effer-
ent activity, thereby causing spontaneous or
stimulus-induced sympathetically maintained
pain. Indirect evidence supports the idea that
there is sympathetic activation of nociceptors
in humans with the clinical features of CRPS.24
However, data from animal research indicate
that, in peripheral nerve injury, sympathetic
efferent activity causes excitation of cutaneous
nociceptors.25–28
The quantitative sudomotor axon reflex test
2 seconds (QSART), discussed in Chapter 36, is a sensi-
Figure 40–6. Response of an identified C polymodal unit
tive test with an approximately 20% coefficient
in symptomatic skin (upper trace) compared with that from of variation. Thus, this test gives reproducible
opposite nonpainful hand (lower trace). The mechanical results in normal subjects and in patients with
stimulus (stroking with a blunt wooden stick) was applied neuropathy. Furthermore, there is no signifi-
to the receptive fields, at a time marked by arrowheads. cant side-to-side difference in normal subjects.
Note the prolonged after-discharge for the C unit in symp-
tomatic skin. (From Cline, M. A., J. Ochoa, and H. E. This allows clinicians to compare the results in
Torebjörk. 1989. Chronic hyperalgesia and skin warming patients to normative values and to determine
caused by sensitized C nociceptors. Brain 112:621–47. By side-to-side differences in a patient with a
permission of Oxford University Press.) unilateral painful condition to obtain useful
information about sympathetic dysfunction or skin, such as warming another part of the body
sympathetic involvement in neuropathic pain or sympathetic blockade of the affected area,
states that fit the definition of CRPS.29–31 do not cause warming of that area of skin.
Patterns of QSART response that occur Patients with sensitized C nociceptors, ery-
in limbs affected by painful peripheral neu- thromelalgia, or topically applied capsaicin
ropathies and CRPS are excessive or persistent secrete vasodilating substances antidromically
sweat responses with reduced latencies or, in from active nociceptors, and as a result, the
some cases, decreased sweat volumes.29–31 An skin is warm in the areas with pain and hyper-
abnormal QSART pattern in a patient com- algesia.16, 35, 36 This can be reproduced in nor-
plaining of pain provides an objective mea- mal human volunteers by performing MCNG
sure indicating a pathophysiologic change in with intraneural microstimulation at intensities
the involved limb that may occur with, but that produce a painful sensation. Initially, the
is not necessarily a causative factor for, the pain may cause vasoconstriction that is readily
neuropathic pain. apparent on infrared thermography as cool-
Measurements of resting sweat output are ing of the skin. Continued activity in the pri-
helpful in conjunction with QSART to show mary nociceptors with microstimulation causes
sudomotor abnormalities in patients with the vasodilatation and warming of the skin that
clinical features of CRPS. These patients tend sympathetic reactivation can override to pro-
to have increased resting sweat output on duce cool skin. This may provide a pathophys-
the involved limb compared with the normal iologic explanation for the variability in the
side.30, 31 temperature of the painful area of skin com-
The sympathetic skin response (SSR) evalu- pared with that of the normal skin in patients
ates sympathetic sudomotor function through with neuropathic pain.37
somatosympathetic reflexes.32 Some reports
have documented SSR abnormalities in sudo- Key Points
motor function in patients with CRPS.33, 34
Because of inherent difficulties in producing • Autonomic studies evaluate cardiovagal,
quantifiable and reproducible data with the adrenergic, and postganglionic sudomotor
SSR technique, it is not a useful neurophysi- function: in doing so, they assess auto-
ologic tool for the assessment of neuropathic nomic, but not somatic, small-fiber func-
pain. tion.
Skin temperature measurements with a • The tests are noninvasive but require
surface thermistor or infrared thermography patient’s preparation and cooperation.
can compare multiple sites on the skin of • The battery is complementary to other
the involved extremity with the correspond- gastrointestinal, urologic, and cardiologic
ing areas on the asymptomatic extremity. assessments of autonomic functions.
Because patients with neuropathic pain or • A specific set of tests, focused on vaso-
CRPS may have alterations in skin tempera- motor and sudomotor functions, can be
ture in conjunction with sensory aberrations, useful in the diagnosis of CRPS, albeit not
these measurements permit examiners to doc- very specific.
ument clinically useful abnormal temperature • The tests are reproducible and can be
patterns or asymmetries caused by various used to monitor disease evolution and
pathophysiologic mechanisms.30 Patients with treatment response.
lesions causing deafferentation pain and vaso-
motor denervation may have relatively warm
skin on the involved side because of vasodi- MICRONEUROGRAPHY
latation. Later in the course of the condition,
denervation supersensitivity results in vasocon- In MCNG, semimicroelectrodes with a tip
striction caused by upregulation of adrenore- diameter of 1–15 μm are inserted percuta-
ceptors on blood vessels that begin to respond neously into an accessible peripheral nerve to
more vigorously to circulating catecholamines. record the activity in a single axon, in a portion
Thus, the skin on the involved side becomes of a fascicle, or in an entire nerve fascicle.38
cooler than that on the normal side. Maneu- MCNG is useful for uncovering the physio-
vers that usually result in reflex warming of the logic mechanisms of neuropathic pain.39 It is
a time-consuming test that requires a highly pain to occur with activation of low-threshold
motivated and observant patient for successful mechanoreceptors in the presence of central
acquisition of useful data. The electrode is con- nociceptor sensitization.
nected via a preamplifier to an amplifier with MCNG may be used to unravel the com-
attached audiomonitors and an oscilloscope to plex story of pain and the sympathetic nervous
permit the examiner to monitor the neural system in humans by directly recording sym-
activity of a peripheral nerve innervating an pathetic efferent activity. With MCNG, Casale
involved area of skin. The recording of skin and and Elam46 demonstrated normal activity in a
muscle sympathetic activity, Aβ low-threshold sympathetic efferent fiber in a nerve innervat-
mechanoreceptors, and Aδ and C nociceptor ing a painful area of skin of a patient with the
afferent activity can provide pathophysiologic clinical symptom complex of CRPS. Our obser-
information about the mechanisms of different vations in several such patients are consistent
types of neuropathic pain. with this finding.
As noted above, MCNG has documented
the occurrence of sensitized C nociceptors in a Key Points
patient with erythromelalgia-type pain.17 Tore-
bjork et al.40 used MCNG to provide evidence • MCNG monitors directly the level of
that an injury or the application of capsaicin to activity of single nerve fibers by intraneu-
the skin causes central sensitization in the area ral recording.
of secondary hyperalgesia outside the actual • Once identified, fibers can be tested
area of capsaicin injection or topical appli- by applying specific stimuli to induce a
cation. Ongoing nociceptive input appears to response (such as a sensory stimulation for
help maintain this sensitization.39 a somatic fiber or inducing hypotension if
With MCNG, investigators have identified recording from a sympathetic fiber).
three previously undescribed types of human • The technique is tedious and time-
C nociceptors that respond only to mechanical, consuming, and requires a very coopera-
heat, or chemical stimuli.41 Some of these units tive subject.
were sensitized to heating or mechanical stim- • Technical limitations have limited its use
uli after chemical stimulation with mustard oil, to research studies; it has no clinical appli-
capsaicin, or tonic pressure.42, 43 These likely cation at this time.
have a role in the primary hyperalgesia that
occurs with chemical irritation or inflammation
and in the secondary hyperalgesia caused by LASER EVOKED POTENTIALS
central sensitization.
Animal experiments have shown that sym- LEPs provide a noninvasive, easily tolerated
pathetic activation of primary afferents occurs means of directly assessing function of the
with direct stimulation of sympathetic nerves.23 central and peripheral portions of the nocicep-
This was not found in MCNG studies of tive system.47 Carmon et al.48 first showed that
human subjects and patients with the clini- stimulation of normal human skin with short-
cal features of CRPS in whom reflex activa- duration infrared CO2 laser pulses produced
tion of sympathetic efferents did not activate a near-field cerebral potential at the vertex.
low-threshold mechanoreceptors.44 In animals, Amplitudes of the cerebral response usually
the activity of sympathetic efferents results in correlate well with the intensity of perceived
neural activity in low-threshold mechanore- pain reported by patients in response to the
ceptors in even the normal state; sympathetic stimulus and with the intensity of the applied
efferents have a similar effect on nociceptors stimulus.49 Wu et al.50 recently reported on two
only after nerve injury.26–28 Although patients patients with hyperalgesia (caused by central
with CRPS symptoms have allodynia on neu- pain in one and peripheral neuropathic pain in
rologic examination, activity in single isolated the other) in whom the LEP responses were
low-threshold mechanoreceptors produced by delayed, desynchronized, and attenuated.
intraneural microstimulation at frequencies up Heat pain–producing lasers, as opposed to
to 30 Hz did not cause pain.45 This suggests transcutaneous electrical stimulation of periph-
that temporal and spatial summation may be eral nerves traditionally used for somatosen-
necessary for spontaneous or stimulus-induced sory evoked potentials, can induce pain
with minimal influence on other sensations skin surface temperature (50◦ C per second)
(Fig. 40–7). Only minimal habituation, adapta- and selectively activates the smallest diame-
tion, or tissue damage tends to occur even with ter nerve terminals of thinly myelinated Aδ
repeated applications of the laser stimulus. and unmyelinated C fibers.46, 50, 51 Laser stim-
Although some laboratories use intracutaneous ulus intensity is best characterized as stim-
electric shock to obtain pain somatosensory ulus energy per unit area, and the average
evoked potentials, most laboratories perform pain threshold in young healthy adults is
LEPs. The laser does not contact the skin 10 mJ/mm2 .55
directly as it produces an invisible, inaudi- LEPs have larger amplitudes than routine
ble, short-duration (20 ms) radiant heat pulse. somatosensory evoked potentials and require
The very superficial layers of skin (20–50 mm) averaging of only 25–40 responses (Fig. 40–8).
are able to absorb this pulse because of its The components of LEPs include late and
long wavelength (CO2 laser, 10.6 μm; and ultralate waveforms with a maximal amplitude
thulium YAG laser, 1.8–2.01 μm).48, 51–54 This over the vertex at Cz (according to the Inter-
type of stimulus produces a rapid increase in national 10–20 System). Laser activation of
Brain potential
N150
Reaction
5 μV
100 ms
P250
Pain rating
0 4 10
1 mm
Epidermis
Dermis
Subcutis
Stimulus
Figure 40–7. The measurement of pain-related cerebral potentials. Pain-inducing stimuli, such as intracutaneous shock
(lower right) or laser heat pulse, specifically activate nociceptive afferents, which conduct information in anterolateral and
dorsal spinal tracts to the thalamus and, from there, to the cerebral cortex. Sensation is estimated on an analogue scale,
with values of 4 and more denoting increasing pain. Stimulus-induced brain potentials appear in the surface EEG and
are visible after averaging more than 40 stimulus repetitions. The negativity (upward deflection) at 150 ms (N150) after
stimulus onset and the positivity at 250 ms (P250) are late components of the evoked potential that reflect the painfulness
of the stimulus applied. (From Bromm, B. 1995. Consciousness, pain, and cortical activity. Advances in Pain Research and
Therapy 22:35–59. By permission of Lippincott Williams & Wilkins.)
–
10 μV
+
–1 0 1 2 3
t (s)
Figure 40–8. Late and ultralate LEPs in a healthy subject. Vertex vs. linked earlobes (negativity upward). Stimulation of
the back of the hand elicited a late positivity at about 400 ms (A). Preferential A-fiber block by pressure to the radial nerve
at the wrist strongly attenuated the late LEP and an ultralate potential appeared (B), indicating that the latter was mediated
by preserved C-fiber input. (From Treede, R.-D., J. Lorenz, K. Kunze, and B. Bromm. 1995. Assessment of nociceptive
pathways with laser-evoked potentials in normal subjects and patients. Advances in Pain Research Therapy 22:377–92. By
permission of Lippincott Williams & Wilkins.)
Aδ fibers that have conduction velocities of P2 component. Contralateral primary and sec-
4–30 m/second in humans38 causes first pain, ondary somatosensory cortex activity appears
with a latency of approximately 500 ms corre- to be the generator of the middle latency (N1)
sponding with the late LEP. The typical wave- component.
form obtained with stimulation of the skin on LEPs are useful clinically to evaluate objec-
the dorsum of the hand has middle latency neg- tively the peripheral and central nociceptive
ative peaks (N1, N170), a negative peak (N2) pathways in patients with neuropathic pain
at a latency of 250 ± 20 ms (mean ± SD), and disturbances of pain perception, such as
and a positive peak (P2) at 390 ± 30 ms. N2 hypalgesia, hyperalgesia, allodynia, and spon-
is maximal at Cz, with extension into the cen- taneous pain.58–60 Some of these patients have
tral leads, but P2 is maximal at Cz and Pz. abnormal summation, or wind-up, consisting
The level of attention, arousal, and distraction of the perception of continuous burning pain
influences these potentials, especially Pz, and instead of the normal individual sharp-pricking
these factors must be taken into account when painful sensations when a repetitive 1-Hz pin-
performing the test.48, 53, 56 prick stimulus is applied to the skin. LEPs
Activation of C fibers (conduction veloc- indicated the involvement of Aβ and Aδ fibers
ity, 0.4–1.8 m/second in humans) results in the in a patient with polyneuropathy, muscle weak-
ultralate components of LEPs. This response ness, impaired sensation (cold, position, and
has a positive peak maximal at the vertex and vibratory), absence of conventional tibial nerve
a latency of about 1400 ms. It is unreliable somatosensory evoked potentials, and large
in recordings unless preferential A-fiber block myelinated fiber loss on sural nerve biopsy.53 In
suppresses the late component.53 The ultralate this patient, LEPs showed small late responses,
wave is easily obtained if the late component is evidence for impaired Aδ function, and large
absent because of disease selectively affecting ultralate responses, with a peak latency of
Aδ and not C fibers. approximately 1600 ms, which is evidence for
Scalp topography and waveforms of the preserved function of C fibers (Fig. 40–9).
late and ultralate LEPs are similar, suggest- In a case of polyneuropathy in which the
ing that they have the same cerebral gener- nerve conduction distance between the hand
ators.57 Spatiotemporal source analysis likely and the foot was 0.8 m, the late and ultralate
indicates that N2 is generated by activity LEP responses corresponded to conduction
mainly bilaterally in secondary somatosensory velocities of 16 m/second and 1.2 m/second,
cortex. A deep dipole in the midline corre- respectively53 (Fig. 40–10). This study con-
sponding to the location of the anterior cin- firmed that Aδ peripheral afferents are respon-
gulate gyrus is primarily responsible for the sible for transmission of the late component
Hand
Conventional average
Latency corrected average
Conventional average
Foot
– Latency corrected average
10 μV
+
0 1000 2000
t (ms)
Subject H.G., 67 years. polyneuropathy. laser EP
Figure 40–9. Late and ultralate LEPs in a 67-year-old man with polyneuropathy. Top traces: Following stimulation of
the right hand, a normal Aδ-fiber-related late potential was recorded. Bottom traces: Following stimulation of the left foot,
the late potential was markedly decreased in amplitude and a C-fiber-related ultralate potential was documented. The
heat-pain threshold for laser stimuli was unremarkable in both areas, but a pronounced temporal summation occurred
with stimulation of the foot. (From Treede, R.-D., J. Lorenz, K. Kunze, and B. Bromm. 1995. Assessment of nociceptive
pathways with laser-evoked potentials in normal subjects and patients. Advances in Pain Research Therapy 22:377–92. By
permission of Lippincott Williams & Wilkins.)
and C-fiber activation for the ultralate com- unmasking appears to provide a cortical cor-
ponent. This patient had marked wind-up, relate for disinhibition of C-fiber responses to
despite hypalgesia in response to a single pin- noxious heat that occurs in persons who dis-
prick stimulus, and there was unmasking of play wind-up when A fibers are impaired.53
the ultralate component of the LEP. This Therefore, this technique can be useful in the
Hand
Foot
5 μV 50 ms
= 16 m/s 670 ms
= 1.2 m/s
+
0 500 1000 1500 2000 2500
t (ms)
Figure 40–10. Late and ultralate LEPs in a 25-year-old man with hereditary motor and sensory neuropathy type I. The
latency differences between hand and foot stimulation indicate that late LEPs are mediated by Aδ fibers and ultralate
LEPs by C fibers. (From Treede, R.-D., J. Lorenz, K. Kunze, and B. Bromm. 1995. Assessment of nociceptive pathways
with laser-evoked potentials in normal subjects and patients. Advances in Pain Research Therapy 22:377–92. By permission
of Lippincott Williams & Wilkins.)
evaluation of nociceptive pathways in general. two types of sensations: a first sharp pain fol-
Also, it can help document Aδ-fiber impair- lowed by a second, duller type pain. The cal-
ment with sparing of C-fiber function. The culated conduction velocities for the first peak
selective loss of small unmyelinated fibers (N550) would fit Aδ fibers (10 m/second) and
can only be documented when A fibers are can be generated at temperatures of 45 C. The
blocked or impaired, because activity in C later components (around 1000 ms) would be
fibers produces the highly variable ultralate the result of C-fiber activation (velocities esti-
LEP response.53 mated at 2–3 m/second) and are visible only
with higher peak temperatures around 52 C.
Reproducibility and reliability of CHEPS are
Key Points comparable to those of LEPs.
• LEPs have been the first technique Varying the attentional target toward differ-
described to test pain pathways, trying ent properties of the stimulus did not cause any
to emulate what somatosensory evoked significant change in CHEPS response ampli-
potentials are for large fiber sensory tudes and latencies, suggesting that CHEPS
modalities. represent a reliable functional measure of the
• Albeit promising, the potential risk of nociceptive pathways.62
superficial burns and the intrinsic diffi- CHEPS amplitudes correlate negatively
culties of the technique have made it with age. Amplitudes and latencies of CHEPS
unsuitable for clinical use. correlate with verbal pain scores: the higher
the rating, the shorter the N1 latency and the
higher the N1–P1 amplitude.63
In a study performed on patients with
CONTACT HEAT EVOKED symptoms of sensory neuropathy compared
POTENTIALS to controls, CHEPS were compared to other
methods used to evaluate small fibers—
LEPs have not achieved wide use in clinical the histamine-induced skin flare response,
practice due to intrinsic technical difficulties intraepidermal fibers (IEF) count, and quanti-
(requiring skilled personnel to calibrate and tative sensory testing. Amplitudes of Aδ evoked
operate the expensive equipment) and pitfalls, potentials were reduced in patients, who also
which include the fact that they are not a nat- showed reduced leg skin flare responses and
ural stimulus and can cause skin burns and reduced IEF compared to controls. The reduc-
hyperpigmentation. Contact heat is a natural tion in flare response and fiber count corre-
stimulus, but previously could not be used as lated with the potential amplitude. The authors
a suitable stimulus for evoked potentials due concluded CHEPS provide a clinically practi-
to its slow rise time. In recent years, heat- cal, noninvasive, and objective measure, and
foil technology has been developed that can can be a useful additional tool for the assess-
elicit CHEPS as it has a rapid rising time of ment of sensory small-fiber neuropathy.64 More
50◦ C/second.61 Peak temperature is reached recently, two children with congenital insen-
360 ms after the offset of the 300-ms duration sitivity to pain were tested: they had normal
stimulus. The stimuli are delivered at random flare response but absent cortical heat evoked
intervals, mean = 10 seconds. Forty trials are potentials, suggesting an abnormality in more
averaged and recordings are performed as per proximal or central pain pathways.65
routine somatosensory evoked potential. The In a recent study to evaluate trigeminal
largest activation occurs at the vertex area. To small-fiber function, Truini et al. found con-
generate well-formed waveforms, the stimulus tact heat stimuli at 51◦ C evoked vertex poten-
intensity has to be able to induce at least mod- tials consisting of an NP complex similar to
erate pain in the subjects, with a VAS rating that elicited by laser pulses, though with a
of 6 or higher (Fig. 40–11). This is generally latency some 100 ms longer. Perioral stimu-
achieved with peak temperatures in the range lation yielded higher pain intensity ratings,
of 50◦ C. Four peaks become visible then: shorter latency, and larger amplitude CHEPS
N450 (at T3), N550 (at Cz), P750 (at Cz), and than supraorbital stimulation. Contact heat
P1000 (at Pz). This stimulus appears to activate stimuli at 53◦ C evoked a blink-like response in
both Aδ and C fibers; it sometimes can induce the relaxed orbicularis oculi muscle and a silent
–10 Warm, nonpainful (l-2)
Amplitude (μV)
–5
–10 Slight pain (l-4)

Amplitude (μV)
–5
–15 Moderate pain (l-6)
–10 Cz/N550
Amplitude (μV)
–5
0
5
10
Cz/P1000
0
Cz/P750
Figure 40–11. Contact heat evoked potentials. Vertex waveforms in relation to stimulus heat energy levels (I-2, I-4, I-6),
resulting in increasing pain sensation on verbal rating score. Amplitude increase correlated to increase of energy levels, and
the results also demonstrated the consistency and reproducibility between Study-I (dark) and Study-II (grey). Prominent
vertex components are marked: Cz/N550, Cz/P750, and Cz/P1000 at I-6 (moderate pain level). (From Chen, A. C. N.,
D. M. Niddam, and L. Arendt-Nielsen. 2001. Contact heat evoked potentials as a valid mean to study nociceptive pathways
in human subjects. Neuroscience Letters 316:79–82. By permission of Elsevier.)
period in the contracted masseter muscle. age and the initial negative latency correlates
In patients with facial neuropathic pain, the with gender, with shorter latencies noted in
CHEPS abnormalities paralleled those seen females. Pain intensity per the verbal rating
with LEPs.66 scale (VRS) was decreased with aging and
A dipolar model explaining the scalp higher in females than males. Higher VRS
CHEPS distribution is very similar to that pre- responses correlated positively with higher
viously described to explain the topography of CHEPS amplitude and a shorter latency of the
evoked potentials to radiant heat stimulation by first negative peak. Thus, age-related changes
laser pulses.67 Since laser stimuli activate the in thermal pain perception and CHEPS should
nociceptive fibers, the strong similarity of the be considered when using this modality of
cerebral dipoles activated by contact heat stim- testing somatosensory function.63
uli and by laser pulses suggests that only noci-
ceptive inputs are involved in the scalp painful
CHEPS building. Therefore, CHEPS record-
Key Points
ing can reliably assess nociceptive pathways,
similarly to LEPs. • CHEPS may be the new generation equiv-
Recent evidence shows that CHEPS evoked alent of LEPs, easier to use with virtually
response amplitudes correlate negatively with no risks.
• Preliminary data show they could be a • Neurophysiologic studies can evaluate the
valuable objective tool to assess Aδ- and integrity and/or dysfunction of the neu-
C-fiber pain pathways. roanatomic substrate of pain, and may
• As some degree of pain has to be induced indirectly provide insight into the patho-
to generate an adequate cortical potential, physiology of pain.
patient’s cooperation is a must.
SUMMARY REFERENCES
Pain is a subjective experience in which the 1. Greenspan, J. D. 2001. Quantitative assessment of
patient’s emotional state has a major role, con- neuropathic pain. Current Pain and Headache Reports
tributing to the challenge pain clinicians have 5:107–13.
2. Claus, D., M. J. Hilz, and B. Neundorfer. 1990.
in quantifying and objectively evaluating this Thermal discrimination thresholds: A comparison of
common complaint. The multiplicity and com- different methods. Acta Neurologica Scandinavica
plexity of the neural mechanisms that produce 81:533–40.
chronic pain make the clinician’s task even 3. Dyck, P. J., J. L. Karnes, D. A. Gillen, P. C. O’Brien,
more challenging. During the last 20 years, I. R. Zimmerman, and D. M. Johnson. 1990. Com-
parison of algorithms of testing for use in automated
research in the assessment of the nociceptive evaluation of sensation. Neurology 40:1607–13.
and autonomic systems has provided useful 4. Peripheral Neuropathy Association. 1993. Quan-
tools for the diagnosis, treatment, and scien- titative sensory testing: A consensus report from
tific investigation of neuropathic pain. QSTs the Peripheral Neuropathy Association. Neurology
43:1050–2.
provide an accurate, reproducible assessment 5. Price, D. D., P. A. McGrath, A. Rafii, and B. Buck-
of the sensory response to well-delineated ingham. 1983. The validation of visual analogue scales
controlled stimuli for evaluating the function as ratio scale measures for chronic and experimental
of small myelinated and unmyelinated fibers. pain. Pain 17:45–56.
These tests can be useful for documenting sen- 6. Price, D. D. 1988. Psychological and neural mecha-
nisms of pain, 18–49. New York: Raven Press.
sory disturbances that may occur in patients 7. Gruener, G., and P. J. Dyck. 1994. Quantitative sensory
with neuropathy who experience hypalgesia, testing: Methodology, applications, and future direc-
hyperalgesia, or allodynia. The inclusion of tions. Journal of Clinical Neurophysiology 11:568–83.
autonomic tests in the evaluation of some 8. Dyck, P. J., P. C. O’Brien, J. L. Kosanke, D. A. Gillen,
and J. L. Karnes. 1993. A 4, 2, and 1 stepping algo-
patients with neuropathic pain may provide rithm for quick and accurate estimation of cutaneous
objective evidence of increased sympathetic sensation threshold. Neurology 43:1508–12.
tone, heightened somatosympathetic reflexes, 9. Dyck, P. J., I. Zimmerman, D. A. Gillen, D. Johnson,
or sympathetic denervation. This information J. L. Karnes, and P. C. O’Brien. 1993. Cool, warm, and
may help the clinician to diagnose more accu- heat-pain detection thresholds: Testing methods and
inferences about anatomic distribution of receptors.
rately the particular type of pain syndrome Neurology 43:1500–8.
the patient has so that the treatment may 10. Darian-Smith, I. 1984. Thermal sensibility. In Hand-
be better directed at the underlying mech- book of physiology: A critical, comprehensive pre-
anism. MCNG, primarily a research tool, is sentation of physiological knowledge and concepts,
Vol. 3, The nervous system, ed. J. M. Brookhart,
a powerful method for directly studying pri- and V. B. Mountcastle, 879–913. Bethesda: American
mary afferent and sympathetic efferent neural Physiological Society.
activity in patients with pain caused by lesions 11. Yarnitsky, D., D. A. Simone, R. M. Dotson, M. A.
of the nervous system. LEPs allow clinicians Cline, and J. L. Ochoa. 1992. Single C nociceptor
and researchers to investigate objectively func- responses and psychophysical parameters of evoked
pain: Effect of rate of rise of heat stimuli in humans.
tion of the peripheral and central nociceptive The Journal of Physiology 450:581–92.
pathways in patients with neuropathic pain. 12. Verdugo, R., and J. L. Ochoa. 1992. Quantitative
somatosensory thermotest. A key method for func-
Key Points tional evaluation of small calibre afferent channels.
Brain 115:893–913.
• Pain is a subjective experience. Only the 13. Willis, W. D., Jr. 1995. From nociceptor to corti-
cal activity. Advances in Pain Research and Therapy
subject suffering from it can quantify it. 22:1–19.
• Neuropsychophysiologic studies can aid 14. Ochoa, J. L., and D. Yarnitsky. 1994. The triple
the quantification of pain experience. cold syndrome. Cold hyperalgesia, cold hypoaesthesia
and cold skin in peripheral nerve disease. Brain study and laboratory evaluation. The Clinical Journal
117:185–97. of Pain 14:282–9.
15. Wahren, L. K., and E. Torebjork. 1992. Quantitative 32. Shahani, B. T., J. J. Halperin, P. Boulu, and J. Cohen.
sensory tests in patients with neuralgia 11–25 years 1984. Sympathetic skin response—A method of assess-
after injury. Pain 48:237–44. ing unmyelinated axon dysfunction in peripheral neu-
16. Culp, W. J., J. Ochoa, M. Cline, and R. Dotson. 1989. ropathies. Journal of Neurology, Neurosurgery, and
Heat and mechanical hyperalgesia induced by cap- Psychiatry 47:536–42.
saicin. Cross modality threshold modulation in human 33. Cronin, K. D., R. L. Kirsner, and V. P. Fitzroy. 1982.
C nociceptors. Brain 112:1317–31. Diagnosis of reflex sympathetic dysfunction. Use of the
17. Cline, M. A., J. Ochoa, and H. E. Torebjörk. 1989. skin potential response. Anaesthesia 37:848–52.
Chronic hyperalgesia and skin warming caused by 34. Rommel, O., M. Tegenthoff, U. Pern, M. Strumpf,
sensitized C nociceptors. Brain 112:621–47. M. Zenz, and J. P. Malin. 1995. Sympathetic skin
18. Wahren, L. K., E. Torebjork, and B. Nystrom. 1991. response in patients with reflex sympathetic dystrophy.
Quantitative sensory testing before and after regional Clinical Autonomic Research 5:205–10.
guanethidine block in patients with neuralgia in the 35. Ochoa, J. 1986. The newly recognized painful
hand. Pain 46:23–30. ABC syndrome: Thermographic aspects. Thermology
19. Nygaard, O. P., R. Kloster, T. Solberg, and S. I. Mell- 2:65–107.
gren. 2000. Recovery of function in adjacent nerve 36. Ochoa, J., M. Cline, W. Comstock, et al. 1987. Painful
roots after surgery for lumbar disc herniation: Use of syndrome newly recognized: Polymodal hyperalge-
quantitative sensory testing in the exploration of dif- sia with cross modality threshold modulation and
ferent populations of nerve fibers. Journal of Spinal rubor (abstract). Abstracts—Society for Neuroscience
Disorders 13:427–31. 13:189.
20. Attal, N., V. Gaude, L. Brasseur, et al. 2000. Intra- 37. Ochoa, J. L., D. Yarnitsky, P. Marchettini, R. Dot-
venous lidocaine in central pain: A double-blind, son, and M. Cline. 1993. Interactions between sym-
placebo-controlled, psychophysical study. Neurology pathetic vasoconstrictor outflow and C nociceptor-
54:564–74. induced antidromic vasodilatation. Pain 54:191–6.
21. Dotson, R. M. 1993. Causalgia—reflex sympathetic 38. Vallbo, A. B., K. E. Hagbarth, H. E. Torebjork, and
dystrophy—sympathetically maintained pain: Myth B. G. Wallin. 1979. Somatosensory, proprioceptive,
and reality. Muscle & Nerve 16:1049–55. and sympathetic activity in human peripheral nerves.
22. Stanton-Hicks, M., W. Janig, S. Hassenbusch, J. D. Physiological Reviews 59:919–57.
Haddox, R. Boas, and P. Wilson. 1995. Reflex sym- 39. Torebjork, E. 1993. Human microneurography and
pathetic dystrophy: Changing concepts and taxonomy. intraneural microstimulation in the study of neuro-
Pain 63:127–33. pathic pain. Muscle & Nerve 16:1063–5.
23. Roberts, W. J. 1986. A hypothesis on the physiolog- 40. Torebjork, H. E., L. E. Lundberg, and R. H. LaMotte.
ical basis for causalgia and related pains. Pain 24: 1992. Central changes in processing of mechanorecep-
297–311. tive input in capsaicin-induced secondary hyperalgesia
24. Campbell, J. N., R. A. Meyer, and S. N. Raja. 1992. Is in humans. The Journal of Physiology 448:765–80.
nociceptor activation by alpha-1 adrenoreceptors the 41. Schmidt, R., M. Schmelz, C. Forster, M. Ringkamp,
culprit in sympathetically maintained pain? American E. Torebjork, and H. Handwerker. 1995. Novel classes
Pain Society Journal 1:3–11. of responsive and unresponsive C nociceptors in
25. Levine, J. D., Y. O. Taiwo, S. D. Collins, and J. K. Tam. human skin. Journal of Neuroscience 15:333–41.
1986. Noradrenaline hyperalgesia is mediated through 42. Schmidt, R., M. Schmelz, H. E. Torebjork, and H. O.
interaction with sympathetic postganglionic neurone Handwerker. 2000. Mechano-insensitive nociceptors
terminals rather than activation of primary afferent encode pain evoked by tonic pressure to human skin.
nociceptors. Nature 323:158–60. Neuroscience 98:793–800.
26. Sato, J., and E. R. Perl. 1991. Adrenergic excitation of 43. Schmelz, M., R. Schmid, H. O. Handwerker, and
cutaneous pain receptors induced by peripheral nerve H. E. Torebjork. 2000. Encoding of burning pain
injury. Science 251:1608–10. from capsaicin-treated human skin in two categories
27. Sato, J., S. Suzuki, T. Iseki, and T. Kumazawa. of unmyelinated nerve fibres. Brain 123:560–71.
1993. Adrenergic excitation of cutaneous nocicep- 44. Dotson, R., J. Ochoa, M. Cline, W. Roberts, D. Yarnit-
tors in chronically inflamed rats. Neuroscience Letters sky, and D. Simone. 1990. Sympathetic effects on
164:225–8. human low threshold mechanoreceptors (abstract).
28. Gold, M. S., D. M. White, S. C. Ahlgren, M. Guo, and Abstracts—Society for Neuroscience 16:1280.
J. D. Levine. 1994. Catecholamine-induced mechan- 45. Dotson, R., J. Ochoa, M. Cline, P. Marchettini,
ical sensitization of cutaneous nociceptors in the rat. and D. Yarnitsky. 1992. Intraneural microstimulation
Neuroscience Letters 175:166–70. of low threshold mechanoreceptors in patients with
29. Low, P. A. 1993. Autonomic nervous system function. causalgia/RSD/SMP (abstract). Abstracts—Society of
Journal of Clinical Neurophysiology 10:14–27. Neuroscience 18:290.
30. Chelimsky, T. C., P. A. Low, J. M. Naessens, P. R. Wil- 46. Casale, R., and M. Elam. 1992. Normal sympathetic
son, P. C. Amadio, and P. C. O’Brien. 1995. Value nerve activity in a reflex sympathetic dystrophy with
of autonomic testing in reflex sympathetic dystrophy. marked skin vasoconstriction. Journal of the Auto-
Mayo Clinic Proceedings 70:1029–40. nomic Nervous System 41:215–19.
31. Sandroni, P., P. A. Low, T. Ferrer, T. L. Opfer- 47. Rossi, P., M. Serrao, G. Amabile, L. Parisi, F. Pierelli,
Gehrking, C. L. Willner, and P. R. Wilson. 1998. Com- and G. Pozzessere. 2000. A simple method for
plex regional pain syndrome I (CRPS I): Prospective estimating conduction velocity of the spinothalamic
tract in healthy humans. Clinical Neurophysiology 58. Antonini, G., F. Gragnani, A. Romaniello, et al. 2000.
111:1907–15. Sensory involvement in spinal-bulbar muscular atro-
48. Carmon, A., J. Mor, and J. Goldberg. 1976. Evoked phy (Kennedy’s disease). Muscle & Nerve 23:252–8.
cerebral responses to noxious thermal stimuli in 59. Agostino, R., G. Cruccu, A. Romaniello, P. Innocenti,
humans. Experimental Brain Research 25:103–7. M. Inghilleri, and M. Manfredi. 2000. Dysfunction
49. Beydoun, A., T. J. Morrow, J. F. Shen, and K. L. of small myelinated afferents in diabetic polyneuropa-
Casey. 1993. Variability of laser-evoked potentials: thy, as assessed by laser evoked potentials. Clinical
Attention, arousal and lateralized differences. Elec- Neurophysiology 111:270–6.
troencephalography and Clinical Neurophysiology 88: 60. Agostino, R., G. Cruccu, G. D. Iannetti, et al. 2000.
173–81. Trigeminal small-fibre dysfunction in patients with
50. Wu, Q., L. Garcia-Larrea, P. Mertens, A. Beschet, diabetes mellitus: A study with laser evoked poten-
M. Sindou, and F. Mauguiere. 1999. Hyperalgesia with tials and corneal reflex. Clinical Neurophysiology
reduced laser evoked potentials in neuropathic pain. 111:2264–7.
Pain 80:209–14. 61. Chen, A. C. N., D. M. Niddam, and L. Arendt-
51. Bromm, B. 1995. Consciousness, pain, and corti- Nielsen. 2001. Contact heat evoked potentials as a
cal activity. Advances in Pain Research and Therapy valid mean to study nociceptive pathways in human
22:35–59. subjects. Neuroscience Letters 316:79–82.
52. Kazarians, H., E. Scharein, and B. Bromm. 1995. 62. Le Pera, D., M. Valeriani, D. Niddam, A. C. Chen, and
Laser evoked brain potentials in response to painful L. Arendt-Nielsen. 2002. Contact heat evoked poten-
trigeminal nerve activation. The International Journal tials to painful and non-painful stimuli: Effect of atten-
of Neuroscience 81:111–22. tion towards stimulus properties. Brain Topography
53. Treede, R.-D., J. Lorenz, K. Kunze, and B. Bromm. 15:115–23.
1995. Assessment of nociceptive pathways with 63. Chao, C. C., S. T. Hsieh, M. J. Chiu, M. T. Tseng,
laser-evoked potentials in normal subjects and and Y. C. Chang. 2007. Effects of aging on contact
patients. Advances in Pain Research Therapy 22: heat-evoked potentials: The physiological assessment
377–92. of thermal perception. Muscle & Nerve 36(1):30–8.
54. Spiegel, J., C. Hansen, and R. D. Treede. 2000. Clini- 64. Atherton, D. D., P. Facer, K. M. Robertset al. 2007.
cal evaluation criteria for the assessment of impaired Use of the novel contact heat evoked potential stim-
pain sensitivity by thulium-laser evoked potentials. ulator (CHEPS) for the assessment of small fibre
Clinical Neurophysiology 111:725–35. neuropathy: Correlations with skin flare responses and
55. Gibson, S. J., S. A. LeVasseur, and R. D. Helme. 1991. intra-epidermal nerve fibre counts. BMC Neurology
Cerebral event-related responses induced by CO2 7(1):21.
laser stimulation in subjects suffering from cervico- 65. Facer, P., D. D. Atherton, K. M. Roberts, et al. 2007.
brachial syndrome. Pain 47:173–82. A new syndrome of congenital insensitivity to pain
56. Tarkka, I. M., and R. D. Treede. 1993. Equiv- diagnosed by skin biopsy and contact heat evoked
alent electrical source analysis of pain-related potentials. PNS Meeting.
somatosensory evoked potentials elicited by a CO2 66. Truini, A., F. Galeotti, E. Pennisi, F. Casa, A. Bia-
laser. Journal of Clinical Neurophysiology 10: siotta, and G. Cruccu. 2007. Trigeminal small-fibre
513–19. function assessed with contact heat evoked potentials
57. Treede, R. D., and B. Bromm. 1988. Reliability and in humans. Pain 132(1–2):102–7.
validity of ultra-late cerebral potentials in response 67. Valeriani, M., D. Le Pera, D. Niddam, A. C. Chen,
to C-fibre activation in man. In Proceedings of and L. Arendt-Nielsen. 2002. Dipolar modelling of
the Vth World Congress on Pain, ed. R. Dubner, the scalp evoked potentials to painful contact heat
G. F. Gebhart, and M. R. Bond, 567–73. Amsterdam: stimulation of the human skin. Neuroscience Letters.
Elsevier. 318:44–8.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART G
Sleep and Consciousness
The steadily increasing recognition of the fre- Electrophysiologic assessment of sleep

quency and severity of sleep disorders has disorders is a superb example of the import-
markedly increased the demand for physio- ance of combining the methods of
logic assessment of sleep. The development clinical neurophysiology for assessing the
of electrophysiologic tools for the assessment condition of a patient. The combination
of sleep has helped define the normal and of newly defined patterns of surface elec-
altered physiology of sleep. Sleep disorders tromyographic recordings, electroencephalo-
include inadequate, excessive, and disordered graphic recordings, and measurements of
sleep. The latter includes excessive or abnor- autonomic function is a critical part of
mal movements during sleep. The recording the assessment of sleep disorders. New
of the surface muscle electromyogram during advances in sleep studies include unat-
these movements while monitoring blood pres- tended polysomnography, split-night record-
sure, pulse, and respiration, as described in this ings, excess daytime sleepiness evaluation,
section, can help physicians identify, character- and the interpretation of electro-oculography
ize, and define the type and severity of sleep and multiple sleep latency testing in the
disorder (Chapter 41). parasomnias.
695
Chapter 41
Assessment of Sleep and Sleep

Disorders
Michael H. Silber, Cameron D. Harris, and Peter J. Hauri
INTRODUCTION Blood Gases

TECHNIQUES USED IN Cardiac Rhythm
STUDYING SLEEP Summary Statistics for Respiratory
Polysomnography Variables
Multiple Sleep Latency Test
ASSESSING MOVEMENTS IN SLEEP
Maintenance of Wakefulness Test
Periodic Limb Movements
Actigraphy
REM Sleep Without Atonia
Portable Monitoring
Bruxism
STAGING OF SLEEP
ASSESSING OTHER
Sleep Stages
PHYSIOLOGIC VARIABLES
Arousals
Core Temperature
Body Position
Esophageal pH Measurements
Summary Statistics for Sleep Variables
PERFORMANCE OF A SLEEP STUDY
ASSESSING RESPIRATION
ASSESSING SLEEP DISORDERS
DURING SLEEP
Disorders of Excessive Somnolence
Definitions
Parasomnias
Airflow
Respiratory Effort SUMMARY
Snoring Sounds
INTRODUCTION 1960s seminal discoveries were made delin-

eating the disturbances in sleep physiology in
Soon after the human electroencephalogram narcolepsy and obstructive sleep apnea. The
(EEG) was discovered in 1929, researchers clinical evaluation of sleep and its disorders
began recording brain electrical activity during began in the early 1970s, eventually leading
sleep. In 1937, Loomis described and classified to the development of the medical specialty
non-rapid eye movement (NREM) sleep, and of Sleep Medicine. The number of clinical
in 1953 rapid eye movement (REM) sleep was sleep laboratories, most within comprehensive
discovered by Aserinsky and Kleitman. In the sleep centers, continues to increase. The goal
697
of this chapter is to review the techniques • Indicated as a way to determine circadian

used to study sleep and discuss the physio- patterns and daily sleep time in patients
logic parameters recorded and the use of these with hypersomnia and may be useful prior
methodologies in elucidating sleep disorders. to an MSLT.
Purpose and Role of Polysomnography Purpose and Role of Portable Monitoring

• Indicated for the diagnosis of sleep- • Indicated to diagnose suspected OSA
related breathing disorders and the ini- in patients unable to be studied in a
tiation of therapy with positive airway sleep laboratory, for example, because
pressure. they are not ambulatory or medically
• Indicated for patients with neuromuscular unstable.
disease and symptoms of sleep disorders • Indicated for follow-up of previously
not diagnosed by history alone. diagnosed OSA to assess response to
• Indicated for the diagnosis of narcolepsy therapy.
and idiopathic hypersomnia. • May be used as an alternative to poly-
• Indicated for the elucidation of parasom- somnography to diagnose OSA in patients
nias that are potentially injurious, unusual, with a high pre-test probability of mod-
or atypical, or do not respond to con- erate to severe OSA in the absence of
ventional therapy, or when the distinction co-morbidities and as part of a compre-
from a seizure disorder is difficult despite hensive clinical sleep evaluation.
standard EEG.
• Indicated for the diagnosis of suspected
periodic limb movement disorder.
TECHNIQUES USED IN
Purpose and Role of the Multiple Sleep STUDYING SLEEP
Latency Test
• The multiple sleep latency test (MSLT) is Polysomnography
a validated objective measure of the ability
to fall asleep. Polysomnograms are recorded during the nor-
• Indicated as part of the evaluation of mal sleeping hours of a patient. Patients come
patients with suspected narcolepsy or idio- to the sleep laboratory 1–2 hours before their
pathic hypersomnia. usual bedtime. After the electrodes and sensors
have been applied, patients may watch televi-
Purpose and Role of the Maintenance of sion or read until they are ready to go to sleep.
Wakefulness Test They sleep 6–8 hours before either awakening
• The maintenance of wakefulness test spontaneously or being awakened by a techni-
cian. The goal of polysomnography is to quan-
(MWT) is a validated objective measure of
tify the amount of time spent in various stages
the ability to stay awake.
• Indicated to assess a person’s ability to of sleep during the night and to document clin-
ically relevant events that disrupt sleep, such
remain awake when failure to do so may
as cardiopulmonary abnormalities or abnormal
constitute a personal or public safety risk.
• May be indicated in patients with exces- motor activity.
The basic format of a comprehensive
sive sleepiness to assess response to
polysomnogram is standardized. The required
treatment.
minimum montage1 includes (1) continuous
Purpose and Role of Actigraphy monitoring of at least three channels of an
EEG, (2) two channels of an electro-oculogram
• Indicated to assist in the diagnosis of (EOG), (3) three channels of respiratory data
circadian rhythm disturbances, including (markers of airflow, respiratory effort, and
delayed sleep phase syndrome. oxyhemoglobin saturation), (4) electrocardio-
• Indicated to characterize circadian rhythms graphic (ECG) recording, and (5) electromyo-
and sleep disturbances in patients with graphy (EMG) of the chin and anterior tibial
insomnia. muscles. Other variables may be monitored as
Assessment of Sleep and Sleep Disorders 699
clinically indicated, including snoring, position sleep disorders may masquerade as sleep-
of the sleeper, additional EEG, surface EMG disordered breathing), and some patients may
recordings from all four limbs, esophageal have an abnormal sleep structure that is not
pressure, and esophageal pH. The patient’s easily scored by the machine. The analysis
nighttime behavior should also be observed of polysomnograms of neonates and infants
and recorded during polysomnography. requires special experience and skill because,
Polysomnograms are recorded with digital for these groups, EEG stages are defined dif-
systems that allow review at variable screen ferently, and respiratory and other behaviors
widths. A 30-second screen window is most during sleep are unique.
frequently used, but longer windows (60–180 An excerpt from a typical polysomnogram
seconds) may aid in evaluating respiratory or is shown in Figure 41–1, and a typical sum-
other periodic events and a shorter window mary of a polysomnographic recording (in
(10 seconds), in assessing EEG abnormalities. this case, from a patient who complained of
Digital systems allow 16–32 channels of data to excessive sleepiness) is shown in Figure 41–2.
be recorded. The usual placement of electrodes and sen-
Polysomnograms are usually interpreted sors and the amplification variables typically
visually. Even though sleep stages and other used for a polysomnogram are summarized in
polysomnographic variables are scored by the Table 41–1.
computer,2 visual inspection of the raw data is Current standards approved by the American
still necessary. The interpretation of many noc- Academy of Sleep Medicine (AASM)1 allow
turnal events requires clinical skill (e.g., other laboratories to choose between two different
Figure 41–1. Typical polysomnogram montage, including channels for eye movement, EEG, chin and leg EMG, air-
flow, snoring sounds, oxyhemoglobin saturation, and respiratory effort. Abdomen, abdominal breathing effort by inductive
plethysmograph; Airflow, nasal pressure, and oronasal thermocouple; ECG, electrocardiogram; EMG, electromyogram;
LOC, left outer canthus; Rib cage, thoracic breathing effort by inductive plethysmograph; ROC, right outer canthus; sono-
gram, sound recording; SpO2 , oxyhemoglobin saturation by pulse oximetry; Sum, electronic summation of rib cage and
abdominal movements.
Figure 41–2. A typical polysomnogram report. NREM, non-REM sleep; REM, rapid eye movement sleep.
EEG montages, a recommended and an alter- simple. However, low-amplitude vertical and
native version. The recommended montage oblique eye movements may be missed, and
consists of F4–M1, C4–M1, and O1–M1, while it is not possible to differentiate eye move-
the alternative montage consists of Fz–Cz, ment direction. In contrast, the alternative
Cz–Oz, and C4–M1. The recommended mon- montage, which is used in our laboratory, dis-
tage uses referential derivations, allowing the plays most eye movements, representing ver-
site of origin of signals to be determined tical eye movements as in-phase and hori-
by their maximal amplitude. The alternative zontal movements as out-of-phase deflections.
montage provides both referential and bipo- This is helpful in the differentiation of rapid
lar representation of sleep activity, with the eye movements of REM sleep (most com-
midline electrodes being less prone to contam- monly horizontal) from blinks of wakefulness
ination by EMG artifact than a mastoid refer- (resulting in vertical movements), especially in
ence. Similarly, recommended and alternative REM sleep without atonia. The chin EMG is
EOG montages are permitted (Fig. 41–3). The recorded with surface electrodes placed above
advantage of the recommended montage is and below the inferior edge of the mandible.
that all eye movements in any direction are The 10-Hz low-frequency filter setting used
represented by signals of the opposite phase, in the EMG derivations reduces movement
making differentiation from EEG and artifact artifact.
Table 41–1 Typical Settings for a Polysomnogram

Signal Sensor type and placement Sensitivity LFF, Hz HFF, Hz
EOG∗ Electrodes placed 1 cm lateral and 5–7.5 μV/mm 0.3 35

1 cm inferior to outer canthi
referenced to Fpz
EEG∗ Electrodes placed at Fz, Cz, Oz, C3, 5–7.5 μV/mm 0.3 35†
C4, M1, and M2
Chin EMG Electrodes placed in the midline 1 cm 2 μV/mm 10 100
above inferior edge of mandible and
2 cm below inferior edge to the right
and to the left of the midline
Leg EMG Electrodes placed over belly of 2 μV/mm 10 100
anterior tibialis muscle of each leg
ECG Electrodes placed on right shoulder 50 mV/mm 0.3 70
and left hip
Oxygen Light absorbance sensor placed on 100 mV/cm‡ DC 5
saturation earlobe or finger
Airflow Thermocouple placed in front of both Transducer- 0.1 15
nares and the mouth, connected in dependent‡
series to provide one signal
Nasal pressure transducer Transducer- 0.1 15
dependent
Breathing Inductors (elastic bands with Patient- DC 15
effort embedded wires) placed around the dependent‡
chest, just under the axillae, and
around the abdomen at the level of
the navel
Snoring Microphone placed in contact with Transducer- 10 100
the skin slightly superior and lateral to dependent‡
thyroid cartilage
LFF, low-frequency filter; HFF, high-frequency filter.
∗ These are the derivations used in our laboratory (AASM alternative derivations). See the text for the AASM recom-
mended derivations.
† The HFF can be set at 70 Hz if more detailed EEG analysis is desired.
‡ Optimum sensitivity may depend on manufacturer or equipment used, or recording conditions.
DC, direct current; ECG, electrocardiogram; EEG, electroencephalogram; EMG, electromyogram; EOG, electro-
oculogram.
Polysomnographic studies are indicated for presence of chronic lung disease unless
the following purposes:3 sleep-related upper airway obstruction
is also suspected. The mere presence
1. Assessment of sleep-disordered breath- of snoring, obesity, systemic hyperten-
ing, including suspected, obstructive, or sion, or nocturnal cardiac arrhythmia
central sleep apnea. Polysomnography without other symptoms is not an indi-
is also indicated for titration of posi- cation for polysomnography. However,
tive airway pressure (PAP) in patients patients with heart failure, coronary
with diagnosed sleep-disordered breath- artery, or cerebrovascular disease or sig-
ing. Polysomnography is not routinely nificant arrhythmias should be ques-
indicated for nocturnal hypoxemia in the tioned about symptoms of sleep apnea
Figure 41–3. AASM recommended and alternative derivations for recording eye movements (EOG). In the recom-
mended derivations, E1 and E2 are placed 1 cm below and 1 cm above the outer canthus of the left and right eyes. The
reference electrode is placed on the right mastoid. In the alternative derivations, E1 and E2 are placed 1 cm below and
1 cm lateral to the outer canthus of the left and right eyes. The reference electrode is Fpz.
and polysomnography performed if they indicated for typical, uncomplicated, and

are present. Patients with neuromuscu- noninjurious parasomnias that can easily
lar disorders and sleep-related symptoms be diagnosed on the basis of the med-
should undergo polysomnography if a ical history or for patients with known
sleep disorder cannot be confidently diag- nocturnal seizures but no other sleep
nosed by history. complaints.
2. Assessment of excessive daytime som- 4. Assessment of periodic limb movements
nolence (suspected narcolepsy or idio- of sleep (PLMS). Polysomnography is
pathic hypersomnia). This assessment indicated when PLMS are suspected
requires polysomnography and a subse- to be the cause of insomnia or exces-
quent MSLT. sive daytime sleepiness, especially in the
3. Assessment of parasomnias. An extended absence of restless legs syndrome. The
polysomnographic montage using addi- frequency of PLMS and their effect on
tional EEG and EMG channels and sleep continuity and architecture need to
time-synchronized audiovisual recording be determined. Polysomnography is not
is indicated for sleep-related violent or indicated for the diagnosis of restless legs
potentially injurious behavior, or behav- syndrome.
ior that is highly disturbing to the
sleep of others. Such a study is also
indicated for sleep behaviors that are Polysomnography is not indicated for the
unusual or atypical because of age at routine evaluation of transient or chronic
onset, frequency or duration of occur- insomnia. However, it may be indicated
rence, or the specifics of the particular if sleep-disordered breathing or PLMS are
motor pattern in question. Paroxysmal thought to comprise a significant compo-
arousals thought to be possible noctur- nent of the insomnia, if the initial diagnosis
nal seizures are a further indication, when is uncertain, or if behavioral or pharmaco-
the results of standard EEG are incon- logic treatment has been unsuccessful. In rare
clusive. Polysomnography is not routinely cases, polysomnography may be performed if
a sleep-state misperception syndrome is sus- latencies by disrupting sleep, no differences

pected. Such a syndrome is defined as a com- in latencies have been found after nights of
plaint of insomnia or excessive sleepiness that home vs. laboratory polysomnography.5 Sec-
occurs without any objective evidence of sleep ond, all treatment with psychotropic drugs that
disturbance. In this case, the polysomnogram can safely be discontinued should be stopped
would demonstrate normal sleep latency, a nor- at least 2 weeks before the study.
mal number of arousals, and normal sleep A mean sleep latency over four or five nap
duration. opportunities is calculated. If one of the sleep
latencies is an outlier (e.g., 3-, 4-, 5-, 4-, and
18-minute latencies), it may be more accurate
Multiple Sleep Latency Test to use the median than the mean sleep latency.6
A value of fewer than 5 minutes generally indi-
The MSLT4 consists of four or five 20-minute cates excessive daytime sleepiness, and a value
rest periods spaced 2 hours apart. The test is greater than 10 minutes falls within the sta-
conducted with the patient lying in bed in a tistically normal range. Values between 5 and
dark room and is almost always recorded on 10 minutes should be regarded as represent-
the day following a night with polysomnogra- ing a continuum, with lower values indicating
phy. The goal of the test is to quantify phys- greater sleepiness than higher values.7 Most
iologic sleepiness during waking hours and to patients with narcolepsy will have mean sleep
determine the occurrence of REM sleep near latencies <8 minutes.8 The presence of REM
sleep onset. For this test, patients are asked sleep occurring within 15 minutes of sleep
to remain in the laboratory for the entire day. onset (sleep onset REM [SOREM]) in any of
They may read, watch television, or engage in the nap opportunities is also recorded. The
other quiet activities, but they must not sleep presence of SOREMs in two or more naps,
or exercise between scheduled naps. Record- or in one nap and the preceding overnight
ings performed during the MSLT are simpler polysomnographic study, is considered abnor-
than those performed during polysomnogra- mal.4 Although this finding is regarded as
phy and typically only EEG, EOG, submental the neurophysiologic marker of narcolepsy, it
EMG, and ECG data are recorded. For each is not specific and other causes need to be
nap, sleep latency and the occurrence of REM considered. These include abrupt withdrawal
sleep are noted. For the MSLT, sleep latency of REM-suppressant medication (e.g., most
is defined as the time between “lights out” and stimulants and antidepressants), moderate or
the beginning of the first epoch of any stage severe OSA syndrome,9 or sleep deprivation.
of sleep. If no sleep occurs during the first If the sleepy patient has no history of cata-
20 minutes in bed, the rest period is discon- plexy, the results of the polysomnography are
tinued and the sleep latency is recorded as normal, and the MSLT shows a pathologi-
20 minutes. After sleep onset has occurred, cally short mean sleep latency with fewer than
patients are allowed to sleep for 15 minutes two SOREMs, the likely diagnosis is idiopathic
to determine whether they will enter REM hypersomnia.8
sleep. Accurate interpretation of the results
depends on fulfillment of certain precondi-
tions. First, adequate amounts of sleep must Maintenance of Wakefulness Test
be obtained for 1–2 weeks before the study
to ensure that the patient is not voluntarily A less frequently used variant of the MSLT is
sleep deprived. This is assessed by having the the maintenance of wakefulness test.4 This is
patient complete a sleep log and often wear a not a test for sleepiness but measures the abil-
wrist actigraph, a device that monitors motor ity of the patient to remain awake. It should not
activity of an arm and allows periods of rest be used to diagnose sleep disorders, but some-
and activity to be differentiated. At least 6, or times it is helpful in assessing the response to
preferably 7, hours of sleep, as measured by stimulants or a patient’s fitness to drive or to
the polysomnogram, should be obtained dur- fly an airplane. Conditions are the same except
ing the night before the MSLT. Although it that the patient sits in a comfortable recliner
might be imagined that a laboratory sleep study or bed with the back and head supported by a
the preceding night might affect the MSLT bed rest in a dimly lit room (7.5-W night light)
and is asked to try remain awake rather than 19 or less correlated with driving impairment.
to sleep. The patient may not use extraordinary A reasonable practical guide is to consider
methods to remain awake, such as slapping the latencies greater than 20 minutes as indicative
face or singing. Each trial lasts until unequiv- of adequate alertness.
ocal sleep occurs, defined as three consecutive
epochs of stage N1 sleep or one epoch of any
other stage, or after 40 minutes if no sleep Actigraphy
occurs. Sleep latency is measured from test
onset until the start of the first epoch scored A small, watch-like device is worn by the
as sleep. Normal mean sleep latency for this patient on the wrist of the nondominant hand,
protocol7, 10 is 30.4 minutes, but the data are usually for an entire week (Fig. 41–4). This
not normally distributed with 42% of all sub- device counts and stores in its memory the
jects remaining awake for 40 minutes on all number of wrist movements that occur for each
four tests. The 15th percentile falls between 16 1-minute epoch. Periods of relative absence
and 25 minutes, depending on the exact def- of such movements are interpreted as sleep
inition used for sleep onset. A recent study11 and periods of high activity as wakefulness.
determining the ability of the MWT to pre- Actigraphy is a valid way of assessing sleep
dict simulated driving performance in patients patterns in normal subjects and in patients
with OSA suggested that mean latencies of with some sleep disorders.12 In particular, there
Figure 41–4. Wrist actigraphy data recorded from 18:00 (6:00 pm) Friday through 18:00 the following Friday. The height
of each vertical black line is proportional to the level of activity for a 1-minute epoch. Areas where the lines are very short
or absent represent periods of immobility and probable sleep.
is good correlation between total sleep time and EMG may add useful additional informa-
measured by actigraphy and polysomnogra- tion, including determination of sleep stage
phy. Actigraphy, in general, shows a closer and total sleep time. If portable monitoring is
agreement with polysomnography than with used, it is essential that it be part of the com-
sleep logs. prehensive clinical assessment of patients by
Actigraphy is indicated for the following knowledgeable physicians and that the studies
purposes: be interpreted by sleep specialists with access
to the raw data.
1. To assist in the diagnosis of circadian The role of portable monitoring for the diag-
rhythm disorders, including delayed sleep nosis of OSA remains controversial. The Amer-
phase disorder, advanced sleep phase dis- ican Academy of Sleep Medicine has reviewed
order, and shift work sleep disorder. this practice and has suggested that under cer-
2. To characterize the sleep patterns in tain, carefully defined conditions, unattended
patients with insomnia, including insom- portable recording for the assessment of OSA
nia with depression. may be acceptable13 . These include suspected
3. To characterize the circadian pattern OSA in patients unable to be studied in a sleep
and daily sleep time in patients with laboratory, for example, because they are not
hypersomnia, especially prior to perform- ambulatory or medically unstable, follow up of
ing an MSLT. previously diagnosed OSA to assess response
to therapy, and as an alternative to polysomno-
graphy to diagnose OSA in patients with a
Actigraphy may be helpful in monitoring treat-
high pre-test probablility of moderate to severe
ment of patients with circadian rhythm dis-
OSA in the absence of co-morbidities and as
orders or insomnia. It may be used in older
part of a comprehensive clinical sleep evalua-
populations as well as in infants and children.
tion. It is not currently recommended for the
routine diagnosis of OSA.14 Portable monitor-
ing has a higher frequency of false positives and
Portable Monitoring negatives than conventional polysomnography
and it is by no means clear that widespread use
Portable monitoring incorporates a range of of the procedure would result in more cost-
techniques used for assessing OSA includ- effective therapy. However, two recent stud-
ing overnight oximetry, partial cardiorespira- ies15, 16 have suggested that diagnosis of OSA
tory studies, and fully portable polysomnog- by portable monitoring in selected patients
raphy with monitoring of EEG, EOG, and carefully followed in academic sleep centers
EMG. It is usually performed in the home by experienced sleep specialists may result in
without a technician in attendance. Overnight similar clinical outcomes after treatment com-
oximetry may be a useful screening technique pared with patients diagnosed by laboratory
in some patients but should not be used as polysomnography. Further larger outcome-
a definitive diagnostic method. Although the based research studies are in progress.
presence of unequivocal repetitive desatura-
tions in an oximetric tracing strongly sug- Key Points
gests sleep-disordered breathing, normal find-
ings on oximetry do not rule out sleep apnea. • The goal of polysomnography is to
Patients, especially younger ones, may have document clinical events that disrupt
sleep apneas serious enough to cause repeated sleep, including respiratory and move-
arousals from sleep without causing significant ment disorders.
oxyhemoglobin desaturation. Alternatively, the • Multiple physiologic sensors are used in
patient may not have slept much during polysomnography, including EEG, EOG,
the night when oximetry was performed or EMG, respiratory monitoring, and ECG.
may have positional sleep apnea and hap- • There are standard derivations for record-
pened to sleep only on the side during the ing EEG, EOG, and EMG in polysomno-
study. Portable monitors used to diagnose OSA graphy.
should at a minimum record oximetry, airflow, • The MSLT, an objective measure of the
and respiratory effort, although EEG, EOG, ability to fall asleep, must be performed
under carefully controlled conditions for work, involving eight task forces and approxi-
the results to be valid. mately 80 experts, culminated in the 2007 pub-
• The MWT measures a person’s ability to lication of the AASM Manual for the Scoring
remain awake and must be interpreted of Sleep and Associated Events: Rules, Termi-
with thorough understanding of norma- nology and Technical Specifications1 and seven
tive data. review articles explaining the evidence and
• Actigraphy is indicated for the assessment rationale behind the new rules.18
of sleep patterns in circadian rhythm dis-
orders, insomnia, and hypersomnia.
• Portable monitoring is not currently rec- Sleep Stages
ommended for the routine diagnosis of
OSA. If used, the results must be inter- Sleep is scored in arbitrary 30-second periods,
preted by a sleep specialist and used as known as epochs. For each epoch, a stage is
part of the comprehensive management of assigned that comprises the greatest portion of
the patient. that epoch. Sleep is divided into NREM sleep
(subdivided into three stages: N1, N2, and N3)
and REM sleep.
Wakefulness (Stage W) is scored if there
is activity in the 8–13 Hz (alpha) range over
STAGING OF SLEEP the occipital region with eye closure, attenu-
ating with eye opening, during the majority
As early as 1937, it was recognized that what of a 30-second epoch (Fig. 41–5). If alpha
is now called NREM sleep showed differ- rhythm is not discernable, stage W can also
ent electrophysiological phenomena at variable be scored in the presence of eye blinks, read-
times during the night with slow eye move- ing eye movements or rapid eye movements
ments, sleep spindles, and increasing rhythmic in association with normal or high chin muscle
slow activity. After the discovery of REM sleep tone.
in 1953, classifications of sleep stages matured, Stage N1 is defined by a relatively
culminating in the 1968 manual, edited by Alan low-amplitude mixed-frequency EEG pre-
Rechtschaffen and Anthony Kales.17 This clas- dominantly in the 4–7 Hz range. There are typ-
sification remained the standard for scoring ically slow, usually horizontal, eye movements
sleep in humans until the American Academy and often decreased tone on the chin EMG
of Sleep Medicine commissioned the develop- (Fig. 41–6). Vertex sharp waves (<0.5-second
ment of a new manual in 2004. Three years of duration, maximal over the central region) may
Figure 41–5. Wakefulness (stage W). Note the prominent alpha rhythm, rapid eye movements, and high EMG activity in
the chin and legs.
Figure 41–6. Stage N1 sleep. Note the vertex sharp waves, the slow rolling eye movements, and the absence of alpha
rhythm.
be seen. It may be difficult to distinguish stage complexes (diphasic sharp waves, initially neg-
W from stage N1 in the approximately 10% ative followed by a positive component, max-
of subjects who do not generate alpha rhythm imal over the frontal region lasting ≥0.5 sec-
during eye closure in wakefulness. In these ond) (Fig. 41–7). K complexes may occur spon-
subjects, stage N1 should be scored when the taneously or be evoked by intrinsic or extrinsic
earliest of the following phenomena occurs: sensory stimuli. In patients with conditions
slow eye movements, vertex sharp waves, or such as obstructive sleep apnea, light sleep may
slowing of the background EEG by at least be highly fragmented by K complexes followed
1 Hz. by runs of alpha rhythm at the termination of
Stage N2 is characterized by the appearance apneas. K complexes associated with arousals
of sleep spindles (trains of 11–16 Hz activ- in the absence of sleep spindles or spontaneous
ity, most commonly 12–14 Hz, maximal over K complexes are insufficient to justify a change
the central region lasting ≥0.5 second) or K from stage N1 to stage N2 sleep.19 Because
Figure 41–7. Stage N2 sleep. Note sleep spindles and K complexes that characterize stage N2 sleep.
Figure 41–8. Stage N3 sleep. More than 20% of the epoch contains 0.5–2 Hz activity of amplitude greater than 75 μV.
sleep spindles and K complexes are discrete at least 20% of the epoch. Sleep spindles may
and intermittent, intervals of low-amplitude persist in stage N3 sleep.
mixed-frequency activity between K complexes REM sleep (Stage R) is defined by a rel-
or spindles are still scored as stage N2 unless atively low-amplitude, mixed-frequency EEG,
there is evidence of a transition to stages W, similar to that seen in stage N1, in combination
N3, or R, an arousal or a major body movement with low chin EMG tone and episodic bursts
followed by slow eye movements. of rapid eye movements (Fig. 41–9). Sawtooth
Stages N3 sleep (also known as slow-wave waves (sharply contoured or triangular, often
sleep) is a combination of the Rechtschaffen serrated, 2–6 Hz waves over the central region,
and Kales stages 3 and 4 (Fig. 41–8). The defin- frequently preceding bursts of rapid eye move-
ing criteria are high-amplitude slow waves (at ments) may be seen in REM sleep. Transient
least 75 μV peak-to-peak measured over the muscle activity, previously referred to as pha-
frontal region, 0.5–2 Hz frequency) comprising sic muscle twitches, consists of short irregular
Figure 41–9. REM sleep (stage R). REM sleep is characterized by rapid eye movements, low chin muscle tone on EMG,
and an EEG pattern similar to that seen in stage N1 sleep. Sawtooth waves may also be seen.
Figure 41–10. Alpha intrusions into non-rapid eye movement sleep. This polysomnogram of a 50-year-old woman with a
complaint of chronic fatigue illustrates intrusion of diffuse alpha activity into slow-wave sleep.
bursts of EMG activity, usually <0.25 second activity superimposed on the normal activity of
duration, superimposed on low EMG tone. NREM sleep (Fig. 41–10). This phenomenon,
The presence of sawtooth waves or transient also called alpha–delta sleep,21 is often associ-
muscle activity is strongly supportive of stage R ated with chronic pain (e.g., in fibromyalgia or
sleep and may be helpful if scoring is in doubt. rheumatoid arthritis), but it also may occur in
Major body movements do not define a sleep patients who have no known medical disorder.
stage but occur when movement and muscle
artifact obscure more than half the EEG of an
epoch. They usually occur in the setting of an Body Position
arousal, so if alpha rhythm is discernable at all
during the epoch or in the preceding or fol- Body position may affect the severity of disor-
lowing epochs, the epoch with the movement dered breathing. It is always scored in clinical
should be scored as stage W. Otherwise the polysomnograms. This is done either with a
epoch is assigned the same stage as the epoch special position indicator worn by the patient
that follows it. or by observing the patient through a closed
circuit video monitor. If the patient does not
spontaneously sleep part of the time on the
Arousals back and part of the time on the side during the
recording, he or she is usually awakened and
Arousals are defined as abrupt shifts in EEG asked to sleep in the other position to assess
frequency to the alpha, theta, or fast beta respiration in both positions.
(>16 Hz) frequency bands. The subject must
be asleep for at least 10 seconds before the
arousal, and the EEG frequency shift must last
for a minimum of 3 seconds. Arousals scored Summary Statistics for Sleep
in REM sleep must also show increased chin Variables
EMG activity lasting at least 1 second. Scoring
these arousals is important for an assessment After all epochs have been scored, summary
of sleep quality: the more arousals there are, statistics are computed (Fig. 41–2). They
the worse one’s perception is of how well one include the following:3
has slept and the sleepier one is during the
subsequent day.20 1. Total recording time is from “lights out”
Alpha intrusion should be distinguished to “lights on” in the morning for the last
from arousals. It is characterized by alpha time.
2. Total sleep time includes all epochs scored To interpret arousal indices, it is impor-
as stages N1, N2, N3, or REM sleep. tant to know how many of the arousals
3. Sleep efficiency is computed as the ratio and awakenings are caused by disordered
of total sleep time to total recording time, breathing or by periodic limb movements.
expressed as a percentage. Arousals that do not meet the criteria for
4. Sleep latency is the lights out time to the hypopneas may be associated with sub-
start of the first epoch of any stage of tle increases in upper airway resistance.
sleep. Markedly increased arousal indices unas-
5. Wake after sleep onset is defined as the sociated with periodic limb movements
time in stage W during total recording or disordered breathing are often seen in
time, excluding sleep latency. patients with pain or other medical disor-
6. Sleep onset is the start of the first epoch ders or in those with psychologic distress,
scored as other than stage W (usually especially anxiety.
stage N1).
7. REM latency is the time from sleep onset Interpretation of the various sleep scores and
until the start of the first epoch scored as indices requires clinical judgment and should
stage R. REM latency decreases with age be made conservatively. Studies have shown
(Table 41–2). Initial REM latency con- that the patient’s sleep in the laboratory on the
siderably shorter than that expected for a first night may be atypically poor (first-night
patient’s age is nonspecific, but may sug- effect).24 Alternatively, some insomniacs sleep
gest consideration for diagnoses of major especially well on the first night in the labo-
depression, sleep deprivation, narcolepsy, ratory (reverse first-night effect).25 Because the
or withdrawal of REM-suppressing med- first night spent in the laboratory is frequently
ications. If narcolepsy is suspected, an atypical, research studies often use a mini-
MSLT is essential. mum of two laboratory nights to assess sleep
8. Time in each stage of sleep and percent- and discard the first as adaptation. Economi-
age of total sleep time spent in each stage cally, however, this is not feasible for clinical
should be recorded. studies. Therefore, clinical studies are rarely
9. Arousal index counts the number of performed to assess sleep architecture alone
arousals per sleep hour. The arousal but usually to assess factors that disrupt sleep,
indices considered normal depend on the such as breathing or movement disorders.
scoring criteria used and may also vary A comparison between the objectively
from laboratory to laboratory. A study of recorded sleep variables as discussed in this
arousals in normal subjects22 revealed a chapter and the patients’ self-reports of their
mean of 4 per hour using the original sleep (obtained by their answers to a ques-
Rechtschaffen and Kales definition17 and tionnaire the following morning) may yield
a mean of 20 per hour using the Amer- clinical insight, especially if there is a large
ican Sleep Disorders Association crite- discrepancy in total sleep time. Such a dis-
ria,23 which are very similar to the new crepancy may be explained by (1) a high
AASM arousal criteria described earlier.1 arousal index or alpha intrusions into sleep
In our laboratory, we consider 15 or or (2) sleep-state misperception, a form of
fewer arousals per hour to be normal. insomnia in which polysomnography reveals
Table 41–2 Normal Lower Bounds for REM Latency (Mayo

Sleep Disorders Center data)
Age (years) Initial REM Latency (minutes)
15–24 70
25–34 60
35–44 45
45–60 35
>60 30
REM, rapid eye movement sleep.
Figure 41–11. Sleep histogram of, A, a 19-year-old woman with no history of a sleep disorder and, B, a 58-year-old man
with severe obstructive sleep apnea. The man had severely fragmented sleep during the diagnostic study, with marked
rebound of rapid eye movement sleep during the CPAP trial. PLM, periodic limb movements of sleep; CPAP, air pressure
(cm H2 O) of continuous positive airway pressure system.
normal sleep while the patient complains of late (possibly suggesting influences of the cir-
severely restricted sleep time. Development of cadian rhythm or psychiatric factors).
sleep over an entire night is best represented
by results of a sleep histogram (Fig. 41–11).
This can indicate whether the sleep distur-
Key Points
bances (e.g., disordered breathing and peri-
odic limb movements) are evenly distributed • The AASM Manual for the Scoring of
over the entire sleep period or are associ- Sleep and Associated Events provides
ated with specific sleep stages, times of the rules for the scoring of sleep stages,
night, or body positions. They can also show arousals, and other physiologic parameters
whether excessive wakefulness occurs early or during polysomnography.
• Sleep is divided into NREM (stages N1, with increasing respiratory effort leading to an
N2, and N3) sleep and REM (stage R) arousal but not fulfilling criteria for an apnea
sleep, each with specific scoring criteria. or hypopnea. Traditionally, respiratory effort is
• Arousals are sudden shifts in EEG fre- monitored using esophageal pressure but this
quency (usually to alpha) lasting at least technique is rarely used today. Decreased nasal
3 seconds, following at least 10 seconds pressure signal amplitude with flattening of the
sleep and, when occurring in REM sleep, inspiratory component (“flow limitation”) and
are associated with increased chin tone. desynchrony of the abdominal and chest induc-
tance plethysmography signals are acceptable
surrogate markers, while a sequence of increas-
ing amplitude snores may be helpful collabora-
ASSESSING RESPIRATION tive information.
DURING SLEEP It is also necessary to determine if an apnea
or hypopnea is obstructive or central. Obstruc-
Most polysomnographic studies are performed tive apnea is defined as a cessation of air-
to assess disordered breathing during sleep. flow in the presence of measurable, often
The basic information obtained from polyso- gradually increasing, respiratory effort. Dur-
mnography about disordered breathing inclu- ing obstructive apnea, paradoxical breathing is
des the frequency and type of breathing often observed, that is, the chest expands as the
disturbances, how severely oxyhemoglobin abdomen contracts, but this is not essential for
saturation is affected, and whether or not there diagnosis (Fig. 41–12A). Central apnea shows a
is an associated cardiac arrhythmias. It is also of cessation of airflow coupled with a lack of respi-
clinical relevance to determine if there is a dif- ratory effort (Fig. 41–12B). Sleep-onset central
ference in the degree of disordered breathing apnea may be relatively benign; it often simply
related to body position or sleep stage. indicates anxious overbreathing during wake-
fulness, with normalization during sleep. Peri-
odic central apnea, or Cheyne–Stokes breath-
Definitions ing, during sleep suggests either poor cardiac
output (longer circulation time between lung
Disordered-breathing events associated with and blood gas sensors in the carotid body) or
sleep are classified as apneas, hypopneas, or problems with neuronal control of respiration.
respiratory effort-related arousals (RERAs).1 Apnea events with an initial central compo-
Apnea is defined as complete cessation of air- nent followed by an obstructive component are
flow, and hypopnea is a partial decrease in called mixed apneas (Fig. 41–12C).
airflow. In theory, apnea and hypopnea are To evaluate disordered breathing events,
distinct, but the difference in their clinical rel- a polysomnographic montage includes airflow
evance is slight, and reliable discrimination of evaluated by both a nasal pressure transducer
the two types of events may be technically and an oronasal thermal sensor, breathing
difficult. An apnea is defined as ≥90% drop effort, snoring intensity, a continuous mea-
in airflow amplitude measured by a thermal sure of oxyhemoglobin saturation, and cardiac
sensor lasting at least 10 seconds. There are rhythm. As mentioned above, the body posi-
two acceptable rules for defining a hypopnea. tion of the sleeper also has to be assessed.
The recommended AASM rule requires ≥30% If continuous positive airway pressure (CPAP)
drop in airflow measured by a nasal pres- is considered as a treatment for disordered
sure transducer lasting at least 10 seconds and breathing during sleep, polysomnography is
associated with ≥4% oxyhemoglobin desatu- used to determine the minimal CPAP pres-
ration. The alternative rule requires ≥50% sure that eliminates all disordered breathing
drop in airflow measured by a nasal pres- (including snoring) in all sleep stages and in
sure transducer lasting at least 10 seconds all positions. Therapeutic trials in the labora-
associated with ≥3% oxyhemoglobin desatura- tory help the patient to adapt to wearing the
tion or an arousal. The alternative definition mask while sleeping under professional super-
is favored for epidemiological research, but vision. During the CPAP trial, pressure must
most clinicians use the recommended rule. be monitored either by continuous recording
A respiratory effort-related arousal requires a on the polygraph or by careful written notation
sequence of breaths lasting at least 10 seconds of pressure changes.
but the constrictive mask is uncomfortable

and tends to disturb sleep. Its use is limited
to research applications. The first method for
detecting airflow during clinical polysomnog-
raphy is the use of thermal sensors (thermo-
couples or thermistors) placed in the airstream
at each nostril and the mouth.26 Thermal sen-
sors detect the temperature change between
inspired and expired air. Although this does
not quantitatively measure how much air is
flowing, changes in signal amplitude indicate
relative increases or decreases in airflow. The
second method measures changes in airway
pressure with a cannula placed in the nos-
trils and connected to a pressure transducer.26
Thermal sensors are more sensitive to small
volumes of airflow than nasal pressure trans-
ducers and therefore should be used to be
certain that an apnea, rather than a hypop-
nea, is present. In contrast, the nasal pressure
signal is more sensitive to minor changes in
airflow or increases in upper airway resistance
and thus should be used to define the pres-
ence of hypopneas or RERAs. As discussed
earlier, both the sensors are required in routine
polysomnograms. When CPAP is used during
polysomnography, thermal sensors do not func-
tion well. Most CPAP systems designed for
laboratory use provide analog signals for pres-
sure and airflow. These signals can be inter-
faced with the polygraph to allow continuous
monitoring of airflow during the CPAP trial.
Figure 41–12. Types of sleep apnea. A, Obstructive

Respiratory Effort
apnea is characterized by complete cessation of airflow,
paradoxical movement of the rib cage and abdomen, and The purpose of the respiratory pump mus-
oxyhemoglobin desaturation. B, Central apnea has simul- cles is to create a pressure gradient between
taneous cessation of airflow and respiratory effort, and the thoracic cavity and the atmosphere. The
oxyhemoglobin desaturation. C, Mixed apnea—an initial
central apnea followed by a later obstructive component.
resulting pressure variations can be measured
(From Kaplan, J. 1991. Diagnosis and therapy of sleep- accurately by placing a transnasopharyngeal
disordered breathing. In Respiratory care: A guide to clin- balloon-tipped, transducer-tipped, or water-
ical practice, ed. G. G. Burton, J. E. Hodgkin, and J. J. filled catheter into the esophagus. Although
Ward, 3rd ed., 279–87. Philadelphia: JB Lippincott Com- this technique is the reference standard for
pany. By permission of Lippincott Williams & Wilkins.)
measuring respiratory effort, it is invasive and
can be an unpleasant experience and, thus, is
not used routinely. Esophageal pressure may
Airflow be used in rare circumstances to definitively
distinguish central from obstructive apneas and
The reference standard for measurement of to identify RERAs.
airflow is a pneumotachometer, a device that The recommended method to detect res-
requires a snug-fitting mask over the face of piratory effort is inductance plethysmography,
the subject. Pneumotachometers provide an a noninvasive technique for detecting res-
accurate quantitative measurement of airflow, piratory effort based on measuring changes
in the dimensions of the thoracic cavity.27 technician’s personal judgment, using a grading
Motion of the diaphragm changes the vol- scale for loudness of snoring in four steps: (1)
ume of the abdominal cavity, and motion of barely audible, (2) audible at the bedside, (3)
the rib cage changes the volume of the tho- audible from the open door to the bedroom,
racic cavity. The inductance plethysmograph and (4) audible through the closed door.
has wires embedded in elastic bands that are
placed around the chest and abdomen to sense
changes in the cross-sectional area of each
of these two cavities. An electrical summa- Blood Gases
tion of the signals from rib cage and abdomen
provides a rough estimate of overall tidal vol- Continuous monitoring of oxyhemoglobin sat-
ume. The signals may be calibrated or uncal- uration is mandatory because it provides infor-
ibrated; however, calibrated signals should be mation about the consequences of respiratory
used to distinguish obstructive from central dysfunction.1 Monitoring of oxyhemoglobin
hypopneas. saturation is performed easily with a pulse
An alternative method for measuring breath- oximeter, a device that measures the light
ing effort is the monitoring of EMG activity in absorption of two wavelengths of red light
intercostal muscles. Because intercostal activ- passed through a capillary bed, such as that
ity is inhibited during REM sleep, electrodes of the earlobe or the nail bed of a finger.
are placed over the sixth or seventh intercostal The wavelengths used match the peak absorp-
space in an attempt to also record diaphrag- tion factors of oxyhemoglobin and deoxyhe-
matic EMG. This technique should only be uti- moglobin. The oximeter then calculates the
lized if inductance plethysmography becomes ratio of oxyhemoglobin to total hemoglobin and
unreliable. translates it into a digital display and an ana-
log voltage that is written out on the polygraph.
Because of lung–ear or lung–digit circulation
time, the nadir of the oxyhemoglobin satura-
Snoring Sounds tion graph usually follows the termination of
the respiratory event (by 7–9 seconds when the
In many sleep laboratories, a small micro- sensor is on the ear). A review of the com-
phone is attached to the throat of the patient pressed signal over the course of the night can
or mounted on the headboard or suspended allow easy identification of periods of maximal
above the bed. The output of the device may be respiratory dysfunction and their relationship
filtered and recorded directly on the polygraph to sleep stage and body position (Fig. 41–13).
or processed through an integrator or sound If CO2 retention is of concern in a patient,
level meter before recording. Digital systems end-tidal (capnography) or transcutaneous
allow for audio recording of the actual sound PCO2 may be recorded. The PCO2 of air sam-
for later review, a technique especially help- pled from the airway at end-expiration (end-
ful for distinguishing snoring from stridor. The tidal) approximates alveolar PCO2 . Transcuta-
recorded signal may be supplemented with the neous PCO2 uses a skin surface electrode to
Figure 41–13. Oximetry strip chart recording. The repetitive desaturations occurring in the early portion of the tracing
are caused by obstructive apneas. Initiation of CPAP eliminates the apnea during the later portion of the recording.
measure the PCO2 in tissue underlying the the number of events per sleep hour
skin. Both the techniques provide data about with 3% or 4% oxyhemoglobin desatura-
trends in arterial PCO2 but should be sup- tions. The occurrence of Cheyne–Stokes
plemented with blood gas analysis of arterial breathing or hypoventilation should be
blood samples for accurate diagnosis of alveolar specified.
hypoventilation. 4. Snoring is often reported on a scale
from 1 to 4 (see the section on Snoring
Sounds).
Cardiac Rhythm
Key Points
The ECG is recorded to detect changes in the • Respiratory events during sleep are classi-
cardiac rhythm related to disordered breathing fied into apneas (obstructive, central, and
during sleep. A single lead II is derived from mixed), hypopneas, and RERAs.
electrodes placed on the torso or the right arm • Airflow is monitored by both a nasal
and left leg.28 Additional leads can be placed if pressure transducer and a thermal sen-
clinically indicated. sor; the former is used to define hypop-
neas and RERAs and the latter to define
apneas.
Summary Statistics for Respiratory • Respiratory effort is most commonly
Variables monitored by inductance plethysmogra-
phy and occasionally by intercostal and
The following respiratory variables are typically diaphragmatic EMG.
reported on the polysomnographic summary • Other respiratory sensors include pulse
sheet: oximeters and microphones to record
snoring.
• The ECG should be recorded using stan-
1. The number of central, obstructive, or
mixed apneas and the number of hypop- dard or modified lead II.
neas should be recorded. The apnea
hypopnea index (AHI) indicates the com-
bined number of apneas and hypopneas
per sleep hour. This is usually reported ASSESSING MOVEMENTS IN
for total sleep time, NREM and REM SLEEP
sleep time, and sleep time in the supine
vs. other positions. AHI is the standard
indicator of severity for obstructive sleep Periodic Limb Movements
apnea. Fewer than 5 events per hour is
considered normal, 5–15 events per hour PLMS, usually of the legs but rarely the arms,
is considered mild, 15–30 is moderate, may occur at regular intervals during sleep.
and more than 30 is severe, although They are recorded with surface electrodes
degree of oxyhemoglobin desaturation applied over the anterior tibialis muscle of both
and associated arousals should also be legs. Either one channel of EMG is recorded
taken into account. from both legs or two channels are recorded,
2. The number of RERAs and the RERA one per leg. The latter arrangement elimi-
index (number of RERAs per hour of nates much of the ECG artifact that is typical
sleep) should be recorded. in a one-channel recording from both legs.
3. Mean and minimum oxyhemoglobin sat- PLMS are scored only if at least four occur
uration values during wakefulness, REM in sequence, with a duration of 0.5–10 sec-
sleep, and NREM sleep are usually given onds each and an interval of 5–90 seconds
or a range is indicated. Optional measures between the onsets of consecutive movements
include the percentage of time during (Fig. 41–14).1 Each movement must have a
sleep that the patient spends with oxy- minimum amplitude of 8 μV, measured as a
gen saturation above 80% or 90%, and change in EMG amplitude over the resting
Figure 41–14. Periodic limb movements of sleep (PLMS). Bursts of anterior tibial surface EMG occurring at approxi-
mately 20-second intervals are accompanied by arousals lasting 3 seconds or longer.
EMG.29 Bilateral leg movements separated by as narcolepsy or idiopathic hypersomnia and

less than 5 seconds between movement onsets may not in themselves cause symptoms.30
are counted as single movements. Periodic However, if a high percentage of the move-
limb movements show flexions of hip, knee, ments cause the sleeper to awaken, PLMS
and ankle joints together with extension of the can potentially contribute to excessive daytime
toes, similar to a withdrawal response. Periodic somnolence, although care should be taken to
limb movements need to be distinguished from first rule out other causes of sleepiness.
gross body movements, which also show EMG Periodic limb movements can also be
artifact in the ECG and EEG leads, and from recorded during wakefulness, especially in
body movements associated with arousal from patients with RLS. Their presence during peri-
sleep apnea. ods of wakefulness on a polysomnogram can
PLMS occur in many normal, especially be helpful diagnostically, and their presence
older, sleepers but may also occur in sev- can be formally quantitated by the suggested
eral disorders, including restless legs syndrome immobilization test for RLS. In this test, the
(RLS). Generally, polysomnography is not indi- patient is asked to lie down in the early
cated for the diagnosis of RLS, which is made evening but to remain awake and not to move
on the basis of the history because the pres- the legs voluntarily. Anterior tibial EMG is
ence of PLMS is neither sufficiently spe- recorded, and the periodic limb movement
cific nor sensitive. When limb movements are index over 30–60 minutes is calculated. More
not associated with arousals, their significance than 40 PLMS per hour have both a sensitiv-
in causing daytime sleepiness is uncertain. ity and specificity of 81% for the diagnosis of
In many cases, PLMS may be an epiphe- RLS.31 The test is often uncomfortable because
nomenon accompanying other disorders such patients with RLS need to move their legs
to relieve the discomfort and thus is rarely of the mini-epochs contain transient muscle
performed. activity, this is considered excessive.1 Experi-
A periodic limb movement index (i.e., the ence and judgment are needed to determine
number of periodic limb movements per sleep whether muscle tone is excessive during REM
hour) is reported, followed by a periodic limb sleep over the course of an entire night, as
movement arousal index. This is the number of adequate normative data are not available.
periodic limb movements per sleep hour that REM sleep without atonia is the neurophysio-
led to arousal. Movements occurring within 0.5 logic marker of REM sleep behavior disorder
second before or after an apnea or hypopnea (RBD) (see the section on Assessing Sleep
should not be counted.1 Disorders). In RBD, the technician in addi-
tion often observes vigorous twitching, quasi-
purposeful movements such as punching in
REM Sleep Without Atonia the air, and vocalization during REM sleep
(Fig. 41–15).
REM sleep without atonia is scored when both
the EEG and the EOG suggest REM sleep
but the chin or anterior tibial EMG does not Bruxism
show the expected muscle atonia. This usu-
ally takes the form of a marked increase in Bruxism can be recognized by a sequence of
transient muscle activity (“phasic twitches”), three or more brief elevations in chin EMG
but sometimes sustained tonic muscle activ- lasting 0.25–2 seconds each or sustained eleva-
ity is present. Transient muscle activity can be tion lasting more than 2 seconds.1 Audio and
quantitated by dividing each 30-second epoch visual recordings are essential in confirming
into ten 3-second mini-epochs. If at least five the diagnosis.
Figure 41–15. REM sleep without muscle atonia. The montage is modified to record EMG activity from all four limbs
from an older man with Parkinson’s disease. Note frequent bursts of activity in the limb and submental EMG leads. Typi-
cally, the movements associated with this activity are related to dream content. L.A.T. = left anterior tibial; R.A.T. = right
anterior tibial; L.E.D. = left extensor digitorum (arm); R.E.D. = right extensor digitorum (arm) (From Daube, J. R., G. D.
Cascino, R. M. Dotson, M. H. Silber, and B. F. Westmoreland. 1998. Continuum: Lifelong learning in neurology [Clinical
Neurophysiology], Vol. 4, Part A, 169. Baltimore: Lippincott Williams & Wilkins. By permission of the American Academy
of Neurology.)
Key Points indicated, polysomnography with or without a

subsequent MSLT is then scheduled.
• Periodic limb movements are defined as
On the appointed night, patients come to the
trains of at least four anterior tibial EMG sleep center at about 7 pm. First, they fill out
bursts, each of amplitude at least 8 μV a questionnaire about their activity during the
above baseline and lasting 0.5–10 seconds, past day, about their intake of medicine, cof-
with a time of 5–90 seconds between the fee, and alcohol, and about emotional issues
start of consecutive movements. that might disrupt sleep. A technician applies
• PLMS are common in restless legs syn-
the electrodes and sensors necessary for the
drome, but are nonspecific and are often polysomnographic study. Patients then watch
asymptomatic. television or read until they are ready to go
• REM sleep without atonia is the neuro-
to bed. After the patients are connected to
physiologic substrate of RBD. the polysomnograph, biocalibrations are made
• Bruxism can be recognized by a charac-
(patients are asked to move and blink their
teristic appearance of elevations in the eyes, move their legs, grit their teeth, breathe
chin EMG in association with audiovisual exclusively through either the nose or the
recordings. mouth, and perform an isovolume maneuver to
maximize the recording of paradoxical breath-
ing). Patients stay in bed for a minimum of
ASSESSING OTHER 6 hours after “lights out” (often longer) except
PHYSIOLOGIC VARIABLES for short trips to the bathroom. A technician
monitors the recording throughout the night,
observes the patients through video moni-
Core Temperature toring, and notes on the record a patient’s
position and any potentially significant events
Monitoring core temperature may provide (e.g., patient vocalization, environmental noise
information about the phase and amplitude of that arouses the patient, and loudness of
the circadian cycle. Theoretically, this may be snoring).
helpful in the diagnosis of circadian rhythm If a diagnosis of OSA syndrome is made early
disorders, but in practice it is used only as a in the night, the second half of the night may
research technique. be used for a therapeutic trial of nasal CPAP.
The AASM recommends that such “split-night”
studies be restricted to patients with apnea–
Esophageal pH Measurements hypopnea indices greater than 40 (or greater
than 20 if apneas are long with major desatu-
Many patients experience reflux of stomach rations) in the first half of the night.3 However,
contents into the esophagus during sleep. economic pressures have forced an increasing
Clearing of this material from the esopha- number of sleep laboratories to perform “split-
gus is markedly delayed during sleep. In spe- night” studies routinely if disordered breath-
cialized laboratories, esophageal pH may be ing of lesser severity is detected in the first
measured to confirm a diagnosis of gastroe- half of the study. No differences in medium
sophageal reflux and to determine any relation- and long-term CPAP use have been found
ship between episodes of reflux and arousals. in patients titrated with a split-night protocol
compared with those undergoing a full-night
CPAP study.32
In the morning, patients fill out another
PERFORMANCE OF A SLEEP questionnaire about their perception of the
STUDY quality of their sleep in the laboratory. A tech-
nician then scores the recordings and com-
Before evaluation, patients are usually sent putes various indices. After the sleep specialist
questionnaires about sleep and a sleep log to has reviewed the record, a follow-up confer-
be filled out for a minimum of 1 week. A sleep ence with the patient is scheduled to discuss
disorder specialist, ideally board certified in the results, to make a final diagnosis, and to
Sleep Medicine, then interviews the patient. If initiate treatment if indicated.
Key Points sleep log or wears an actigraph for a week. If a

cause, such as sleep apnea, is found, the MSLT
• Polysomnography should be performed by
scheduled for the following day is canceled and
a sleep specialist as part of a comprehen- the problem is treated. If the polysomnogram
sive clinical assessment of patients with is normal and nighttime sleep is adequate (6 or
suspected sleep disorders. 7 hours), MSLT is performed. Interpretation of
• Split-night polysomnograms may be cost-
the MSLT is discussed earlier.
effective for the diagnosis and initiation of
treatment for OSA, especially in patients
with moderate to severe disease.
Parasomnias
Parasomnias are undesirable, mainly motor,
ASSESSING SLEEP DISORDERS phenomena that occur during sleep. They
include disorders of arousal (e.g., sleepwalking
and sleep terrors), REM-associated parasom-
Disorders of Excessive nias (e.g., RBD and nightmare disorder), and
Somnolence other parasomnias (e.g., sleep-related groaning
and sleep-related eating disorder).8 The indi-
Excessive daytime somnolence is defined as the cations for studying parasomnias are discussed
tendency to fall asleep very easily. The abil- earlier under Polysomnography. The labora-
ity to fall asleep quickly differentiates exces- tory assessment of parasomnias requires more
sive daytime somnolence from chronic fatigue, sophisticated recording systems and additional
malaise, or low-grade depression, in which pat- interpretive abilities than routine polysomnog-
ients may be exhausted and lie in bed for most raphy.34 Additional polygraphic channels are
of the day but are usually unable to fall asleep needed, including one or more channels for
quickly when asked to do so. Excessive day- recording arm EMG (usually over the exten-
time somnolence may be caused by an extrinsic sor digitorum muscles) and additional EEG
or intrinsic cause. The most common extrinsic channels (preferably 16), to rule out seizures
cause is voluntary sleep deprivation (insuffi- that can mimic parasomnias (Fig. 41–16). Dig-
cient sleep syndrome), often found in people ital recording systems with a 10-second screen
who work two jobs or have other reasons for size and high-resolution, time-synchronized
not allowing themselves enough time in bed. If videotape recording under infrared light are
insufficient sleep syndrome is suspected, it is essential. Physicians interpreting parasomnia
often useful to ask patients with excessive day- studies should be familiar with both rou-
time somnolence to stay in bed 1 hour longer tine polysomnographic interpretation and ictal
for an entire week, regardless of how long they EEG patterns.
have stayed in bed until that time, and observe Disorders of arousal (sleepwalking, sleep
whether excessive daytime somnolence wanes. terrors, or confusional arousals) are charac-
Other extrinsic causes include the use of sedat- terized by sudden arousals from slow-wave
ing medications, alcohol or recreational drugs, sleep, followed by abnormal motor behavior.
shift work, or environmental noise (including They are most common in children but may
the snoring of a spouse).33 commence or persist in adulthood. Although
The most common intrinsic cause of exces- the conditions are usually benign, occasionally
sive sleepiness is obstructive sleep apnea. patients may suffer or inflict severe injuries.
Other causes include central sleep apnea, The polysomnogram shows a rapid, usually
narcolepsy, idiopathic hypersomnia, disorders unprovoked arousal from slow-wave sleep. This
of circadian rhythm such as delayed sleep is sometimes preceded by a series of hyper-
phase disorder, and PLMS (if these are asso- synchronous delta waves without epileptiform
ciated with a high percentage of arousals).8 activity, seen in the scoring channel in 47% of
Polysomnography (with or without a subse- events.35 The actual event is usually obscured
quent MSLT) is needed to diagnose most by movement artifact, but the record may show
intrinsic causes (circadian rhythm disorders are tachycardia and varied EEG patterns, includ-
exceptions). Initially, polysomnography is per- ing rhythmic delta activity, irregular mixed
formed, usually after the patient completes a frequency activities (predominantly delta and
Figure 41–16. Parasomnia recording using an expanded EEG montage, illustrating arousal parasomnia. There is a sudden
partial arousal from stage N3 non-rapid eye movement sleep without epileptiform activity. Note that the arousal is preceded
by a series of hypersynchronous delta waves and that some slow activity continues throughout the arousal. In this 9-year-old
boy, the episodes of sleep terror were precipitated by stridor (visible as deflections on the sonogram channel) as a result
of vocal cord paresis following surgery for a posterior fossa medulloblastoma. LOC, left outer canthus; ROC, right outer
canthus; sonograph, recording of upper airway sound. (From Daube, J. R., G. D. Cascino, R. M. Dotson, M. H. Silber,
and B. F. Westmoreland. 1998. Continuum: Lifelong learning in neurology [Clinical Neurophysiology], Vol. 4, Part A, 166.
Baltimore: Lippincott Williams & Wilkins. By permission of the American Academy of Neurology.)
theta), or alpha rhythm. Compared with nor- REM sleep behavior disorder is character-
mal age-matched controls, the polysomnogram ized by an abnormal persistence of muscle
of patients with disorders of arousal shows a tone during REM sleep with dream enact-
higher percentage of slow-wave sleep, more ment behavior.37 This usually takes the form
frequent arousals from slow-wave sleep, and of arm flailing and kicking with vocaliza-
a more even distribution of slow-wave sleep tions. If the patient is wakened, a violent
through the night.36 Although these findings dream is often recalled. Injuries to the patient
shed interesting light on the pathogenesis of and bed partner are common. The condition
the disorder, they are not specific enough to occurs most frequently in older men and is
be of diagnostic help. The polysomnographic often associated with neurodegenerative dis-
appearances of sleep terrors, sleepwalking, and ease such as Parkinson’s disease, dementia with
confusional arousals are identical, and video Lewy bodies, or MSA (synucleinopathies). The
recording is essential to delineate fully the polysomnogram shows abnormally increased
nature of the event. Even if a typical episode transient muscle activity in REM sleep and
is not recorded the night of the study, careful occasionally a persistent tonic EMG. Even
review of the tracing often reveals the presence if no gross movements are recorded the
of minor confusional arousals. night of the study, muscle tone during REM
sleep usually remains high and can be recog- REFERENCES

nized. It is essential to record an additional
upper-extremity EMG channel, because not 1. Iber, C., S. Ancoli-Israel, A. Chesson, et al. 2007. The
all skeletal muscles are involved in any one AASM Manual for the Scoring of Sleep and Associated
Events: Rules, Terminology and Technical Specifica-
patient. tions, 1st ed. Westchester, IL: American Academy of
Sleep Medicine.
Key Points 2. Hirshkowitz, M., and C. A. Moore. 1994. Issues in
computerized polysomnography. Sleep 17:105–12.
• Excessive somnolence may be caused by 3. Kushida, C. A., M. R. Littner, T. I. Morgenthaler,
insufficient sleep syndrome, environmen- et al. 2005. Practice parameters for the indications for
tal sleep disorder, hypersomnia induced polysomnography and related procedures: An update
for 2005. Sleep 28:499–521.
by drugs, sleep apnea, narcolepsy, idio- 4. Standards of Practice Committee of the American
pathic hypersomnia, and circadian rhythm Academy of Sleep Medicine. 2005. Practice parame-
disorders. ters for clinical use of the multiple sleep wakefulness
• The elucidation of excessive somnolence test and the maintenance of wakefulness test. Sleep
28:113–21.
needs a careful history and the judicious 5. Kingshott, R. N., and N. J. Douglas. 2000. The effect
use of sleep studies such as actigraphy, of in-laboratory polysomnography on sleep and objec-
polysomnography, and the MSLT. tive daytime sleepiness. Sleep 23:1109–13.
• Polysomnography to elucidate the para- 6. Benbadis, S. R., M. C. Perry, B. R. Wolgamuth, et al.
somnias must include additional EEG and 1995. Mean versus median for the multiple sleep
latency test. Sleep 18:342–5.
EMG derivations and time-synchronized 7. Arand, D., M. Bonnet, T. Hurwitz, et al. 2005.
audiovisual monitoring. Interpretation The clinical use of the MSLT and MWT. Sleep 28:
must be performed by physicians trained 123–44.
to recognize electroencephalographic sei- 8. American Academy of Sleep Medicine. 2005. The
International Classification of Sleep Disorders: Diag-
zures as well as polygraphic manifestations nostic and Coding Manual, 2nd ed. Westchester, IL:
of sleep disorders. American Academy of Sleep Medicine.
• Even if characteristic events do not occur 9. Aldrich, M. S., R. D. Chervin, and B. A. Malow. 1997.
the night of the study, the diagnosis of Value of the multiple sleep latency test (MSLT) for the
REM sleep behavior disorder can be diagnosis of narcolepsy. Sleep 20:620–9.
10. Doghramji, K., M. M. Mitler, R. B. Sangal, et al. 1997.
made by identifying increased muscle A normative study of the maintenance of wakeful-
tone in REM sleep, and often the diagno- ness test (MWT). Electroencephalography and Clini-
sis of a disorder of arousal can be made by cal Neurophysiology 103:554–62.
identifying minor confusional arousals. 11. Sagaspe, P., J. Taillard, G. Chaumet, et al. 2007. Main-
tenance of wakefulness test as a predictor of driving
performance in patients with untreated obstructive
sleep apnea. Sleep 30:327–30.
12. Morgenthaler, T. I., C. Alessi, L. Friedman, et al. 2007.
SUMMARY Practice parameters for the use of actigraphy in the
assesment of sleep and sleep disorders: An update for
2007. Sleep 30:519–29.
This chapter reviews the techniques that are 13. Collop, N. A., W. M. Anderson, B. Boehlicke, et al.
available to study sleep and disorders that 2007. Clinical guidelines for the use of unattended
may occur during sleep. Multiple physiologic portable monitors in the diagnosis of obstructive sleep
apnea in adult patients. Journal of Clinical Sleep
parameters can be assessed using one or a com- Medicine 3:737–747.
bination of polysomnography, multiple sleep 14. Chesson, A. L., R. B. Berry, and A. Pack. 2003.
latency or maintenance of wakefulness tests, Practice parameters for the use of portable mon-
actigraphy, and portable monitoring. The use itoring devices in the investigation of suspected
of these tests is helpful to not only identify dif- obstructive sleep apnea in adults. A joint project
sponsored by the American Academy of Sleep
ferent stages of sleep, but to assess sleep disor- medicine, the American Thoracic Society and the
ders, such as abnormal movements in sleep and American College of Chest Physicians. Sleep 26:
disorders manifesting with excess somnolence. 907–13.
The field of Sleep Medicine and the associ- 15. Mulgrew, A. T., N. Fox, N. T. Ayas, and C. F. Ryan.
2007. Diagnosis and initial management of obstruc-
ated sleep neurophysiology continues to grow tive sleep apnea without polysomnography: A ran-
and expand our understanding of sleep-related domized validation study. Annals of Internal Medicine
disorders. 146:157–66.
16. Whitelaw, W. A., R. F. Brant, and W. W. Flemons. 28. Caples, S. M., C. L. Rosen, W. K. Shen, et al. 2007.
2005. Clinical usefulness of home oximetry compared The scoring of cardiac events during sleep. Journal of
with polysomnography for assessment of sleep apnea. Clinical Sleep Medicine 3:147–54.
American Journal of Respiratory and Critical Care 29. Walters, A. S., G. Lavigne, W. Hening, et al. 2007.
Medicine 171:188–93. The scoring of movements in sleep. Journal of Clinical
17. Rechtschaffen, A., and A. Kales. 1968. A manual of Sleep Medicine 3:155–67.
standardized terminology, techniques, and scoring sys- 30. Montplaisir, J., M. Michaud, R. Denesle, and
tem for sleep stages of human subjects. Bethesda, A. Gosselin. 2000. Periodic leg movements are not
MD: National Institute of Neurological Disease and more prevalent in insomnia or hypersomnia but are
Blindness. specifically associated with sleep disorders involv-
18. Iber, C., S. Ancoli-Israel, M. Chambers, and S. F. ing a dopaminergic mechanism. Sleep Medicine 1:
Quan. 2007. The new sleep scoring manual—The 163–7.
evidence behind the rules. Journal of Clinical Sleep 31. Montplaisir, J., S. Boucher, A. Nicolas, et al. 1998.
Medicine 3:107. Immobilization tests and periodic leg movements in
19. Silber, M. H., S. Ancoli-Israel, and M. H. Bonnet. sleep for the diagnosis of restless leg syndrome. Move-
2007. The visual scoring of sleep in adults. Journal of ment Disorders 13:324–9.
Clinical Sleep Medicine 3:121–31. 32. McArdle, N., A. Grove, G. Devereaux, et al.
20. Bonnet, M. H. 1986. Performance and sleepiness 2000. Split-night versus full-night studies for sleep
as a function of frequency and placement of sleep apnoea/hyponea syndrome. The European Respiratory
disruption. Psychophysiology 23:263–71. Journal 15:670–5.
21. Hauri, P., and D. R. Hawkins. 1973. Alpha-delta sleep. 33. Beninati, W., C. D. Harris, D. L. Herold, and J. W.
Electroencephalography and Clinical Neurophysiol- Shepard. 1999. The effect of snoring and obstructive
ogy 34:233–7. sleep apnea on the sleep quality of bed partners. Mayo
22. Mathur, R., and N. J. Douglas. 1995. Frequency of Clinic Proceedings 74:955–8.
EEG arousals from nocturnal sleep in normal subjects. 34. Aldrich, M. S., and B. Jahnke. 1991. Diagnostic
Sleep 18:330–3. value of video-EEG polysomnography. Neurology
23. Sleep Disorders Atlas Task Force of the American 41:1060–6.
Sleep Disorders Association. 1992. EEG arousals: 35. Schenck, C. H., J. A. Pareja, A. L. Patterson, and
Scoring rules and examples. Sleep 15:173–84. M. W. Mahowald. 1998. Analysis of polysomnographic
24. Agnew, H. W. J. 1966. The first night effect: An EEG events surrounding 252 slow-wave sleep arousals in
study of sleep. Psychophysiology 2:263–6. thirty-eight adults with injurious sleepwalking and
25. Hauri, P. J., and E. M. Olmstead. 1989. Reverse first sleep terrors. Journal of Clinical Neurophysiology 15:
night effect in insomnia. Sleep 12:97–105. 159–66.
26. Redline, S., R. Budhiraja, V. Kapur, et al. 2007. 36. Espa, F., B. Ondze, P. Deglise, et al. 2000. Sleep archi-
The scoring of respiratory events in sleep: Reliabil- tecture, slow wave activity, and sleep spindles in adult
ity and validity. Journal of Clinical Sleep Medicine 3: patients with sleepwalking and sleep terrors. Clinical
169–200. Neurophysiology 111:929–39.
27. Staats, B. A., H. W. Bonekat, C. D. Harris, and K. 37. Olson, E. J., B. F. Boeve, and M. H. Silber. 2000.
P. Offord. 1984. Chest wall motion in sleep apnea. Rapid eye movement sleep behavior disorder: Demo-
The American Review of Respiratory Disease 130: graphic, clinical and laboratory findings in 93 cases.
59–63. Brain 123:331–9.
SECTION 2
ELECTROPHYSIOLOGIC
FUNCTION
PART H
Intraoperative Monitoring
The central and peripheral nervous systems are with the level and type of surgical procedure:
at risk for damage during surgical procedures, supratentorial (Chapter 42), posterior fossa
particularly vascular, orthopedic, and neurosur- (Chapter 43), spinal column (Chapter 44), and
gical procedures. Although some types of dam- peripheral nerve (Chapter 45).
age may be expected because of the nature of As surgeons have become more familiar
the procedure, other types can occur unexpect- with the benefits of intraoperative monitor-
edly. In either case, damage may be reversible ing of neural function, the demand for it
if the surgeon is made aware of the change has increased. This change is reflected in
and takes appropriate action. Standard clini- the expanded discussion in this edition of
cal tools cannot assess neural function during the compound muscle action potentials, motor
surgical procedures. Therefore, surgeons have evoked potentials, electromyographic record-
had relatively little information on which to ings, carotid stump, and compressed spectral
base decisions about modifying the procedure array, and in the revision of the discussion on
in response to impending damage to neural intraoperative monitoring during nerve entrap-
tissue. ment procedures. As post-traumatic brachial
Most electrophysiologic measurements plexus reconstruction methods have devel-
described in this book can be made intra- oped, there has been an increasing call for neu-
operatively to monitor neural function. Elec- rophysiologic monitoring during the surgery.
troencephalography can be used to monitor With increased experience with intraoperative
the status of cortical function; somatosensory monitoring, more specific criteria have been
evoked potentials, to monitor sensory pathways developed to prevent damage.
in the periphery, spinal cord, and brain; audi- Continuous electrophysiologic monitoring
tory evoked potentials, to monitor peripheral can be helpful in the intensive care unit.
and central auditory pathways; nerve conduc- Patients in an intensive care unit either have
tion studies and electromyography, to monitor a major neurologic disease or are at risk
peripheral nerve damage; and motor evoked of having one develop as a complication
potentials, to monitor descending motor path- of another disorder. Electrophysiologic tech-
ways in the brain stem and spinal cord. Each niques can be used to monitor neural func-
of these techniques has been modified, so it tion in this setting just as they can be in
can be used in the operating room for a wide the operating room to identify early or oth-
variety of surgical procedures. These moni- erwise unrecognizable neural damage. The
toring techniques provide helpful guidance to chapters in this section illustrate the appli-
the surgeon during the procedure and have cations of electrophysiologic techniques both
reduced morbidity associated with certain pro- in the operating room and in the intensive
cedures. The optimal monitoring methods vary care unit.
725
Chapter 42
Cerebral Function Monitoring

Elson L. So and Frank W. Sharbrough
INTRODUCTION SEP RECORDING DURING

TECHNICAL FACTORS IN CAROTID ENDARTERECTOMY
INTRAOPERATIVE EEG OTHER MONITORING
MONITORING TECHNIQUES DURING
EFFECTS OF ANESTHESIA ON CAROTID ENDARTERECTOMY
ELECTROENCEPHALOGRAPHY EEG MONITORING DURING
SYMMETRICAL EEG PATTERNS CARDIAC SURGERY
DURING ANESTHESIA BISPECTRAL ANALYSIS OF EEG
PREOPERATIVE FOCAL FOR MONITORING DEPTH OF
ABNORMALITIES SEEN WITH ANESTHESIA
ANESTHESIA EEG MONITORING FOR
CLINICAL APPLICATIONS EPILEPSY SURGERY
EEG CHANGES DURING SUMMARY
CAROTID ENDARTERECTOMY
INTRODUCTION causes of ischemia occur less commonly than

during operations on the carotid artery.
Among the first applications of intraoperative Purpose and Role of Cerebral Function
recording of cerebral electrical activity were Monitoring
corticography and, later, acute depth stud-
ies during epilepsy surgery. More recently, • To monitor cerebral electrical activity dur-
scalp electroencephalographic (EEG) recording carotid endarterectomy and cardiac
ings (with or without special computer process- bypass surgery.
ing) have been used routinely in many medical
centers to monitor cerebral electrical activ-
ity during endarterectomy1 and cardiac bypass TECHNICAL FACTORS IN
operations.2 EEG monitoring during the latter INTRAOPERATIVE EEG
procedure is limited by the effect of hypother- MONITORING
mia, which often suppresses EEG activity,3
making it ineffective as a monitoring tool for To ensure reliable EEG recordings during
ischemia. EEG monitoring for cardiac surgery intraoperative monitoring, it is important to
is further limited because easily correctable pay special attention to technical factors:4
727
Figure 42–1. EEG with isoflurane anesthesia. Left, EEG recorded with routine paper speed (30 mm/second). Common
patterns of anesthetic—ARF alpha pattern, ATS pattern, and PAS pattern—are seen in the left hemisphere. In the right
hemisphere, there is an increase in the amount of irregular slowing, especially in temporal distribution, and reduction in
the normal ARF alpha pattern. Right, The same changes can be appreciated qualitatively even at a reduced paper speed of
5 mm/second (one-sixth the usual paper speed). Note that this patient has a focal abnormality while under anesthesia, even
though the patient had had only a right cerebral transient ischemic attack (TIA) without residual deficit. (From Daube,
J. R., C. M. Harper, W. J. Litchy, and F. W. Sharbrough. 1990. Intraoperative monitoring. In Current practice of clinical
electroencephalography, ed. D. D. Daly, and T. A. Pedley, 2nd ed., 743. New York: Raven Press. By permission of Mayo
Foundation for Medical Education and Research.)
application of electrodes with collodion for by altering the timescale in digital EEG
stability, an adequate number of electrodes (at recordings; for example, slowing the paper
least 8 and preferably 16 channels), filtering of speed initially to a rate of 15 mm/second, which
background noise with a 60-Hz notch filter, use is half the speed used in standard EEG record-
of adequate sensitivity, and proper grounding ings8 (Figs. 42–1 to 42–4). Even a longer
for patient safety. The timescale used to display timescale display, corresponding to a paper
EEG data is a particularly important techni- speed of 5 mm/second (one-sixth of that used
cal factor, because a large amount of data are in standard recordings), is adequate for detect-
generated during intraoperative EEG monitoring important EEG changes of ischemia dur-
ing, and it must be compressed without essen- ing intraoperative clamping (Fig. 42–5). If
tial data being lost. Computer processing with unusual EEG changes are suspected while a
spectral analysis presented as a compressed slower timescale is being used, returning to
spectral array has been used for this purpose.5 the regular timescale or paper speed of 30
There are reports that suggest that computer- mm/second will permit prompt identification
ized quantitative analysis of EEG may com- of abnormal patterns (Fig. 42–1). Moreover,
plement visual analysis in detecting the EEG EEG changes may be detected early by dis-
changes.6, 7 However, data can be compressed playing both ongoing EEG and baseline EEG
visually without using spectral analysis simply using a split-screen view for visual comparison.
Cerebral Function Monitoring 729
Figure 42–2. The effect of painful stimulation (skin incision) on the EEG during levels of anesthesia below minimal
alveolar concentration; such stimulation tends to reduce the amount of slow activity and to accentuate the amount of fast
activity seen with a given concentration of anesthetic agent. Paper speed is 15 mm/second (half the usual speed). (From
Daube, J. R., C. M. Harper, W. J. Litchy, and F. W. Sharbrough. 1990. Intraoperative monitoring. In Current practice of
clinical electroencephalography, ed. D. D. Daly, and T. A. Pedley, 2nd ed., 742. New York: Raven Press. By permission of
Mayo Foundation for Medical Education and Research.)
Therefore, it is essential that EEG recording simple commands for checking motor strength.
commence before anesthetics are administered, If the electrophysiologist is not present in the
and recorded subsequently on a continuous operating room, technical provisions should
basis. At our institution, the EEG monitoring is be made to permit reliable and immediate
continued until the patient has recovered suf- review of the digital recording at a site readily
ficiently from anesthetic effect to respond to accessible to the electrophysiologist.
Figure 42–3. A–C, EEG of a 54-year-old man undergoing left carotid endarterectomy. The EEG result demonstrates
the type of reduction in the faster anesthetic components and the retention of rhythmic slowing that occurs with less
severe decrease in cerebral blood flow (CBF) below the critical level. This change is easy to identify despite significant
artifact affecting the posterior electrodes. Paper speed is 15 mm/second (half the usual paper speed). (From Daube, J. R.,
C. M. Harper, W. J. Litchy, and F. W. Sharbrough. 1990. Intraoperative monitoring. In Current practice of clinical electroen-
cephalography, ed. D. D. Daly, and T. A. Pedley, 2nd ed., 742. New York: Raven Press. By permission of Mayo Foundation
for Medical Education and Research.)
Figure 42–4. A–C, EEG of a 74-year-old man undergoing right carotid endarterectomy. The EEG in B demonstrates
the more severe type of EEG “suppression” that occurs in association with a more severe decrease in cerebral blood
flow below the critical level. Paper speed is 15 mm/second (half the usual paper speed). (From Daube, J. R., C. M. Harper,
W. J. Litchy, and F. W. Sharbrough. 1990. Intraoperative monitoring. In Current practice of clinical electroencephalography,
ed. D. D. Daly, and T. A. Pedley, 2nd ed., 743. New York: Raven Press. By permission of Mayo Foundation for Medical
Education and Research.)
Figure 42–5. EEG of a 68-year-old man undergoing left carotid endarterectomy. The EEG shows the type of dramatic,
rapid attenuation of all components often associated with cerebral blood flow of less than 6 or 7 mL/100 g per minute.
Paper speed is 5 mm/second (one-sixth the usual paper speed). (From Daube, J. R., C. M. Harper, W. J. Litchy, and
F. W. Sharbrough. 1990. Intraoperative monitoring. In Current practice of clinical electroencephalography, ed. D. D. Daly,
and T. A. Pedley, 2nd ed., 744. New York: Raven Press. By permission of Mayo Foundation for Medical Education and
Research.)
730
Key Points in amplitude, and slows in frequency toward

the alpha range.10 During this phase, the faster
• EEG data during carotid endarterectomy
rhythm may be intermixed with bursts of high-
must be compressed and displayed at
amplitude, frontal-intermittent, rhythmic delta
settings that will permit review of large activity, the FIRDA pattern. With a slower
amount of data.
• If the electrophysiologist is not present in rate of induction with nonbarbiturate inhala-
tion agents, there is less tendency toward inter-
the operating room, technical provisions mittent rhythmic bursting.
should be made to permit reliable and
immediate review of the digital recording Light steady-state anesthesia. When a light
at a site readily accessible to the electro- level of steady-state anesthesia is reached, a
physiologist. characteristic anterior maximum, rhythmic fast
• EEG recording should commence before (ARF) pattern, usually in the lower beta- or
anesthetics are administered, and recorded alpha-frequency range, is seen. This pattern is
subsequently on a continuous basis. present with most agents, including halothane,
• EEG monitoring should be continued enflurane, isoflurane, and even thiopental (see
until the patient has recovered suffi- Fig. 42–1). The frequency of this pattern tends
ciently from anesthetic effect to respond to slow with increasing concentrations of the
to simple commands for checking motor agent. This anesthetic ARF pattern is strik-
strength. ingly similar to postanoxic, widespread, ante-
rior maximum alpha-coma pattern. The origin
of the postanoxic ARF pattern is uncertain.
EFFECTS OF ANESTHESIA ON However, evidence suggests that the anesthetic
ELECTROENCEPHALOGRAPHY ARF pattern, usually in the alpha-frequency
range, is a drug-induced beta-variant pat-
SYMMETRICAL EEG PATTERNS tern,10 which with anesthesia becomes more
DURING ANESTHESIA widespread, higher in amplitude, and slower in
frequency—from the upper to the lower beta
Although much has been written about the range and finally into the alpha or even theta
effects of different anesthetic agents on the range. In humans, the anesthetic ARF pattern
EEG during anesthesia, most of these agents tends to be more continuous than normal sleep
produce similar EEG patterns when used spindles.
at concentrations below their minimal alve- In addition to the alpha-frequency ARF
olar concentration, which is the level nec- anesthetic pattern, there are often anterior
essary to prevent movement in response to maximum, triangular, slow (ATS) waves, which
a painful stimulus in approximately 50% of commonly are diphasic and have a sharply con-
patients. Common anesthetic agents, includ- toured, initially negative phase that is followed
ing thiopental, halothane, enflurane, isoflu- by a more rounded positive phase (Fig. 42–1).
rane, and nitrous oxide, produce certain EEG These ATS waves characteristically have a
changes that are related to the degree and duration of less than 1 second and may occur
timing of anesthesia. either as a single transient or in brief trains.
Subanesthetic Concentrations. Subanesthetic When preceded by a waxing ARF alpha pat-
concentrations characteristically produce max- tern, they may produce a “mitten-like” pat-
imal beta activity in the anterior midline. This tern.9 In addition to the ATS waves, more pos-
pattern is most commonly seen with thiopental terior, arrhythmic, slow (PAS) waves, which are
but may occur less prominently with halothane, often of lower amplitude, may become promi-
enflurane, isoflurane, and 50% nitrous oxide nent. Duration of these individual polymorphic
activity.9 slow waves is usually longer than 1 second.
The maximum of this activity is often diffi-
Rapid Induction. Rapid surgical inductions cult to identify clearly, but it is always more
performed with thiopental are accompanied posterior and may be more prominent over
by the following characteristic sequence of the temporal regions (Fig. 42–1). This pattern
changes: the subanesthetic pattern of beta is least obvious with halothane, more obvi-
activity rapidly becomes widespread, increases ous with enflurane, and most prominent with
isoflurane. Strictly, nitrous oxide alone is not that does not affect the normal symmetrical
potent enough to be an anesthetic agent at alpha pattern. Despite such a normal sym-
atmospheric pressure; nonetheless, it poten- metrical alpha pattern, the anesthetic record
tiates the effects of other inhalation agents. may show a major decrease in the ARF pat-
During “balanced anesthesia” with 50% nitrous tern and an increase in the irregular slowing in
oxide in combination with another agent, the the anterior distribution. In these patients, the
PAS pattern tends to be more prominent than drug-induced beta activity seen during the pre-
when only a single agent is used. anesthetic state is nearly always decreased on
the side of a reduced anesthetic ARF pattern,
Key Points which is further evidence that the anesthetic
• Most of anesthetic agents produce similar alpha-frequency ARF pattern is related more
directly to drug-induced beta activity than to
EEG patterns when used at concentra-
the normal waking alpha rhythm.
tions below their minimal alveolar concen-
This focal abnormality seen with anesthesia
tration.
• Recognizable EEG patterns during gen- consists of a unilateral decrease in the ampli-
tude of the ARF anesthetic pattern by more
eral anesthestic effect include ARF, ATS,
than 30%–40%. This commonly is associated
and the PAS.
• Subanesthetic concentrations characteris- with increased wave length and amplitude of
persistent polymorphic slowing on the side of
tically produce maximal beta activity in the
the decreased amplitude. The latter generally
anterior midline.
• With rapid induction, the beta-frequency can be distinguished from the usual PAS pat-
tern during anesthesia because it has a longer
activity may be intermixed with bursts
wave length and is more irregular and higher in
of high-amplitude, frontal-intermittent,
amplitude on the pathologic side.
rhythmic delta activity, the FIRDA pat-
In most patients, focal baseline EEG abnor-
tern.
• With a slower rate of induction with non- mality correlates with preoperative deficits.
However, a small percentage of patients with
barbiturate inhalation agents, there is less
such baseline abnormalities have experienced
tendency toward intermittent rhythmic
only transient ischemic attacks, presumably
bursting.
• With light steady-state anesthesia, a char- caused by a hemodynamic mechanism. These
patients usually have normal findings on com-
acteristic ARF pattern is seen.
puted tomography, with an ipsilateral decrease
in retinal artery pressure and a low-baseline
cerebral blood flow thought sufficient to cause
PREOPERATIVE FOCAL EEG abnormalities in the absence of resid-
ABNORMALITIES SEEN WITH ual neurologic signs or computed tomographic
ANESTHESIA abnormalities6 (Fig. 42–1).
The anesthetic EEG also tends to activate an
Most patients undergoing endarterectomy abnormality when intermittent rhythmic slow
show a symmetrical baseline pattern of the waves are present in the temporal region on
type described above. However, depending on the side of ischemia. This intermittent abnor-
patient selection, 30%–40% of patients may mality is often converted into a more obvious
have a focal abnormality of varying sever- and persistent focal slowing along with reduc-
ity.11 In general, a preoperative focal anesthetic tion in the ARF pattern during anesthesia.
EEG abnormality correlates with a preoper- A more posterior lesion, which produces sig-
ative waking EEG abnormality. Occasionally, nificant abnormality in the alpha pattern, may
in exceptional patients, anesthesia activates an leave the anesthetic ARF pattern symmetri-
abnormality that was either not apparent or cal, without obvious focal slowing. In such a
less apparent during the waking trace.12 How- case, preanesthetic beta activity is also usually
ever, anesthesia may obscure an abnormality symmetrical.
that was present during the waking trace and The nonlocalizing FIRDA pattern or more
may also activate an abnormality that is mini- persistent generalized slowing is easily recog-
mal or not apparent during the waking state, nized as abnormal during the waking state,
as in the case of an anterior hemispheric insult but it cannot be identified as abnormal during
anesthesia because these patterns commonly Intraoperative monitoring techniques can be

occur in most patients at some stage of anes- divided into those that detect cerebral dys-
thesia, whether or not they have had symptoms function (EEG, somatosensory evoked poten-
of a central nervous system lesion. tials (SEPs), and neurologic examination) and
those that measure the integrity of cerebral
perfusion (carotid stump pressure, transcra-
Key Points nial Doppler, and cerebral blood flow mea-
• As many as 30%–40% of patients may surements). McCarthy and colleagues14 have
have a focal abnormality of varying sever- confirmed that when a major EEG change
ity on the preoperative baseline EEG. occurs during carotid cross-clamping, the risk
• Anesthesia may obscure an abnormality of stroke is higher than when there is no change
that was present during the waking trace (4.5% vs. 1.4%). Selective shunt placement
and may also activate an abnormality that in patients with major EEG changes signifi-
is minimal or not apparent during the cantly decreases the rate of stroke.15 The rate
waking state. of intraoperative stroke when no monitoring is
• If there is preoperative unilateral anterior used is six times higher than that when EEG
hemispheric insult, the anesthetic record monitoring is used to select patients for shunt
may show a major decrease in the ARF placement.16
pattern and an increase in the irregular
slowing in the region. Key Points
• Drug-induced beta activity seen during
the preanesthetic state is nearly always • The greatest susceptibility to cerebral
decreased on the side of a reduced anes- ischemic injury occurs when the carotid
thetic ARF pattern. artery is cross-clamped just before it is
• When intermittent rhythmic slow waves incised.
are present in the temporal region on the • When a major EEG change occurs during
side of ischemia, the intermittent slow carotid cross-clamping, the risk of stroke
waves are often converted by anesthesia is higher than when there is no change.
to a more obvious and persistent focal • Selective shunt placement in patients
slowing along with reduction in the ARF with major EEG changes significantly
pattern. decreases the rate of stroke.
• The rate of intraoperative stroke when
no monitoring is used is six times higher
than that when EEG monitoring is used
CLINICAL APPLICATIONS to select patients for shunt placement.
Intraoperative monitoring is useful only if it

can accurately and promptly forewarn surgeons
of the occurrence of cerebral complications EEG CHANGES DURING
in time for corrective measures to be insti- CAROTID ENDARTERECTOMY
tuted. The greatest susceptibility to cerebral
ischemic injury occurs when the carotid artery EEG changes commonly occur with carotid
is cross-clamped just before it is incised. To artery clamping—some change occurs in
avoid this potential complication, some sur- approximately 25% of patients. These changes
geons have routinely placed a shunt from the almost always occur within 20–30 seconds after
common carotid artery to the internal carotid clamping and are associated with decreased
artery to bypass the site of the clamp. However, cerebral blood flow below a critical level,
the rate of embolic stroke with routine place- which varies with the anesthetic agent. With
ment of a shunt is nearly 10 times greater than halothane, the critical level is between 15 and
that with selective use of a shunt.13 Thus, reli- 18 mL/100 g per minute; it is slightly lower
able and accurate techniques of intraoperative with enflurane and isoflurane.17 The reasons
monitoring are needed to select patients who for these differences are uncertain. Roughly,
require a shunt to decrease the risk of cerebral the severity and rapidity of the onset of EEG
ischemic injury during carotid cross-clamping. change vary in direct proportion to the degree
Figure 42–6. EEG of a 75-year-old man undergoing right carotid endarterectomy. Left side of the figure shows no asym-
metry at baseline following anesthesia administration but before carotid clamping. Right side of the figure shows right
hemisphere slowing when blood pressure dropped to 51/27 without carotid clamping.
of lowering of blood flow below the criti- reversible effects of embolization from the
cal level. Minor changes consist of a 25%– operative site. However, in approximately 1%
50% decrease in the faster components and of patients undergoing endarterectomy, a focal
an increase in amplitude and wave length of EEG change develops intraoperatively that
slower components (Fig. 42–3). Changes that is not associated with carotid artery clamp-
occur with more severe decreases in blood ing and persists throughout the procedure;
flow (in the range of 6 or 7 mL/100 g per ultimately, it is associated with a new neuro-
minute or less) are associated with an even logic deficit in the immediate postoperative
greater reduction of anesthetic faster compo- period. These changes almost always prove to
nents, along with decreased amplitude of the be caused by embolization. Another cause of
slower components, producing a lower ampli- EEG change during carotid endarterectomy is
tude and featureless EEG on the side of clamp- hypotension that occurs with or without brady-
ing (Figs. 42–4 and 42–5). Although up to 25% cardia (Fig. 42–6).
of EEGs may show some change, only 1%–3%
show the more severe degree of change.
Focal transient changes that occur at times Key Points
other than during clamping can be seen in up • EEG changes associated with decreased
to 10% of patients. In most, this is caused by cerebral blood flow below a critical level
transient asymmetrical effects of changing lev- almost always occur within 20–30 seconds
els of anesthesia on a preexisting focal abnor- after clamping.
mality and is of no consequence.17 Emboliza- • Although up to 25% of EEGs may show
tion can occur in association with shunting. some change, only 1%–3% show the more
Some EEG changes likely are caused by the severe degree of change.
• Severe EEG change consists of marked neurological deficits during carotid clamping,
reduction of theta- and alpha-frequency but only 19 (60%) had ischemic EEG changes.
waves, along with markedly decreased None of the 314 patients developed a postop-
amplitude of delta waves, producing a low erative stroke.22
amplitude and featureless EEG on the A method of monitoring the integrity of
side of clamping (Figs. 42–4 and 42–5). vascular flow during carotid surgery is intrac-
• Focal transient changes that occur at arotid injection of xenon.4 The determination
times other than during clamping can be of cerebral blood flow with this technique
seen in up to 10% of patients. In most, is clearly influenced by the Paco2 level and
this is caused by transient asymmetrical requires that xenon be delivered directly to
effects of changing levels of anesthesia on the internal carotid artery while the external
a preexisting focal abnormality and is of no carotid artery is clamped. Blood flow is usually
consequence. determined only three or four times intraop-
eratively: before clamping, immediately after
clamping, immediately after placing a shunt
SEP RECORDING DURING (if one is used), and at the end of the proce-
CAROTID ENDARTERECTOMY dure. Blood flow measurement is complemen-
tary to EEG findings. Focal EEG changes as a
Intraoperative changes in an SEP recording result of ischemia caused by decreased perfu-
are measured and quantified more easily than sion pressure distal to a clamped carotid artery
EEG changes. Moreover, SEPs are less likely are always associated with low blood flow, as
to be influenced by factors such as anes- measured with the xenon technique. An unex-
thetic effects, preoperative cerebral abnormal- pected EEG change that persists and is asso-
ities, and recording artifacts. Nonetheless, a ciated with “normal blood flow” is essentially
major drawback is that, after carotid cross- pathognomonic of embolization. The presence
clamping, SEP changes are not detected as of normal blood flow after embolization is
promptly as EEG changes, because it often explained on the basis of the so-called look-
takes a minute or longer to complete the aver- through phenomenon.23 A simplified interpre-
aging and analysis of SEP signals,18 but nearly tation of this phenomenon is as follows: if
all EEG changes occur within 20–30 seconds the ischemia is a result of embolic occlu-
after cross-clamping. A study that compared sion of one-half of the blood vessels to a
the two techniques concluded that they are region (with the other one-half being patent),
complementary.19 injection of xenon produces a normal or,
at times, an increased flow and washout of
xenon through the patent blood vessels. Totally
occluded vessels receive no xenon and thus
OTHER MONITORING do not contribute to the overall measurement
TECHNIQUES DURING of flow.
CAROTID ENDARTERECTOMY Carotid stump pressure determination is a
measurement of the back pressure of flow at
Intraoperative neurologic examination has the distal carotid stump after cross-clamping.
been advocated as a method of determining Although low carotid stump pressure values
intolerance to carotid cross-clamping. How- are more likely to be associated with clin-
ever, the method cannot be used in nearly ically important EEG changes24 and higher
25% of the patients who prefer or require stroke rates,14 the specificity of carotid stump
general anesthesia.20 Although intraoperative pressure measurement is only approximately
neurologic examination has not been shown to 60%–80%.14, 15, 24 Also, carotid stump pressure
be superior to EEG monitoring in decreas- measurement is not as reliable as xenon blood
ing stroke rate, the combination of the two flow measurement.25 Studies that compared
techniques reportedly can reduce the need EEG with carotid stump pressure monitoring
for shunt placement.21 In a retrospective suggested that EEG monitoring is more accu-
study of 314 patients who were awake during rate,15 and the use of carotid stump pressure
carotid endarterectomy and were monitored measurement alone may result in a high rate of
by neurological examination, 32 developed unnecessary shunts.24
Transcranial Doppler (TCD) insonates the EEG MONITORING DURING

temporal bone window to measure blood flow CARDIAC SURGERY
velocity at the middle cerebral artery ipsilateral
to the carotid cross-clamping. A review of the EEG monitoring has not been used widely
method concluded that it complements EEG during cardiac surgery because the profound
monitoring, probably because the two tech- hypothermia that is induced intraoperatively
niques measure different potential effects of suppresses most of the EEG activities needed
carotid cross-clamping.26 An advantage of TCD for monitoring. Diffuse EEG changes are dif-
is that it can detect abnormalities of intracra- ficult to distinguish from global ischemia even
nial emboli from carotid debris and hyperper- when the hypothermia is mild or when Paco2
fusion patterns after carotid repair.27 TCD has is high. To address the limitations of EEG
also been reported to be useful in identifying monitoring in cardiac surgery, a recent review
patients who are at risk of developing neurolog- proposed that EEG monitoring be used con-
ical symptoms during the postsurgical recovery comitantly with TCD and NIRS.2
period.28 However, a study that compared both
carotid stump pressure and TCD with neu-
rologic monitoring showed that neither could BISPECTRAL ANALYSIS OF EEG
reliably predict the need for shunt. The pos- FOR MONITORING DEPTH OF
itive predictive value for either was relatively ANESTHESIA
low and could result in unnecessary use of a
shunt.29 EEG monitoring had been attempted for guid-
Near-infrared spectrophotometry (NIRS) is ing depth of anesthetic sedation in order to
a noninvasive optical method for continu- avoid patient recall and awareness during and
ously measuring cerebral oxygenation. Its mea- immediately after surgery. However, the lim-
surement is relatively unaffected by anes- ited frequency and amplitude domain infor-
thestic effect, altered cerebral metabolism, or mation of unprocessed EEG is inadequate
markedly decreased perfusion pressure. Tech- for providing additional information regarding
niques have been developed to isolate the anesthetic depth. With the advent of readily
intracranial component from the extracranial available and powerful mathematical comput-
component of NIRS signals, but further inves- ing, EEG signals can now be processed with
tigations are needed to determine whether high-level statistics during surgery to yield a
NIRS could independently indicate the need single number index derived from multivari-
for shunt placement during carotid endarterec- ate analysis of several EEG signal character-
tomy.30, 31 istics. This derived index based on the most
commonly employed mathematical algorithm
Key Points is called the bispectral index (BIS).32 A num-
ber of BIS studies have yielded encourag-
• An unexpected EEG change that persists
ing but inconclusive results, until the recent
and is associated with “normal blood flow” randomized blinded study of BIS compared
is essentially pathognomonic of emboliza- against the standard of end-tidal anesthetic
tion. gas (ETAG) for decreasing anesthesia aware-
• The combination of intraoperative neu-
ness.33 Two of the 967 BIS patients and two of
rologic examination and EEG monitoring the 974 ETAG patients in this study reported
can reportedly reduce the need for shunt anesthesia awareness. The study provided the
placement. strongest evidence that BIS does not con-
• Studies that compared EEG with carotid
tribute further to the current practice of deter-
stump pressure monitoring suggest that mining depth of anesthesia for avoiding anes-
EEG monitoring is more accurate and the thetic awareness.
use of carotid stump pressure measure-
ment alone may result in a high rate of
unnecessary shunt. EEG MONITORING FOR
• An advantage of TCD is that it can detect EPILEPSY SURGERY
abnormalities of intracranial emboli from
carotid debris and hyperperfusion pat- The principles and techniques in EEG mon-
terns after carotid repair. itoring for epilepsy surgery are different than
those for carotid endarterectomy or cardiac grams with cerebral blood flow measurements during
surgery (see Chapter 15). Whereas EEG mon- carotid endarterectomy. Stroke 4:674–83.
itoring during extracerebral surgeries on the 2. Edmonds, H. L. Jr., R. A. Rodriguez, S. M. Audenaert,
E. H. Austin III, S. B. Pollock Jr., and B. L. Ganzel.
cardiovascular system is mainly for the purpose 1996. The role of neuromonitoring in cardiovascu-
of detecting and avoiding impending adverse lar surgery. Journal of Cardiothoracic and Vascular
effects of these surgeries on the brain, EEG Anesthesia 10:15–23.
monitoring in epilepsy surgery has the primary 3. Quasha, A. L., F. W. Sharbrough, T. A. Schweller, and
J. H. Tinker. 1979. Hypothermia plus thiopental: Syn-
goal of guiding the location and maximizing the ergistic EEG suppression (abstract). Anesthesiology
extent of brain tissue resection while minimiz- 51(Suppl):S20.
ing the risk of invading cortical areas that are 4. Sundt, T. M. Jr., F. W. Sharbrough, J. C. Trautmann,
critical for functions such as language or motor and G. A. Gronert. 1975. Monitoring techniques
skills. for carotid endarterectomy. Clinical Neurosurgery
22:199–213.
5. Chiappa, K. H., S. R. Burke, and R. R. Young. 1979.
Results of electroencephalographic monitoring dur-
SUMMARY ing 367 carotid endarterectomies. Use of a dedicated
minicomputer. Stroke 10:381–8.
Intraoperative electrophysiologic monitoring 6. Visser, G. H., G. H. Wieneke, and A. C. van Huffelen.
1999. Carotid endarterectomy monitoring: Patterns of
of cerebral function during cardiovascular spectral EEG changes due to carotid artery clamping.
surgery requires a thorough knowledge of the Clinical Neurophysiology 110:286–94.
effect of anesthetic agents on electrophysi- 7. Minicucci, F., M. Cursi, C. Fornara, et al. 2000.
ologic signals. Although there are variations Computer-assisted EEG monitoring during carotid
among anesthetic agents and their effects on endarterectomy. Journal of Clinical Neurophysiology
17:101–7.
the EEG, most of the agents produce simi- 8. Yanagihara, T., and D. W. Klass. 1979. Discrepancy
lar changes that can be recognized and dis- between CT scan and EEG in hemodynamic stroke
tinguished from the effects of ischemia. Suc- of the carotid system. Transactions of the American
cess of the monitoring also depends heavily on Neurological Association 104:141–4.
9. Stockard, J., and R. Bickford. 1975. The neurophys-
the technical aspects of recording, such as the iology of anaesthesia. Monographs on Anesthesiology
timescale of the EEG display and adjustment 2:3–46.
of the anesthetic agent used. 10. Sharbrough, F. W. 1982. Nonspecific abnormal EEG
Currently, intraoperative EEG monitor- patterns. In Electroencephalography: Basic principles,
ing is used mostly during carotid surgery clinical applications and related fields, ed. E. Nie-
dermeyer, and F. Lopes da Silva, 135–54. Baltimore-
because of its favorable sensitivity and speci- Munich: Urban & Schwarzenberg.
ficity in promptly detecting cerebral intoler- 11. Sundt, T. M. Jr., F. W. Sharbrough, R. E. Anderson,
ance to carotid cross-clamping. Recent stud- and J. D. Michenfelder. 1974. Cerebral blood
ies have continued to demonstrate convinc- flow measurements and electroencephalograms dur-
ing carotid endarterectomy. Journal of Neurosurgery
ingly the usefulness of intraoperative monitor- 41:310–20.
ing in decreasing the risk of stroke in carotid 12. Hansotia, P. L., F. W. Sharbrough, and J. Beren-
endarterectomy. For the most part, the advent des. 1975. Activation of focal delta abnormality with
of non-EEG monitoring techniques has not methohexital and other anesthetic agents (abstract).
replaced EEG monitoring. Many recent stud- Electroencephalography and Clinical Neurophysiol-
ogy 38:554.
ies have shown that these non-EEG monitor- 13. Salvian, A. J., D. C. Taylor, Y. N. Hsiang, et al. 1997.
ing techniques complement EEG monitoring, Selective shunting with EEG monitoring is safer than
largely because the aspects of intraoperative routine shunting for carotid endarterectomy. Cardio-
cerebral hypoperfusion or ischemia that these vascular Surgery 5:481–5.
14. McCarthy, W. J., A. E. Park, E. Koushanpour, W. H.
tests measure are different from those that Pearce, and J. S. Yao. 1996. Carotid endarterectomy.
EEG measures. To date, studies that have com- Lessons from intraoperative monitoring—A decade of
pared the different modalities used in intra- experience. Annals of Surgery 224:297–305.
operative monitoring have lacked the scientific 15. Lacroix, H., G. Van Gertruyden, J. Van Hemelrijck,
rigor of randomized controlled studies. A. Nevelsteen, and R. Suy. 1996. The value of carotid
stump pressure and EEG monitoring in predicting
carotid cross-clamping intolerance. Acta Chirurgica
Belgica 96:269–72.
REFERENCES 16. Plestis, K. A., P. Loubser, E. M. Mizrahi, G. Kantis,
Z. D. Jiang, and J. F. Howell. 1997. Continu-
1. Sharbrough, F. W., J. M. Messick Jr., and T. M. Sundt ous electroencephalographic monitoring and selec-
Jr. 1973. Correlation of continuous electroencephalo- tive shunting reduces neurologic morbidity rates in
carotid endarterectomy. Journal of Vascular Surgery contralateral carotid artery: Another look at selec-
25:620–8. tive shunting. American Journal of Surgery 170:
17. Sundt, T. M. Jr., F. W. Sharbrough, D. G. Piepgras, 148–53.
T. P. Kearns, J. M. Messick Jr, and W. M. 25. McKay, R. D., T. M. Sundt, J. D. Michenfelder,
O’Fallon. 1981. Correlation of cerebral blood flow et al. 1976. Internal carotid artery stump pressure and
and electroencephalographic changes during carotid cerebral blood flow during carotid endarterectomy:
endarterectomy: With results of surgery and hemody- Modification by halothane, enflurane, and innovar.
namics of cerebral ischemia. Mayo Clinic Proceedings Anesthesiology 45:390–9.
56:533–43. 26. Ackerstaff, R., and F. Moll. 1998. Use of EEG and
18. Guerit, J. M., C. Witdoeckt, M. de Tourtchaninoff, TCD for assessment of brain function during opera-
et al. 1997. Somatosensory evoked potential mon- tions on carotid artery. In Peri-Operative monitoring
itoring in carotid surgery. I. Relationships between in carotid surgery, ed. S. Horsch, and K. Ktenidis,
qualitative SEP alterations and intraoperative events. 110–20. Darmstadt: Steinkopf Verlag.
Electroencephalography and Clinical Neurophysiol- 27. Fiori, L., G. Parenti, and F. Marconi. 1997. Combined
ogy 104:459–69. transcranial Doppler and electrophysiologic monitor-
19. Fiori, L., and G. Parenti. 1995. Electrophysiologi- ing for carotid endarterectomy. Journal of Neurosurgi-
cal monitoring for selective shunting during carotid cal Anesthesiology 9:11–6.
endarterectomy. Journal of Neurosurgical Anesthesi- 28. Nielsen, M. Y., H. H. Sillesen, L. G. Jorgensen, and
ology 7:168–73. T. V. Schroeder. 2002. The haemodynamic effect of
20. Stoughton, J., R. L. Nath, and W. M. Abbott. carotid endarterectomy. European Journal of Vascular
1998. Comparison of simultaneous electroencephalo- and Endovascular Surgery 24(1):53–8.
graphic and mental status monitoring during carotid 29. Belardi, P., G. Lucertini, and D. Ermirio. 2003.
endarterectomy with regional anesthesia. Journal of Stump pressure and transcranial Doppler for pre-
Vascular Surgery 28:1014–21. dicting shunting in carotid endarterectomy. Euro-
21. Fiiorani, P., E. Sbarigia, F. Speziale, et al. 1997. Gen- pean Journal of Vascular and Endovascular Surgery
eral anaesthesia versus cervical block and perioper- 25(2):164–7.
ative complications in carotid artery surgery. Euro- 30. Al-Rawi, P. G., P. Smielewski, and P. J. Kirk-
pean Journal of Vascular and Endovascular Surgery patrick. 2001. Evaluation of a near-infrared spec-
13:37–42. trometer (NIRO 300) for the detection of intracra-
22. Hans, S., and O. Jareunpoon. 2007. Prospective nial oxygenation changes in the adult head. Stroke
evaluation of electroencephalography, carotid stump 32(11):2492–500.
pressure, and neurologic changes during 314 carotid 31. de Letter, J. A., H. T. Sie, B. M. Thomas, et al.
endarterectomies performed in awake patients. Jour- 1998. Near-infrared reflected spectroscopy and elec-
nal of Vascular Surgery 45:511–55. troencephalography during carotid endarterectomy—
23. Donley, R. F., T. M. Sundt, R. E. Anderson, and F. W. In search of a new shunt criterion. Neurological
Sharbrough. 1975. Blood flow measurements and Research 20(Suppl 1):S23–7.
the “look through” artifact in focal cerebral ischemia. 32. Bard, J. W. 2001. The BIS monitor: A review and
Stroke 6:121–31. technology assessment. AANA Journal 69(6):477–83.
24. Harada, R. N., A. J. Comerota, G. M. Good, H. 33. Avidan, M. S., L. Zhang, B. A. Burnside, et al. 2008.
A. Hashemi, and J. F. Hulihan. 1995. Stump Anesthesia awareness and the bispectral index. NEJM:
pressure, electroencephalographic changes, and the 1097–108.
Chapter 43
Brain Stem and Cranial Nerve

Monitoring
Brian A. Crum
INTRODUCTION Middle Cranial Fossa

METHODS Posterior Cranial Fossa
Electromyography Head and Neck Surgery
Nerve conduction studies
Evoked potentials SUMMARY
APPLICATIONS
INTRODUCTION during an operation when normal anatomy

is altered, making accurate identification of
Cranial nerves can be injured during surgi- nerves difficult. Finally, information from
cal procedures performed in the middle and intraoperative cranial nerve monitoring may
posterior cranial fossae as well as in the head lead to an altered surgical plan to preserve
and neck region. Damage results from com- neurological function at a time when clinical
pression, stretch, abrasion, or ischemia of the assessment is not possible.
nerve. If axonal disruption occurs, recovery is
limited resulting in significant clinical deficits Purpose and Role of Brain Stem
(i.e., facial paralysis or deafness). Cranial nerve and Cranial Nerve
function can be monitored during anesthesia Monitoring
by recording spontaneous or stimulus-evoked
electrical activity directly from the nerve or • Utilize electrophysiological studies, for exa-
the cranial muscles. Activity in other pathways mple, BAEP, SEP, EMG, and CMAP to
in the brain stem can be monitored by fol- monitor neural structures during surgery.
lowing changes in evoked potentials of various • Determine neural integrity and potential
sensory and motor pathways. These methods damage at a time when clinical assessment
can detect damage to either the intra-axial is not possible.
or the extra-axial portion of cranial nerves • Assist surgical decisions based on pres-
(Table 43–1). This monitoring is also useful ence, absence, or quality of electrophysio-
for real-time localization of cranial nerves logical signals.
739
Table 43–1 Cranial Nerve Modality Used for Monitoring

II Flash VEPs
III, IV, VI Extraocular muscle needle EMG
V Masseter EMG, NAP from first division
VII Facial muscle EMG, CMAP, Lateral Spread Response
VIII BAEPs, NAP
IX Soft palate EMG
X Cricothryoid and vocalis EMG
XI Trapezius EMG
XII Hypoglossal EMG
METHODS appearance.1 Neurotonic discharges provide

surgeons immediate feedback about nerve
Electromyography location and potential injury to nerves in the
surgical field. Neurotonic discharges can be
Special small-diameter flexible wire electrodes recorded in situations that require neuromus-
can be used to record motor unit activity cular blockade by titrating the dose of the neu-
from within muscles. These electrodes are romuscular blocking agent such that a motor
less traumatic to local tissue and are more response is obtained with an amplitude of at
easily secured than the standard monopolar least 25% of the baseline amplitude.2, 3
or concentric needle electrodes. Currently, Neurotonic discharges have various forms,
subcutaneously placed EEG electrodes are but all of them consist of MUPs that fire
utilized in our laboratory for muscle record- in a rapid and irregular manner (Fig. 43–1).
ings. Electrodes placed in muscle record var- The bursts, or trains, of MUPs fire at fre-
ious spontaneous and stimulus-evoked activ- quencies of 30–100 Hz. At times, multiple
ity arising from individual muscle fibers or discharges firing asynchronously and indepen-
motor units. Movement and electric artifacts, dently are recorded from a single muscle. Neu-
fibrillation and fasciculation potentials, and rotonic discharges are often precipitated by
random motor unit potential (MUP) activ- mechanical stimulation of the axonal mem-
ity related to inadequate anesthesia are reg- brane of peripheral nerves. They are sensi-
ularly recorded from muscle intraoperatively. tive indicators of nerve irritation and occur
Electric stimulation of the innervating nerve in virtually all monitored patients.4, 5 Sharp
produces a response by activation of MUPs, transection of a nerve may not produce neu-
and those in the area immediately surround- rotonic discharges,6 and damaged nerves are
ing the electrode are recorded. Mechanical less likely to produce neurotonic discharges
irritation (abrasion, stretch, or compression), than healthy nerves. In addition, irrigation
saline irrigation, and ischemia produce high- of a nerve with saline frequently produces
frequency bursts of MUPs, called neurotonic long trains of neurotonic discharges lasting
discharges that have a characteristic sound and 2–60 seconds. Therefore, the density and
Neurotonic discharges
Orbicularis oculi
Orbicularis oris
200 μV
100 ms
Figure 43–1. Neurotonic discharges recorded from facial nerve innervated muscles during posterior fossa surgery.
Brain Stem and Cranial Nerve Monitoring 741
frequency of neurotonic discharges recorded Nerve action potentials (NAPs) are recorded
during surgery correlates only roughly with the directly from mixed or sensory nerves in
severity of postoperative neurologic deficit.7 the surgical field or subcutaneously. Although
They do, however, correlate with activity of NAPs are lower in amplitude and more diffi-
close proximity to the nerve. cult to record than CMAPs, they may provide
Intraoperative electromyography (EMG) is useful information when sensory nerves are
performed with the same sweep speed and involved or when CMAPs cannot be recorded.
filter settings as standard diagnostic needle The amplitude of the CMAP or NAP is pro-
EMG. Sensitivities are 50–200 μV/division, fil- portional to the number of axons conducting
ter settings are 30–20,000 Hz, and sweep speed the response. Therefore, when a goal of mon-
is from 10 to 100 ms/division. Recordings itoring is to determine the number of intact
are possible from almost any cranial muscle, axons, the amplitude or area of the response
including extraocular and facial muscles, mus- can be measured and compared with values
cles of mastication and tongue, and pharyngeal recorded earlier intraoperatively or with pre-
and laryngeal muscles. The activity from multi- operative baseline measurements. The use of
ple muscles is often monitored simultaneously neuromuscular blocking agents will reduce the
with a multichannel recording instrument. As amplitude of the CMAP, but not the NAP.
with standard EMG, auditory signals are very The use of these agents must be kept in mind
important in the analysis of the origin and when using CMAP monitoring, though, this
relationship of the potentials to intraoperative type of monitoring can still be accomplished.
events. The IOM team must be aware of the degree of
blockade (measured as a reduction in CMAP
amplitude from the first shock to the fourth in
Key Points a train of stimuli) and any changes that occur
• EMG records motor unit action potentials during monitoring.
from muscle. Several different stimulators are used to acti-
• Neurotonic discharges are signals of pos- vate peripheral nerve axons intraoperatively.
sible nerve damage. Handheld stimulators of various sizes and con-
• Neurotonic discharges are short or long figurations that can be gas-sterilized are com-
bursts of rapidly firing motor units. mercially available. Stimulators that are insu-
• EMG and NCS assist the surgeon in local- lated to the very tip of the electrode have
izing a nerve in the operative field. fewer problems with current shunting, but they
may also produce subthreshold stimuli if they
are not applied properly to the surface of the
nerve. Other stimulators have a hooked config-
Nerve conduction studies uration that allows a nerve or fascicles within
a nerve to be separated from surrounding tis-
Two types of nerve conduction studies (NCS) sue. This reduces artifact from the stimulus
can be performed on cranial nerves intra- or surrounding muscles and allows the nerve
operatively. Compound muscle action poten- elements of interest to be stimulated selec-
tials (CMAPs) represent activity in motor tively. Bipolar stimulators have the cathode
axons and muscle fibers. Whenever possible, and anode attached to the same handle and
CMAPs recorded from the skin surface over- within several centimeters of each other. This
lying the motor point are used because they provides a localized stimulus that reduces the
give more quantifiable information about the risk of current spread to adjacent nerves. The
total number of functioning motor axons in disadvantage of the bipolar stimulator is that
the nerve than CMAPs recorded from intra- activation may be inadequate if the nerve is
muscular electrodes. The optimal surface elec- distant or there is too much fluid in the sur-
trodes are 5-mm disks (similar to those used gical field. Monopolar stimulators use a single
in routine NCS) that are applied firmly to the handheld cathode placed on the nerve, with
skin with collodion or tape. CMAP recordings a separate anode placed some distance away,
are made with filter settings of 2 Hz–20 kHz, usually a needle in the edge of the surgical
sweep speeds of 1–10 ms/division, and sensitiv- field or a distant surface electrode. Monopolar
ities of 100 μV–5 mV/division. stimulation reduces the chance of inadequate
stimulation, but increases the likelihood of AUDITORY EVOKED POTENTIALS

current spread to other nerves and shock
Auditory evoked potentials reflect electric
artifact.
activity in the auditory nerve and brain stem
The size of the stimulating electrodes
varies depending on the nerve stimulated. in response to cochlear stimulation. Brain
stem auditory evoked potentials (BAEPs) are
Small cranial nerves require stimulator tips
as small as 1 mm, and larger peripheral recorded from surface electrodes placed either
nerves may require 2–3 mm electrodes to pro- in the external ear canal or on the ear pinna and
referenced to the vertex. Five distinct wave-
vide adequate stimuli. If a surgical forceps
is modified for use as a bipolar stimula- forms can be recorded: wave I originates from
tor, the surgeon can dissect tissue with the the lateral auditory nerve and wave II is from
the medial auditory nerve. Wave III comes
stimulator.
from depolarization in the cochlear nucleus
and waves IV and V come from bilateral activa-
Key Points tion in the lateral lemniscus and inferior olive.
The electrocochleogram can be recorded from
• NCS can be performed by recording
a needle electrode placed through the tym-
CMAPs over muscle or by recording panic membrane into the wall of the middle ear
direct NAPs over nerve. cavity. The first major negative waveform (N1)
• CMAPs can provide quantifiable informa-
of the electrocochleogram represents activity
tion about the status of motor axons. in the lateral portion of the auditory nerve and
• Direct nerve stimulation and selection of
is analogous to wave I of the BAEP. Auditory
equipment pose technical challenges that NAPs can also be recorded from a small cotton-
depend on several factors, for example, wick electrode placed directly on the nerve in
size and depth of the nerve, and amount the cerebellopontine angle at surgery. Auditory
of nerve exposed. evoked potentials are not affected significantly
by general anesthetics.
BAEPs appear to be the best technique for
most surgical procedures, because they pose
Evoked potentials fewer technical problems, can be used dur-
ing the entire surgical procedure, monitor the
VISUAL EVOKED POTENTIALS entire auditory system, and correlate well with
postoperative hearing status.9–11 The electro-
Visual evoked potentials (VEPs) reflect elec- cochleogram is better defined in patients with
tric activity in optic pathways in response to acoustic neuromas, but technical difficulties
photic stimulation of the retina. The most are common and only the lateral portion of the
reproducible response is recorded as a broad auditory nerve located in the auditory canal of
positivity over the occipital head region approx- the temporal bone is monitored. Recording of
imately 100 ms after the stimulus. This P100 auditory NAPs provides immediate feedback,
waveform represents electrical activity in the but technical problems are common and the
occipital cortex. It is well defined in all awake technique can be used only when the audi-
patients when a pattern-reversal illuminated tory nerve is exposed in the surgical field. If
stimulus is given to the retina. VEPs can also be they can be reliably recorded, they show better
recorded in anesthetized patients with a strobe preservation of hearing.11
flash stimulus applied through the closed eye- Sudden changes in the latency and ampli-
lid or by way of specially designed contact tude of BAEPs occur when the nerve is tran-
lenses. However, flash-evoked VEPs are greatly sected or avulsed or the internal auditory
attenuated by general anesthesia and, thus, artery is damaged leading to nerve ischemia.12
have limited use intraoperatively.8 Recovery after sudden loss of BAEPs is
unusual, but correlation of the change with
Key Points intraoperative events may help prevent hear-
ing loss in future cases. More commonly,
• VEPs are not used intraoperatively due to there are gradual changes in BAEPs reflect-
their susceptibility to anesthesia. ing excessive traction on or manipulation of
wave I wave V
1 cm = 0.1 μV
1 cm = 1.5 ms
Figure 43–2. BAEP recorded through surgery. The initial response is at the bottom and last at the top of the figure. Grad-
ual prolongation of waves I and V and the I–V interpeak latency all of which improve by the conclusion of the monitoring.
Sensitivity 0.1 μV/cm, time base 1.5 ms/cm.
the nerve and often resolve with adjustment extensive. Monitoring of motor evoked poten-
of retractors or modification of the surgical tials (MEPs) may enhance the sensitivity of
approach (Fig. 43–2). Cooling of the auditory brain stem monitoring by detecting early com-
nerve during surgical exposure will also lead promise of pyramidal tract neurons, though
to mild increase in the latencies of wave II to this is a minor risk in most surgeries Previously,
wave V. use of MEPs to monitor brain stem function
was limited by sensitivity to anesthetics.13 Cur-
Key Points rently, however, anesthesia regimens can be
used to eliminate this effect (intravenous nar-
• BAEPs are relatively simple and useful for cotics, benzodiazepines, fentanyl, ketamine).
monitoring peripheral and central audi- The main limitation of MEPs is the short dis-
tory pathways. tance between cranial nerves and their mus-
• BAEPs are not affected by general anes- cles to the cortex which introduces significant
thesia. stimulus artifact.
• Sudden loss of the BAEP suggests irre-
versible damage to the nerve, often by
ischemia in the internal auditory artery Key Points
distribution. • SEPs and MEPs can monitor motor and
• More commonly, the latency of wave V of sensory pathways during surgery in which
the BAEP is prolonged during surgery and the brain stem is at risk.
this improves by the end of surgery.
SOMATOSENSORY EVOKED APPLICATIONS

POTENTIALS AND MOTOR
EVOKED POTENTIALS
Middle Cranial Fossa
Median, ulnar, and tibial somatosensory evoked
potentials (SEPs) can be used to monitor activ- VEPs have been used to monitor the visual sys-
ity in the medial lemniscus in the brain stem. tem during removal of tumors in the region
Most often median SEPs are utilized. Changes of the pituitary and hypothalamus and during
may occur only when sensory pathways are operations on vascular lesions that may affect
involved or when damage to the brain stem is the optic nerves, chiasm, or tracts. Because of
the sensitivity to general anesthetics, variability (Fig. 43–3). An NAP can be recorded directly
in latency and amplitude of the response, and from the ophthalmic division of the trigeminal
the poor correlation with postoperative visual nerve with a small cotton-wick electrode.16
function, visual evoked responses are not a
reliable monitor of the function of the visual Key Points
pathway during surgery.8 • For middle cranial fossa surgery, IOM
Tumors and vascular lesions of the orbit, with EMG of oculomotor and abducens
sella, sphenoid or cavernous sinus, and petrous innervated muscles is most common.
portion of the temporal bone can damage cra-
nial nerves directly or distort normal anatom-
ical relationships, making it difficult to distin- Posterior Cranial Fossa
guish between normal and abnormal nervous
system structures. Types of cases that may The trigeminal, abducens, facial, auditory,
benefit from monitoring include surgery for vestibular, glossopharyngeal, vagus, spinal
tumors such as meningiomas, lymphomas, car- accessory, and hypoglossal nerves can be
cinomas, pituitary tumors, and vascular lesions injured during posterior fossa surgery. The
such as carotid or ophthalmic aneurysms. The abducens is monitored for resection of tumors
oculomotor, trochlear, and abducens nerves of the floor of the fourth ventricle. The trigem-
can be monitored with EMG or CMAPs.14 inal, facial, and auditory nerves are most at
Wire electrodes are placed in the extraocular risk during acoustic neuroma resection or
muscles after the patient is anesthetized. This microvascular decompression or neurectomy
is more easily accomplished in muscles sup- for trigeminal neuralgia, hemifacial spasm,
plied by the oculomotor and abducens nerves. or vertigo (Table 43–2). The brain stem
Neurotonic discharges are recorded in the may be compressed by cerebellopontine mass
appropriate muscle when its nerve is mechani- lesions larger than 3 cm in diameter. The
cally stimulated in the surgical field. The sur- facial nerve and the motor division of the
geon may also use a handheld stimulator to trigeminal nerve are monitored with elec-
identify selected cranial nerves by recording trodes placed in muscles of facial expres-
a CMAP in the appropriate target muscle15 sion and mastication, respectively. Mechani-
cal stimulation produces neurotonic discharges
in the respective muscles. Electric stimula-
tion in the operative field can be used to
identify the various nerves in the cerebel-
lopontine angle. The amplitude of CMAPs
recorded over the nasalis or mentalis mus-
cles correlates with the number of functioning
axons in the nerve18 (Fig. 43–4). Preserva-
tion of the facial CMAP at the end of the
operation when stimulating the proximal por-
Figure 43–3. CMAPs from extraocular muscles obtained
tion of the facial nerve, just after leaving the
by direct electric stimulation of the oculomotor nerve in brain stem, predicts good recovery of facial
the surgical field in the region of the cavernous sinus. nerve function within 1 year postoperatively.7
Table 43–2 Most Common Surgical Procedures and Their Monitoring Plans
Acoustic neuroma Facial and trigeminal muscle EMG, facial
CMAP, BAEPs
Hemifacial spasm Facial and trigeminal muscle EMG, Lateral
Spread Response
Trigeminal nerve microvascular Facial and trigeminal EMG, BAEPs
decompression
Parotid surgery Facial EMG
Thyroid or parathyroid surgery Laryngeal EMG
CMAP
Preop
surface
EMG 1 mv
1:10 10 ms
Frontalis
Orbicularis
Mentalis
1:50
Frontalis
Orbicularis
Mentalis
3:10
Frontalis
Orbicularis
200 μV
Mentalis
100 ms
Figure 43–4. Monitoring of EMG potentials and facial CMAPs intraoperatively for acoustic neuroma in a 55-year-old
woman. Examples of neurotonic discharges observed at various times intraoperatively and gradual loss of facial CMAPs
indicate iatrogenic injury of the facial nerve. (From Daube, J. R., and C. M. Harper. 1989. Surgical monitoring of cranial
and peripheral nerves. In Neuromonitoring in surgery, ed. J. E. Desmedt, 118. Amsterdam: Elsevier Science Publishers.
This monitoring can still be accomplished nerve injury, thereby improving future surgical
successfully with up to 50% neuromuscu- results. When brain stem compression is
lar blockade.18 Stimulus intensity required to present, monitoring SEPs or MEPs may also
obtain an evoked response from facial inner- be useful.
vated muscles can be predictive of outcome; The lateral spread response (LSR) is an elec-
a stimulus intensity of 0.05 mA or less pre- trophysiological finding in patients with hemi-
dicts normal facial nerve functioning postoper- facial spasm, and can be used in IOM for
atively.19 The ratio of the amplitude obtained this condition. The LSR is defined as being
with proximal facial nerve stimulation to that able to evoke a CMAP from muscles of one
obtained with distal facial nerve stimulation facial nerve branch by stimulation of a dif-
can also be used for prediction of facial nerve ferent branch (Fig. 43–7). This is tested by
functioning after acoustic neuroma surgery20 recording over orbicularis oculi, for exam-
(Fig. 43–5). ple, and stimulating the mandibular branch
BAEPs can be monitored simultaneously percutaneously. This is a common finding in
with EMG and CMAPs.10 Changes in BAEPs hemifacial spasm and the disappearance of the
correlate well with the postoperative level LSR during surgery is a strong predictor of
of hearing.10, 12, 21 Gradual changes are often elimination of the hemifacial spasm postop-
reversible by altering the surgical approach eratively (Fig. 43–8). The persistence of the
or by moving retractors.22, 23 Sudden loss of LSR,however, does not preclude a good result
BAEPs is usually irreversible12 (Fig. 43–6). from surgery as most of these patients still
Changes that correlate with postoperative experience resolution, just not at such a high
function help determine the mechanism of rate.24
A
MEN–P
4 ms
MEN–O 100 μV
B MEN–P
4 ms
MEN–O 500 μV
Figure 43–5. Facial CMAP at the conclusion of acoustic neuroma surgery. The proximal response (A) is approximately
10% of the distal response (B). Sensitivity 100 μV/division, time base 4 ms/division for (A). Sensitivity 500 μV/division, time
base 4 ms/division for (B). Patient had significant postoperative facial weakness.
Figure 43–6. Monitoring of BAEPs during operation for acoustic neuroma. The sudden loss of the response (at 1:52)
correlated with inadvertent coagulation of the internal auditory artery. (From Harper, C. M., and J. R. Daube. 1989. Surgical
monitoring with evoked potentials: The Mayo Clinic experience. In Neuromonitoring in surgery, ed. J. E. Desmedt, 281.
Amsterdam: Elsevier Science Publishers. By permission of the publisher.)
Surgeries of the jugular foramen, foramen hypoglossal nerve (with electrodes placed in
magnum, and clivus may also place the facial, the tongue), in addition to monitoring of the
auditory, glossopharyngeal, vagal, spinal acces- vagus and spinal accessory nerves, is useful dur-
sory, and hypoglossal nerves at risk. Facial ing operations to remove chordomas, menin-
nerve and auditory monitoring is as mentioned giomas, and other lesions in the region of the
above. Needle EMG electrodes can be placed clivus and foramen magnum.
in the soft palate for monitoring of glossopha-
ryngeal nerve function. The cricothyroid is
used for the external laryngeal nerve and the Key Points
vocalis for recurrent laryngeal nerve monitor- • Cranial nerves V and VII through XII can
ing. Electrodes in trapezius allow monitoring be helpful in monitoring of posterior fossa
of spinal accessory nerve. Electric stimulation surgery.
can be used to distinguish between rootlets • The most common procedures are for
of the glossopharyngeal and vagus nerves in acoustic neuroma resection and microvas-
patients undergoing neurectomy for glossopha- cular decompression for trigeminal
ryngeal neuralgia.25, 26 EMG monitoring of the neuralgia or hemifacial spasm.
Lateral spread response Lateral spread response

Stimulate zygomatic Record orbicularis oris
Orbicularis oculi
Vessel off
Vessel on
Mentalis
Vessel off
Stimulate L Mandibular branch
200 μV
4 ms
Orbicularis oculi
Figure 43–8. The LSR disappears when the blood vessel
is lifted off the facial nerve and returns when the vessel is
in contact with the facial nerve.
Mentalis
nerve that course through the gland. Each

Stimulate L Zygomatic branch branch can be monitored selectively by plac-
Figure 43–7. Stimulation of the mandibular branch ing wire electrodes in the frontalis (temporal
should only lead to an evoked response in mentalis and branch), orbicularis oculi (zygomatic), orbicu-
stimulation of the zygomatic branch should only activate laris oris (buccal), and mentalis (mandibular)
orbicularis oculi. In hemifacial spasm there this a “spread” muscles (Table 43–2). Mechanical irritation
of activation to muscles supplied by other branches of the
facial nerve when one branch is stimulated. of the nerve produces neurotonic discharges
in the target muscle. Electric stimulation in
the surgical field can locate and, therefore,
prevent damage to branches of the facial
nerve. EMG monitoring has been shown to
• The facial nerve is monitored by free- decrease the risk of facial nerve injury during
running EMG (watching for neurotonic parotidectomy.27, 28
discharges) and facial nerve CMAP ampli- The recurrent branch of the laryngeal nerve
tude or threshold. can be injured during carotid endarterec-
• BAEPs are used to improve hearing tomy,29 anterior cervical diskectomy,30 or resec-
preservation. tion of thyroid tumors.31 Wire electrodes
• The LSR can be helpful in hemifacial can be placed by direct laryngoscopy into
spasm surgery, though preservation of this the vocalis muscle for monitoring the recur-
response does not preclude an excellent rent laryngeal nerve. Monitoring has been
surgical outcome. shown to decrease the frequency of vocal
• For skull base surgery, EMG of appro- cord paralysis associated with these proce-
priate muscles helps the surgeon identify dures.28
specific cranial nerves and avoid damage The spinal accessory nerve is at risk dur-
to these nerves. ing neck dissections and lymph node biop-
sies. Needles can be placed into the trapezius
muscle for intraopertive monitoring, both for
Head and Neck Surgery neurotonic discharges and for assistance with
nerve localization. This has proven useful in
Operations for neoplasms of the parotid gland reducing the risk of injury to this nerve in neck
may injure one or more branches of the facial surgeries.32
Key Points 8. Raudzens, P. A. 1982. Intraoperative monitoring of

evoked potentials. Annals of New York Academy of
• EMG monitoring can help to protect Sciences 388:308–26.
the facial, laryngeal, and spinal accessory 9. Radtke, R. A., C. W. Erwin, R. Wilkins, and
nerves during extracranial head and neck M. Rozear. 1987. Intraoperative brain stem audi-
tory evoked potentials (BAEPs): Significant decrease
surgery. in postoperative auditory deficit (abstract). Neurology
37(Suppl 1):219.
10. Harper, C. M., S. G. Harner, D. H. Slavit, et al.
1992. Effect of BAEP monitoring on hearing preser-
vation during acoustic neuroma resection. Neurology
SUMMARY 42:1551–3.
11. Jackson, L. E., and J. B. Roberson Jr. 2000. Acoustic
Various modalities are available for monitor- neuroma surgery: Use of cochlear nerve action poten-
ing the function of the cranial nerves and tial monitoring for hearing preservation. The American
brain stem during intracranial or extracra- Journal of Otology 21:249–59.
12. Neu, M., C. Strauss, J. Romstock, B. Bischoff, and
nial head and neck operations. After consid- R. Fahlbusch. 1999. The prognostic value of intraop-
eration of the surgical risks, a multimodal- erative BAEP patterns in acoustic neuroma surgery.
ity approach can be tailored to the needs of Clinical Neurophysiology 110:1935–41.
each patient. Close communication between 13. Zentner, J. 1989. Noninvasive motor evoked poten-
tial monitoring during neurosurgical operations on the
the IOM team and the surgical team is vital in spinal cord. Neurosurgery 24:709–12.
order to obtain appropriate electrophysiologi- 14. Schlake, H. P., R. Goldbrunner, C. Milewski, et al.
cal information and provide useful feedback at 1999. Technical developments in intra-operative mon-
a time when clinical detection of nerve injury is itoring for the preservation of cranial motor nerves and
impossible. IOM has been shown to decrease hearing in skull base surgery. Neurological Research
21:11–24.
the incidence of cranial nerve injury during 15. Sekiya, T., T. Hatayama, N. Shimamura, and S. Suzuki.
posterior fossa surgery. 2000. Intraoperative electrophysiological monitor-
ing of oculomotor nuclei and their intramedullary
tracts during midbrain tumor surgery. Neurosurgery
47:1170–6.
16. Stechison, M. T., A. Moller, and T. J. Lovely. 1996.
REFERENCES Intraoperative mapping of the trigeminal nerve root:
Technique and application in the surgical management
1. Harner, S. G., J. R. Daube, and M. J. Ebersold. 1986. of facial pain. Neurosurgery 38:76–81.
Electrophysiologic monitoring of facial nerve during 17. Goldbrunner, R. H., H. P. Schlake, C. Milewski, J. C.
temporal bone surgery. Laryngoscope 96:65–9. Tonn, J. Helms, and K. Roosen. 2000. Quantitative
2. Blair, E. A., E. Teeple Jr., R. M. Sutherland, T. Shih, parameters of intraoperative electromyography pre-
and D. Chen. 1994. Effect of neuromuscular blockade dict facial nerve outcomes for vestibular schwannoma
on facial nerve monitoring. The American Journal of surgery. Neurosurgery 46:1140–6.
Otology 15:161–7. 18. Lennon, R. L., M. P. Hosking, J. R. Daube, and J. O.
3. Brauer, M., D. Knuettgen, R. Quester, and M. Doehn. Welna. 1992. Effect of partial neuromuscular blockade
1996. Electromyographic facial nerve monitoring dur- on intraoperative electromyography in patients under-
ing resection for acoustic neurinoma under moderate going resection of acoustic neuromas. Anesthesia and
to profound levels of peripheral neuromuscular block- Analgesia 75(5):729–33.
ade. European Journal of Anaesthesiology 13:612–15. 19. Neff, B. A., J. Ting, S. L. Dickinson, and D. B. Welling.
4. Romstock, J., C. Strauss, and R. Fahlbusch. 2000. 2005. Facial nerve monitoring parameters as a pre-
Continuous electromyography monitoring of motor dictor of postoperative facial nerve outcomes after
cranial nerves during cerebellopontine angle surgery. vestibular schwannoma resection. Otology & Neuro-
Journal of Neurosurgery 93:586–93. tology 26(4):728–32.
5. Kombos, T., O. Suess, B. C. Kern, T. Funk, T. Pietila, 20. Isaacson, B., P. R. Kileny, H. El-Kashlan, and A. K.
and M. Brock. 2000. Can continuous intraoperative Gadre. 2003. Intraoperative monitoring and facial
facial electromyography predict facial nerve function nerve outcomes after vestibular schwannoma resec-
following cerebellopontine angle surgery? Neurologia tion. Otology & Neurotology 24(5):812–17.
Medico-Chirurgica (Tokyo) 40:501–5. 21. Wazen, J. J. 1994. Intraoperative monitoring of audi-
6. Nelson, K. R., and H. C. Vasconez. 1995. Nerve tran- tory function: Experimental observations and new
section without neurotonic discharges during intra- applications. Laryngoscope 104:446–55.
operative electromyographic monitoring. Muscle & 22. Hatayama, T., and A. R. Moller. 1998. Correlation
Nerve 18: 236–8. between latency and amplitude of peak V in the
7. Harner, S. G., J. R. Daube, M. J. Ebersold, and brain stem auditory evoked potentials: Intraopera-
C. W. Beatty. 1987. Improved preservation of facial tive recordings in microvascular decompression oper-
nerve function with use of electrical monitoring during ations. Acta Neurochirurgica 140:681–7.
removal of acoustic neuromas. Mayo Clinic Proceed- 23. Matthies, C., and M. Samii. 1997. Management
ings 62:92–102. of vestibular schwannomas (acoustic neuromas): The
value of neurophysiology for intraoperative monitor- 28. Brennan, J., E. J. Moore, and K. J. Shuler. 2001.
ing of auditory function in 200 cases. Neurosurgery Prospective analysis of the efficacy of continuous
40:459–66. intraoperative nerve monitoring during thyroidec-
24. Kong, D. S., K. Park, B. G. Shin, J. A. Lee, and tomy, parathyroidectomy, and parotidectomy. Oto-
D. O. Eum. 2007. Prognostic value of the lateral laryngology and Head and Neck Surgery 124:
spread response for intraoperative electromyography 537–43.
monitoring of the facial musculature during microvas- 29. Markand, O. N., R. S. Dilley, S. S. Moorthy, and
cular decompression for hemifacial spasm. Journal of C. Warren Jr. 1984. Monitoring of somatosen-
Neurosurgery 106(3):384–7. sory evoked responses during carotid endarterectomy.
25. Taha, J. M., J. M. Tew Jr., R. W. Keith, and Archives of Neurology 41:375–8.
T. D. Payner. 1994. Intraoperative monitoring of the 30. Jellish, W. S., R. L. Jensen, D. E. Anderson, and J. F.
vagus nerve during intracranial glossopharyngeal and Shea. 1999. Intraoperative electromyographic assess-
upper vagal rhizotomy: Technical note. Neurosurgery ment of recurrent laryngeal nerve stress and pha-
35:775–7. ryngeal injury during anterior cervical spine surgery
26. Taha, J. M., and J. M. Tew Jr. 1995. Long-term with Caspar instrumentation. Journal of Neurosurgery
results of surgical treatment of idiopathic neuralgias of 91(Suppl 2):170–4.
the glossopharyngeal and vagal nerves. Neurosurgery 31. Timon, C. I., and M. Rafferty. 1999. Nerve moni-
36:926–30. toring in thyroid surgery: Is it worthwhile? Clinical
27. Terrell, J. E., P. R. Kileny, C. Yian, et al. 1997. Clinical Otolaryngology 24:487–90.
outcome of continuous facial nerve monitoring during 32. Midwinter, K., and D. Willatt. 2002. Accessory nerve
primary parotidectomy. Archives of Otolaryngology— monitoring and stimulation during neck surgery. The
Head and Neck Surgery 123:1081–7. Journal of Laryngology and Otology 116(4):272–4.
Chapter 44
Spinal Cord Monitoring

Jeffrey A. Strommen
INTRODUCTION ELECTROMYOGRAPHY AND

GENERAL PRINCIPLES OF NERVE CONDUCTION STUDIES
INTRAOPERATIVE MONITORING Recording Techniques
EQUIPMENT AND ELECTRICAL SAFETY Physiologic and Technical Effects
MONITORING METHODS— Applications and Interpretation of
SOMATOSENSORY EVOKED Findings
POTENTIALS
TYPES OF SPINAL SURGERIES
Stimulation Techniques
Primary Spine Disease
Cervical Spine Disease
Physiologic and Technical Issues
Thoracic Spine Disease and
Application and Interpretation of
Scoliosis Surgery
SEP Changes
Lumbosacral Spine Disease
MOTOR EVOKED POTENTIALS Primary Neural Disease
Stimulation Techniques Dorsal Rhizotomy
Recording Techniques Cauda Equina and Tethered Cord
Physiologic/Technical Considerations Vascular Diseases
Applications and Interpretation of
SUMMARY
MEP Changes
INTRODUCTION sufficiently early to allow prompt correction of

the cause. Selection of the appropriate modal-
The integrity of neural structures including the ities is customized to the patient dependent
brain stem, spinal cord, and peripheral nerves on the clinical status and the structures felt
can be effectively monitored with multimodal to be at potential risk. In spinal procedures,
techniques during a wide variety of surgeries. this generally involves monitoring the central
These techniques are utilized with the goal of somatosensory pathway with somatosensory
preserving function and preventing injury to evoked potentials (SEPs), corticospinal path-
vital neural structures at a time when clinical ways with motor evoked potentials (MEPs),
examination is not possible. The primary goal and the spinal root with electromyography
of intraoperative monitoring is to prevent new (EMG) or compound muscle action potentials
neurologic deficits by identifying impairment (CMAPs).
751
Neurophysiologic monitoring in the operat- remain intact under anesthesia from those that
ing room (OR) setting can be challenging given may be altered as a result of normal variations,
the significant electrical interference encoun- patient age, underlying disease, or other fac-
tered as well as the effect of physiologic vari- tors. Monitoring can thus provide reassurance
ables including temperature, anesthetics, and to the surgeon of intact neural function dur-
hypotension, and variable responses with pre- ing the course of an operation, allowing greater
existing damage to neurologic structures. This intervention than would have been contem-
chapter will review the techniques and appli- plated without monitoring.
cations of neurophysiologic monitoring of mul- Reversible alterations in recordings occur
tiple modalities during spinal surgery. These when a manipulation results in a nondestruc-
methods have evolved and will continue to tive change in neural function that can be
change with time and advancing technology in recognized by monitoring the function contin-
both monitoring and surgical methods. uously. Examples include irritation of neural
tissue or mild local compression. Early recog-
Purpose and Role of Intraoperative Spinal nition of these alterations allows surgeons to
Cord Monitoring modify their procedures to reduce the likeli-
• Prevent new neurologic deficits by iden- hood of a persistent deficit.
Destructive manipulation, such as when tis-
tifying impairment sufficiently early to
sue is severed, is also readily recognized, but
allow prompt correction of the cause.
• Monitor spinal cord sensory pathways not reversible. In these instances, the changes
may occur rapidly and irreversibly. Monitoring
with SEPs, corticospinal pathways with
provides immediate evidence not only of the
MEPs, and spinal nerves with EMG and
damage, but also of the severity of the damage.
CMAP.
Recognition of irreversible changes can also
be useful by teaching the surgeon about the
mechanism of injury and helping to predict the
GENERAL PRINCIPLES OF nature and severity of the postoperative deficit.
INTRAOPERATIVE Selective recordings can localize the dam-
MONITORING age within the neural structures at risk by
demonstrating which nerves or tracts are still
The ideal monitoring system is a mechanism functional and which are not. The neural struc-
that provides the surgeon rapid feedback of tures and their associated causative risks during
nerve or spinal cord function with reliable, spine surgery include the following:
easily interpreted data while not interfering
with the surgical procedure. If this feedback • Spinal cord—ischemia, slow compression,
is provided in a rapid and reliable fashion, the stretching, and direct trauma.
surgeon can take appropriate action to prevent • Nerve roots and spinal nerve—stretching,
or reverse the potential neurologic injury. For blunt trauma, pinching, and ischemia.
example, some situations such as compression • Cauda equina—stretching, blunt trauma,
of the spinal cord may be reflected by a grad- pinching, and ischemia.
ual or subtle change in the recorded potentials.
Since these changes are typically reversible Damage may occur at one or more levels of
and revert to baseline when the alteration is the spine; thus optimally, each level at risk
reversed, monitoring is best accomplished by should be monitored. The modalities and vari-
demonstrating normal function early in a pro- ables of monitoring change with the level. The
cedure and testing it repeatedly in search for level most commonly monitored is the thoracic
changes that signal impending damage. The level.
monitoring system must be able to monitor Monitoring is of benefit in many surgical
multiple structures and the same structure procedures on the spine.1–3 Immediately post-
with multiple techniques to provide the rapid operatively, persistent neurologic deficit devel-
and accurate feedback to the surgeon in a ops in less than 0.5% of patients who undergo
relatively hostile electrical environment. corrective operations for scoliosis or other sur-
Electrophysiologic testing early in a pro- gical procedures on the spine, but this deficit
cedure will distinguish those functions that can be devastating. Of the complications that
Spinal Cord Monitoring 753
occur, one-half are complete paraplegia and values or normal control values, the patient
one-half are incomplete paraplegia, with one- will generally serve as his or her own con-
third of the patients having no recovery of func- trol with a change in values during the pro-
tion.4 With surgical monitoring, some patients cedure being indicative of potential neurologic
can be considered surgical candidates who oth- injury. One must therefore establish reliable
erwise might not be because of the risk of an baseline recordings before critical portions of
adverse outcome. the procedure as well as monitor all poten-
There is inherent limitation with any mon- tial confounding factors so the surgeon can be
itoring system. False-positive results are not informed of a true potential change due to
infrequent and likely reflect either a subclinical surgical manipulation.
lesion or more likely technical factors that have
artificially affected the potentials. Identifying Key Points
changes requires that a well-defined set of
• Monitoring methods must be able to pro-
baseline values be obtained during the initial,
low-risk portions of the surgery. The variations vide rapid and reliable feedback on both
due to extraneous factors must be identified so sudden and gradual changes.
• Multimodal monitoring of structures at
that the surgeon can be assured that alterations
of responses are related to the surgical pro- risk is required to avoid false-negative
cedure and not to changes in blood pressure, results.
• Monitoring requires assessment both at
artifacts, anesthesia, or other factors. False-
negative results, where the patient experi- baseline and throughout the surgical pro-
ences neurologic deficit without an identifiable cedure.
• Avoidance of false-positive changes requi-
change in the recordings, occur much less fre-
quently. In most cases, this is likely related to res monitoring and accounting for:
◦ Physiologic variables—anesthesia, neu-
the involvement of critical structures that were
not directly monitored during the procedure. romuscular block, temperature, muscle
An example of this would be paralysis after activity, and blood pressure.
◦ Electrical interference—60-cycle, noisy
spine surgery when the SEP remained stable
but MEPs were not monitored. This can also or dislodged electrodes and cautery.
occur when monitoring is discontinued prema-
turely. As previously stated, some abnormali-
ties may be seen immediately after an injury EQUIPMENT AND ELECTRICAL
while others, particularly those indicative of SAFETY
mild nerve compression, may not become man-
ifest for up to an hour after the step in the Two International Federation of Clinical Neu-
procedure that caused the injury. Monitoring rophysiology publications provide specific rec-
must therefore continue throughout the oper- ommendations for assuring quality and safety
ative procedure, even after a so-called critical during surgical monitoring.5, 6 The equipment
period in the operation has passed. used for surgical monitoring is typically the
Intraoperative monitoring differs from rou- same as that used in outpatient testing with
tine electrodiagnostic studies in several ways, some modifications. Evoked potential equip-
both from a technical standpoint and in regard ment should include capabilities for adding,
to interpretation. The most prominent factor subtracting, storing, smoothing, automatic arti-
is related to a hostile electrical environment fact rejection, and simultaneous display of mul-
when devices such as cautery, 60-Hz artifact, tiple traces from several channels. EMG equip-
respirators, or warmers may attenuate or comment should allow audio presentation as well
pletely obliterate the response. The appear- as visual, and should have automatic artifact
ance and reproducibility of potentials recorded rejection to minimize operative interference,
intraoperatively are affected greatly by the type particularly that due to cautery. The common
and level of anesthesia, blood pressure, tem- mode rejection ratio should be at least 85 dB,
perature, and other physiologic variables. The for elimination of 50- or 60-Hz line interfer-
underlying disease process itself will likely also ence, which is a common problem in the OR.
have affected the baseline responses. Due to All equipment should conform to OR safety
these factors, rather than utilizing absolute specifications and careful attention must be
paid to electrical safety when using equipment MONITORING

purchased commercially or made in house. METHODS—SOMATOSENSORY
Monitoring equipment must not be able to EVOKED POTENTIALS
allow a current of more than 100 mA to pass
through the patient if there is a failure of SEPs have become an accepted method of
equipment grounding. This is measured as the monitoring spinal cord function during a wide
chassis leakage current across a functioning variety of surgical procedures on the spinal
ground plug. For electrically sensitive patients, cord. Although they are most commonly used
such as those with catheters leading to the during correction of scoliosis, any spine surgery
heart or great vessels, the limit should be may warrant SEP monitoring if there is risk of
10 mA. These current limits are substantially cord damage. These include monitoring dur-
larger than would occur under normal oper- ing surgery for cervical spondylosis and for
ating circumstances with properly function- intramedullary and extramedullary spinal cord
ing equipment. The most common danger is lesions such as tumors and vascular malforma-
from improper and malfunctioning grounding. tions. As shown in animal models, ischemia
A patient should be electrically grounded at or compression of the spinal cord prolongs,
only one site, usually a large ground plate. That reduces, and then obliterates SEP in pro-
ground is used for the diathermy and for simi- portion to the amount of damage. A large
lar routine OR equipment. All OR machinery multicenter study of scoliosis surgery showed
should be properly grounded; in that case a that the incidence of postoperative neurologic
second patient ground is unnecessary. Current deficits was 0.46% with SEP monitoring and
neurophysiology equipment should have opti- 1.04% without SEP monitoring.7
cal isolation of each patient contact, preventing Optimal use of intraoperative SEP requires a
conduction of inadvertent electrical currents number of specific decisions about methods of
between the patient and the equipment. An stimulation and recording. These vary with the
optically isolated isoground is used with many level of the nervous system to be monitored,
such pieces of equipment. In general, this can the structures at risk, the type of surgery being
be used safely in the OR, even on a patient performed, the extent of preoperative deficit,
who has a true ground plate already in place. and the choice of anesthetic.
In such a patient, any ground current will
travel out the true ground plate, without sub-
stantial leakage current traveling through the
isoground circuit. The neurophysiologist needs Stimulation Techniques
to take the responsibility for assuring that the
leakage current of equipment is tested and Similar to SEP performed in the outpatient set-
safe. At most hospitals, biomedical engineers ting, stimulation of a peripheral nerve is used
are available to check equipment for proper to generate the responses. Stimulation may
grounding on the main power cord and for be performed on any of the major peripheral
leakage current along any connection to the nerves and can be either unilateral or bilateral.
patient. Commonly stimulated nerves include the tibial
nerve at the ankle, peroneal nerve at the knee,
Key Points and median or ulnar nerves at the wrist. Uni-
lateral stimulation of nerves whose fibers are
• Equipment requirements for optimal most at risk or which are most likely to show a
recognition of changes include change is preferred, although occasionally the
◦ Auditory and visual signal display. presence of a preoperative deficit or a particu-
◦ Automatic artifact rejection. lar anesthetic agent may attenuate SEP voltage
◦ Simultaneous, eight-channel, multi- sufficiently that bilateral stimulation is needed.
modality recording and display. Bilateral stimulation, however, can mask asym-
◦ Sequential stimulation and recording on metric changes in the SEP that would be easily
single sweep traces. revealed by unilateral stimulation.
• Monitoring equipment safety requirem- Surface electrodes embedded in a plas-
ent—100 mA current limit through the tic strip or platinum EEG needle electrodes
patient in the event of equipment or may be used for stimulation and should be
grounding failure. fixed firmly over the nerve. Stimulus intensity
must be supramaximal for large sensory fibers. surface electrodes on the scalp, similar to those
Levels that lead to a muscle twitch in distal used for laboratory SEP recording, or sub-
muscles are used and should be determined dermal needle electrodes. Surface electrodes
prior to neuromuscular blockade. When reli- must be firmly attached with collodion and
able responses cannot be obtained from more filled with conductive gel to ensure stability
distal stimulation sites in the leg, stimulation of and low impedance throughout long surgical
the sciatic nerve in the proximal thigh or cauda procedures. Subdermal needle electrodes are
equina may be successful. being more frequently utilized and have better
Rates of stimulation may affect the elicited impedances. Electrodes are placed at standard
responses. Rapid stimulation rates are optimal sites on the scalp and along the peripheral
in order to decrease the time needed to obtain nerve and spine.
a response. Although SEP can be recorded For scalp recording, electrodes are placed
with stimulation rates of 5 Hz or even 10 Hz at Cz–Fz for tibial responses and C3 or
in most awake patients, under anesthesia the C4 -Fz for median and ulnar responses. Larger,
scalp SEP fatigues at rates greater than 3 Hz. more reliable tibial SEP may be seen with
It may be necessary to stimulate at rates as low C3–C4 recordings than with the standard
as 0.5 or 1.0 Hz in children and adolescents, Cz–Fz recordings, and should always be tested
especially at deeper levels of anesthesia. to ensure that the optimal potential is selected.
Direct stimulation in the operative field For recording the potentials at the cord
gives direct activation of ascending and level, needle electrodes are placed adjacent to
descending SEP. It is often not easy to be the peripheral nerve or lamina. Percutaneous
certain whether such potentials are motor needle electrodes of 30–75 mm placed directly
or sensory, since cord stimulation may action the lamina outside the surgical field or in
vate either or both. Direct cord stimulation the intraspinous ligament at any spinal level in
methods typically utilize subarachnoid, epidu- the surgical field can record well-defined, reli-
ral, spinous process, and intraspinous ligament able potentials. Esophageal or nasopharyngeal
stimulation and recording. Epidural electrodes electrodes at the cervical levels are necessary
inserted between the spine and the dural sac to record cervical cord potentials outside the
obtains large readily recorded potentials. Sim- surgical field in cervical spine surgery. Either
ilar recordings have been made of descend- of these can provide a good stable recording
ing activity by stimulating and recording the anterior to the spinal cord.
stimuli from the spinal cord directly. Epidural recordings can be made at any
level in the operating field with multilead cable
Key Points electrodes, strip electrodes, fine wires, or nee-
dles in the intraspinous ligament. Small wick
• Nerves stimulated should be selected to electrodes or plastic embedded platinum elec-
maximize recognition of damage to path- trodes are used to record directly from the
ways at risk and to obtain reliable SEP. spinal cord or epidural space, and from the
◦ Tibial nerve at the ankle (peroneal surface of the cortex. For each of these active
nerve at the knee) or more proximal. electrodes an appropriate reference must be
◦ Ulnar nerve (median nerve) at the wrist chosen. Nearby reference electrodes reduce
or more proximal. noise, but distant electrodes enhance signal
◦ Unilateral (bilateral). amplitude. In general, active and reference
◦ Other peripheral nerves based on struc- electrodes should be of the same material and
tures at risk, including direct stimula- firmly affixed to minimize noise.
tion of nerves in the operative field. In addition to recording from the scalp and
• Stimulation intensities should be suffi- the cord, recordings should be made from
cient to elicit a muscle twitch in appropri- peripheral sites along the sensory pathways to
ate muscles. ensure adequate stimulus input. Adherence to
this important principle will minimize the inci-
dence of false-positive changes and make trou-
Recording Techniques bleshooting for technical errors more efficient.
Figures 44–1 and 44–2 demonstrate techni-
Recording the responses during intraoperative cally reliable approaches utilized for SEP mon-
SEP monitoring is performed using standard itoring during spinal surgeries.
give larger responses but are associated with

more technical problems during the surgical
procedure, such as mechanical artifact. Such
recordings require more technical expertise for
obtaining satisfactory results and require that
the surgeon be familiar with and cooperative
with the recording procedure.
As with routine SEP, the low-amplitude
responses obtained with SEP requires averag-
ing multiple responses. The number of stimuli
averaged should be just enough to obtain reli-
able recordings in order to give the surgeon
feedback as rapid as possible. However, when
averaging large numbers of stimuli, a transient
or gradual change in the evoked potential may
be obscured. As with other evoked potentials,
stable baseline studies must be obtained prior
to critical stages and then followed continu-
ously when these structures are at risk.
Signal amplification and filter settings are
Figure 44–1. Standard placement of electrodes for mon- similar to those used for diagnostic SEP
itoring SEPs during thoracolumbar spine surgery. The sci- recordings, although at times the gain must
atic response is recorded with a needle electrode near the be reduced or the bandpass restricted because
nerve. Cervical responses are recorded from an electrode of the high noise environment. Amplification
on the lamina of the spine of C7 and from an esophageal
electrode. Scalp responses are recorded from standard ver- of 5–10 μV/cm, sweep speeds of 2–10 ms/cm,
tex electrodes. (From Daube, J. R., C. M. Harper, W. J. low-frequency filters (LFF) of 30–100 Hz, and
Litchy, and F. W. Sharbrough. 1990. Intraoperative moni- high-frequency filters (HFF) of 2000–3000 Hz
toring. In Current practice of clinical electroencephalogra- are generally satisfactory.
phy, ed. D. D. Daly, and T. A. Pedley, 2nd ed., 747. New
York: Raven Press. By permission of Lippincott Williams &
Wilkins.) Key Points
• Electrodes must be firmly attached.
Recordings made in the surgical field can be • Leg sensory pathway requires testing both
performed in cases where the spine is open C3–C4 and C3 and C4 to FZ record-
and the dura mater exposed. These recordings ings at the beginning to determine which
provides the most reliable recording.
• Sensitivities and sweep speed settings
should be adequate to best identify the
Fz
C4′
SEP (0.5–10 μV/cm and 2–10 ms/cm).
C3′
• Filter settings that best eliminate arti-
fact with a minimum of reduction of the
Nasopharyngeal-Fz SEP (low of 30–100 Hz and high of 2000–
3000 Hz) are required.
• Stable baseline recordings should be
obtained after the patient is anesthetized,
but before the critical period of surgery.
Esophagus-Fz Physiologic and Technical Issues

Figure 44–2. Electrode placement for standard median Noise. Electrical noise is the most common
or ulnar SEPs during cervical spine surgery. The cortical
response is best recorded at C3 –C4 ; cervical potentials
technical problem encountered during intra-
are recorded via either a nasopharyngeal or an esophageal operative monitoring. Early in the procedure,
electrode. every effort must be made to identify and
Figure 44–3. Series of averages of 200 SEP displayed as a stack of sweeps reading sequentially from top (arrow). Left,
Loss of scalp amplitude due to depth of anesthesia. Right, Preserved cervical response throughout the surgery.
eliminate all sources of noise, especially 60 (Figs. 44–3 and 44–4). Since the reduction
cycle. Maximizing the signal to noise ratio is in amplitude is directly related to depth of
an ongoing challenge in the electrically hostile anesthesia, the level of anesthesia should be
environment of the OR but with proper elec- kept as light as possible. This is especially true
trode placement, impedance matching, and fil- in the presence of disease and in children and
tering, reproducible recordings can generally adolescents. The anesthetic effect varies with
be obtained with 250 stimuli. The recording the agent used: it is least with fentanyl and
system should suppress input during cautery nitrous oxide (NO), more with isoflurane and
and reject high-amplitude artifact. enflurane, and greatest with halothane. Both
propofol and midazolam provide stable median
Muscle Activity. EMG activity from surround- N20 and tibial P40 latencies and amplitudes for
ing muscle can also produce unwanted artifact
SEP monitoring of cord function during spine
if neuromuscular activity is not blocked. EMG surgery.9, 10 There may be some increase in
activity may be so prominent as to obscure
latency and decrease in amplitude immediately
SEP. This most commonly occurs when the
after induction.
level of anesthesia is low.
Newer agents including desflurane and
Anesthesia. Anesthesia reduces the cortical sevoflurane lead to similar reduction of ampli-
SEP recorded from the scalp and to a tudes and the addition of NO to these agents
much lesser extent SEP recorded elsewhere8 leads to marked reduction.11 The primary
Figure 44–4. Effect of anesthesia on tibial (cervical cord stimulation) scalp SEP during scoliosis surgery on a 10-year-old
boy. Left, Baseline recordings before inhalational anesthetic. Right, Loss of the scalp SEP with preservation of the cervical
response after a short period of inhalational anesthetic illustrating the effect of inhalational anesthetics on the anterior horn
cell pool in children and adolescents.
effect of anesthesia on the scalp SEP is reduction in SEP can occur because of the
a gradual prolongation of the latency and accumulation of subdural air. This change is
reduction in the amplitude. These effects recognized readily by comparing the standard
increase the longer the period of anesthesia. vertex electrode recording with that of elec-
Rarely, the scalp response is enhanced after the trodes placed just above the ear, where there
induction of anesthesia. In a small proportion is less subdural air.
of cases, predominantly in children and ado- Differentiating changes caused by technical
lescents, the response is lost immediately after factors from those caused by pathway dam-
the induction of anesthesia. Premedication has age requires that the alteration in the ampli-
little effect on the SEP. tude or latency be consistent at both the neck
Since the anesthetic sensitivity of spinal cord and scalp recording sites and the peripheral
evoked potentials is less than scalp record- response be intact. Accuracy of monitoring can
ings, a combination of spinal and scalp record- be improved with appropriate use of periph-
ings provides the advantages of determining if eral recordings, including monitoring arm SEP
changes in responses may be related to level over Erb’s point with median or ulnar nerve
of anesthesia. The percutaneous electrodes on stimulation and monitoring leg SEP from the
the lamina allow continued recording of spinal sciatic nerve at the gluteal fold or the N22 lum-
potentials when reproducible scalp potentials bar potential from T12–L1. If there is damage
cannot be obtained. Although in most spine to the central pathways during the procedure
surgeries, a reduction of scalp amplitude with leading to loss of the spine and scalp responses,
preservation of the cervical spine potential the peripheral response should still be present.
would suggest anesthetic effects; this is not If the peripheral response is also absent there
necessarily true in upper cord or brainstem is either a peripheral process or more likely
procedures where the cervical response will malfunctioning of the stimulator (Fig. 44–5).
be maintained despite a clinically significant
insult. Occasionally, because of a patient’s pre- Key Points
operative neurologic deficit, scalp potentials • Multiple levels of the peripheral and cen-
cannot be recorded. In many of these patients, tral somatosensory pathways should be
SEP may be recorded at the neck. monitored to localize sites of technical
Blood Pressure. In addition to the level of anes- problems or damage.
• Inhalation agents are the most com-
thesia, blood pressure alterations may prolong
the latency and reduce the amplitude of SEP, mon physiologic cause for cortical SEP
especially if the mean blood pressure is less reduction.
◦ Simultaneous spinal cord recordings
than 70 mm Hg.
that are not affected are needed to best
Variability of Responses. A frequent problem recognize inhalation agent effects.
during surgical monitoring is SEP variability • Extraneous sources of false positives and
on sequential recordings. SEPs may change for false negatives must be controlled.
many reasons other than surgical damage to the ◦ External and equipment noise should
sensory pathway. During cervical surgery with be eliminated early.
the patient in the sitting position, a marked ◦ Cautery suppression should be available.
Figure 44–5. Peripheral stimulation failure during spine surgery. A, Intact scalp, neck, and sciatic recordings. B, Abrupt
loss of all response due to mechanical displacement of the stimulator at the ankle. C, SEP return with reestablishment of
stimulation.
◦ Inhalational anesthetics should be as of not only the anesthetic levels and physio-
light as possible. logic variables, but must be in communication
◦ Blood pressure reduction can reduce with the surgeon at all times since mild changes
SEP amplitudes. during critical portions of the procedure may
◦ SEP variation during baseline record- reflect significant neural compromise.
ings defines later limits. Small, consistent signal changes from two
sites of stimulation provide evidence of com-
pression before more major changes appear.
Infrequent and unusual nonsurgical causes of
Application and Interpretation of rapid changes in SEP must always be con-
SEP Changes sidered before concluding that the operation
is the cause. These causes may be any of
Ideally, the ordinary background variability of the physiologic changes in temperature, blood
the potentials recorded during intraoperative pressure, or subdural air.
SEP is no more than 30% in amplitude and Several patterns of change in SEP have
1.0 ms in latency. While a 50% fall in amplitude been observed that correlate with postopera-
from baseline and a 5% change in latency, if the tive neurologic function. The change may be
concentration of anesthetic and other physio- late or gradual, emphasizing the importance
logic factors have remained stable, are gener- of continuing the monitoring until the patient
ally guidelines for high likelihood of damage,12 is awake. Gradual changes in SEP may be
no absolute change in amplitude can be con- caused by retraction, compressive hematoma,
sidered evidence of spinal cord damage. Sub- or by ischemia of the spinal cord or periph-
cortical potentials are less likely to be affected eral nerves (Fig. 44–6). Less frequently, SEP
by anesthesia and provide a means to interpret change abruptly, usually in relation to an acute
the significance of cortical changes. Because contusion of the spinal cord or vascular infarct.
of this, the neurophysiologist must be aware When the loss of SEP is abrupt, the site of
Figure 44–6. Gradual loss of SEPs during stabilization procedure for cervical spine fracture in a 60-year-old man.
Responses were lost within a few minutes after wiring C5–C7, but they returned quickly after the wires were removed.
The patient awoke with no deficit. B, bilateral; L, left; R, right. (From Daube, J. R., C. M. Harper, W. J. Litchy, and F. W.
Sharbrough. 1990. Intraoperative monitoring. In Current practice of clinical electroencephalography, ed. D. D. Daly, and
T. A. Pedley, 2nd ed., 765. New York: Raven Press. By permission of Lippincott Williams & Wilkins.)
Left
Tibial
1:00
◦ Intact peripheral response with loss of
both neck and scalp SEP.
◦ No possible technical or physiologic
Left 1:50 causes.
• An amplitude reduction of greater than
Epidural 50% or a latency prolongation greater than
Left 2:25 T10–11 baseline variation (generally a 5% latency
increase) is considered abnormal.
T11–12
Bilateral 2:29
MOTOR EVOKED POTENTIALS

Bilateral 2:35
T12–L1
Historically, emphasis has been placed on
SEP monitoring during spine surgery with the
assumption that a significant neurologic injury
Bilateral 3:15
L1–L2 that results in motor deficit would be sufficient
enough to also involve the recorded sensory
pathways. In reality, an injury sparing the pos-
2.5 μV 2.5 μV
terior columns, whether due to compression
20 ms 20 ms
or vascular insult, will not significantly affect
Figure 44–7. Tibial SEPs during thoracic spine surgery. the somatosensory pathways yet can result in
Note the relatively sudden loss of scalp potentials, unre- devastating motor deficits. A complimentary
lated to anesthesia, at 2:29. An epidural recording elec-
trode inserted as shown on the right recorded the tibial
method of recording the MEP is available to
spinal cord SEP. The SEP was not recordable above T11. monitor the integrity of the corticospinal tract.
The patient awoke with paraplegia likely due to spinal A variety of methods of stimulation and record-
cord infarct. (From Harper, C. M., and J. R. Daube. 1989. ing have been reported over the years. This
Surgical monitoring with evoked potentials: The Mayo section will focus on the more commonly used
Clinic experience. In Neuromonitoring in surgery, ed.
J. E. Desmedt. Amsterdam: Elsevier. By permission of the methods with only brief comments on other
publisher.) options.
injury should be localized by recording with Stimulation Techniques

epidural or direct spinal electrodes, followed
by careful inspection of the involved level Motor evoked potentials are reliably obtained
for hematoma or other potentially reversible intraoperatively with direct stimulation of the
causes of spinal cord injury (Fig. 44–7). If spinal cord or cerebral hemispheres. MEPs
the loss of SEP is abrupt and not reversible, can be elicited with either electric or magnetic
paraplegia is likely to occur. Improvement in stimulation. In an anesthetized patient, mag-
the amplitude of SEP intraoperatively is usu- netic stimulation has no advantage over electric
ally associated with improved neurologic status stimulation and has major disadvantages: the
postoperatively. Postoperative motor deficits coil is cumbersome and hard to immobilize rel-
without associated changes in SEP intraoperative to the skull, the apparatus is expensive
atively are infrequent but well documented.13 (particularly if trains of stimuli are to be given),
The motor deficits may occur as part of the and magnetic MEPs are more sensitive than
anterior spinal artery syndrome or they may electric MEPs to anesthetic agents.14 There-
be caused by direct injury to the ventral spinal fore, electrical stimulation is the most common
cord, ventral horn cells, or nerve roots. method utilized.
Transcranial electrical stimulation is per-
Key Points formed by stimulation via subdermal elec-
trodes whereas spinal cord stimulation can
• Recognition of spinal cord damage requires: be performed using direct epidural, laminar
◦ SEP amplitude loss beyond that seen needle, nasopharyngeal or esophageal elec-
during baseline recordings. trodes. In the past, spinal cord stimulation was
Figure 44–8. Electrically induced MEPs can be evoked with stimulation between C3 and C4 on the scalp, or between a
nasopharyngeal and a laminar electrode. The former is more commonly used. Recordings are made from surface electrodes
of multiple leg muscles bilaterally.
easier to perform; however, with recent tech- Spinal cord stimulation can be accomplished
nologic advances transcortical electrical stimu- with a nasopharyngeal/esophageal active and
lation (TCES) has become the technique most laminar needle electrode. Other options are
frequently utilized. stimulation in the operative field with inter-
The stimulation technique for MEP has spinous or epidural electrodes in the surgical
been described in Chapter 25. In summary, field using a distant anode in the subcutaneous
with TCES anodal stimulation is given with a tissue, but are more technically difficult.
short duration (0.05 ms), rapid rise time stim-
ulus using subcutaneously placed EEG elec-
trodes at C3 and C4 (Fig. 44–8). Several (2–5)
Key Points
stimuli with an interstimulus interval of 1–4 ms
are given with intensity of 200–800 V. These • TCES of MEPs requires optimization of a
parameters are varied until a reproducible number of choices:
response can be recorded in all the muscles ◦ Contraindications to TCES must be
examined. The polarity of the simulation is also considered for each patient.
switched to assure maximal anodal stimulation. ◦ Stimulating electrodes at C3 and C4
Contraindications to TCES include the pres- should have each electrode tested as
ence of a pacemaker, infusion pumps, cochlear anode to identify lowest threshold for
implants, or a history of epilepsy, seizures, skull activation.
fracture or defect, major head trauma, stroke, ◦ Stimulus parameters:
other intracranial disease, or aneurysm clips Two stimuli are sometimes satisfac-
and other retained metal fragments. tory, but 3–5 may be needed.
Inter-stimulus intervals of 3 ms are also be recorded in patients with a preexisting

usually satisfactory, but longer or neural deficit and an abnormality in the corti-
shorter intervals can be tried to cospinal tracts may be identified promptly. The
enhance responses. disadvantages of spinal cord recordings are that
Stimulus intensities of 200–800 V they are feasible only when epidural leads can
may be needed. be inserted, which usually requires a posterior
• Cervical cord stimulation can be used if approach to the spinal cord; they do not iden-
cranial stimulation is contraindicated: tify the side responsible for any deterioration in
◦ Stimulation between a percutaneous the recorded volleys and they are not as reliably
needle electrode on a cervical lam- recorded from the lumbar cord.
ina and a nasopharyngeal electrode are
most efficient and reliable.
◦ Paired pulses with a 3-ms interval can PERIPHERAL NERVE RECORDINGS
evoke potentials under 110 V.
Nerve action potentials can be recorded in
the region of the sciatic or tibial nerve in
response to stimulation of the cerebral cor-
Recording Techniques tex or the spinal cord, but they are less well
defined with cortical stimulation. Due to the
Motor evoked potentials can be recorded small responses, 100 or more responses need
directly from the spinal cord as a cord-evoked to be averaged. If inhalation anesthesia is used,
potential, from a peripheral nerve as a nerve the neurogenic MEP obtained with spinal cord
action potential, or from muscle as a CMAP. stimulation may reflect activity primarily of
posterior column axons, because spinal motor
SPINAL CORD RECORDING neuron activity is reduced. With the neuromus-
cular block that typically is used, a significant
Stimulation of the spinal cord, even at thresh- component of the neurogenic potential may be
old, activates both motor and sensory fibers. the end plate potential from surrounding mus-
Therefore, the MEPs recorded from the spinal cle as well as potentials from the motor and
cord in response to direct spinal cord stimula- sensory fibers in the peripheral nerve.
tion (cord-to-cord) are a mixture of potentials
in ascending and descending pathways, and
this mixture may be less sensitive to spinal MUSCLE RECORDINGS
insult than monitoring of selective pathways.
Recordings may be made within the surgical CMAPs are the potentials that are most
field, and these are most useful for operations commonly recorded during MEP monitoring.
on the spinal cord (e.g., tumors or arteriove- CMAPs may be recorded with surface elec-
nous malformations). Epidural bipolar cardiac trodes placed over the muscle in response to
pacing electrodes with an inter-electrode sep- stimulation of either the cerebral cortex or the
aration of 2–3 cm are suitable for recording spinal cord, but the potentials are different for
descending MEP (the same electrode can be the two modes of stimulation. Narcotic anes-
used to record SEP directly from the spinal thesia is necessary for CMAPs to be elicited,
cord or to stimulate the spinal cord). The and even then the CMAPs produced by sin-
directly recorded potentials can help to localize gle stimuli to the motor cortex are too variable
the area of damage or record responses that are for monitoring motor function reliably. Pairs or
too small to be obtained with other methods. trains of stimuli, especially to the spinal cord,
Responses recorded from the spinal cord produce enough temporal summation of exci-
during transcranial MEP monitoring have the tatory input to activate most motor neurons.
advantage of being relatively immune to the Large, stable CMAPs can be recorded, but this
effects of anesthetic agents, and not affected by often requires partial neuromuscular block to
muscle activity since full neuromuscular block- control movement (Fig. 44–9). The level of
ade will not affect the responses. In addition, neuromuscular block is monitored best with
SEPs can be recorded reliably in the same recording of CMAPs in response to peripheral
sweeps if the cerebral cortex and peripheral nerve stimulation along with the CMAPs from
nerve are stimulated simultaneously. MEP may central stimulation.
B
MEP during scoliosis surgery
A
Spinal stimulation: Paired ISI 3 ms, 80 mA, dur 1ms
ESOPHAGEAL-LEFT CERVICAL
Fz
left rect. fem.
C3′
left ant. tib
left soleus
right rect. fem
right ant. tib.
right soleus
C–7
1 mV 10 ms
Esophagus-Fz Neuromuscular block 10%
Figure 44–9. MEPs elicited by stimulation of the spinal cord with paired stimuli applied between a laminar needle and
esophageal electrodes. A, Esophageal stimulating or recording electrode location with tibial SEP shown from the scalp and
esophageal electrodes. B, MEP from esophageal stimulation recorded as surface CMAPs from leg muscles bilaterally, with
partial neuromuscular block. Ant. tib., anterior tibialis; rect. fem., rectus femoris.
Compound muscle action potentials are more sensitive to anesthesia than spinal evoked
recorded best from multiple muscles in both CMAPs.
legs. A major advantage of the technique is
that monitoring can be adapted by choosing Key Points
muscles to suit the specific clinical need. For
example, muscles innervated by specific nerve • MEP recordings can be made from spinal
roots when the operation is low spinal or seg- cord, peripheral nerve, or muscle.
mental or when a nerve root is known to be ◦ CMAP from muscle are easiest to per-
at risk. The signal-to-noise ratio for CMAPs is form and show only a limited variability
sufficient for single trials to be recorded with- from shock to shock.
out averaging. Although the reproducibility of Surface, subcutaneous, or intramus-
evoked CMAPs is good, it may not be as high cular electrodes can be used.
as that of MEP in epidural recordings. Sufficient depth of recording is
The major advantages of MEP monitoring needed to assure a response.
with CMAPs are that unilateral dysfunction ◦ Peripheral nerve MEP recordings are
can be identified, evoked potentials with spinal of unique value only in determining the
cord stimulation are resistant to anesthesia, presence of viable motor axon in a nerve
there is no intrusion into the operative field, root during traumatic brachial plexus
CMAPs evoked with spinal stimulation can be reconstruction.
recorded simultaneously with SEP, cortically ◦ Direct, intraoperative, spinal cord MEP
evoked potentials can be adapted for virtu- recordings
ally all spinal and cerebral operations, and Can localize pathology along the
CMAP recording is equally useful for oper- spinal cord but

ations on the low spinal cord, cauda equina, Cannot distinguish unilateral from
or nerve root. Disadvantages include variabil- bilateral abnormalities.

ity in the CMAPs evoked with spinal stimula- • Neuromuscular blocking agents are nee-
tion with the level of neuromuscular blockade, ded to minimize limb and body movement
unable to perform with spinal stimulation dur- during surgery.
ing cervical spine surgery, and cortically evoked ◦ Continuous monitoring with repeti-
CMAPs intrinsically are more variable and tive stimulation can distinguish loss of
potential from a change in neuromus- Applications and Interpretation of

cular block, from one due to change in MEP Changes
central anesthetic level.
The threshold values of MEP changes that
warn of imminent neurologic damage vary in
the literature. Given the significant variability
Physiologic/Technical of the response due to physiologic and techni-
Considerations cal factors many laboratories utilize an all-or-
none response. Therefore, if a response was
Anesthesia. The primary factor affecting the present and disappears during a critical stage
MEPs is anesthetic agents. These agents can of the procedure there is high likelihood of
suppress the response at multiple sites but par- neural damage assuming that technical or anes-
ticularly those that involve synaptic transmis- thetic effects are excluded. Some authors have
sion. Specifically these are the cortex, anterior used a value of 80% reduction of one out of six
horn cell, and neuromuscular junction.15 Halo- lower extremity muscles showing a sensitivity
genated inhalation agents easily abolish MEPs of 95%.16 When relatively strict criteria such as
by blockade at either the cortex or the anterior this are used, there is a clear increase in false-
horn cell. This appears to be more prominent positive responses with this same study show-
with transcranial magnetic stimulation. If used ing a specificity of 91%. Changes in latency of
at all, the concentration of these agents need spinal cord or muscle responses have not been
to remain very low (<0.5%). There are no useful intraoperatively.
reports of the effect of propofol or midazolam The development of reproducible MEP
on MEP. techniques has advanced the ability to monitor
the motor tracts during spinal procedures. As
Muscle Activity and Neuromuscular Blockade. previously mentioned it has been assumed that
Since the responses obtained during record- if the SEPs are stable during a procedure, the
ing the CMAPs over the muscles are motor motor tracts should be spared. Although this is
responses, the use of neuromuscular block- true in most instances, cases of paralysis in the
ing agents will suppress or eliminate these setting of normal SEP can occur. By utilizing
responses. Therefore, either no blockade or multimodal monitoring of both motor and sen-
very controlled neuromuscular blockade up sory pathways the risk of catastrophic paralysis
to 50% may be used when monitoring the can be reduced.
CMAP. Since there is little or no neuromus- The utility of transcranial electrical stimu-
cular blockade, there will generally be some lation during surgery has been evaluated by a
movement with transcortical stimulation and number of investigators. Although the success
the surgeon should be notified before the stim- rate for establishing reliable MEP during high-
ulus is applied. risk surgical procedures has been hindered in
Blood pressure and Temperature. Temperature the past by lack of equipment and appropri-
also may have a mild effect in that there may be ate protocols, a recent study found that MEP
a gradual increase in stimulation threshold. could be successfully monitored in 94.8% of
cases for upper extremities and 66.6% of lower
extremities.17 In this same study it was noted
that extremes of age (<7 or >64) and presence
Key Points
of a spinal cord lesion reduced the success of
• MEPs are exquisitely sensitive to inhala- obtaining reliable responses.
tional agents. In a series of reports, Levy et al.18, 19
◦ Baseline surgical recordings are needed described their experience with MEP moni-
initially without inhalation agents. toring during a variety of neurosurgical pro-
◦ Levels of anesthetic should be changed cedures. They used a modification of the
only slowly and should be present. bipolar technique with the anode placed on
• Neuromuscular junction block should be the hard palate in some patients and direct
monitored throughout to identify ampli- cortical stimulation with anode and cathode
tude reduction due to a change in placed via a burr hole in others. Responses
level. were recorded either from the spinal cord or
from the peripheral nerve. They reported a • Inhalation of anesthetic agents can abolish
good correlation between changes in the MEP responses; thus, very low levels of inhala-
and postoperative motor function. In over 100 tion of anesthetic agents with supplemen-
cases there were no patients with normal MEP tal narcotic anesthesia and partial neu-
who developed new motor deficits. romuscular blockade allows for optimal
Zentner et al.20 reported their experience recordings.
with MEP monitoring in 50 neurosurgical
operations on the spinal cord. The bipolar tech-
nique of Merton and Morton was used with ELECTROMYOGRAPHY AND
CMAPs recorded over thenar and anterior tib- NERVE CONDUCTION STUDIES
ial muscles at latencies of 20 ms and 30 ms
respectively. Using a change in amplitude of Damage to cervical or lumbosacral nerve roots
50% as a limit of significance, they found a or motor neurons may occur during surgical
false-positive rate of 20% with no false nega- procedures on the spine. For example, anterior
tives. Of the 4 patients with new motor deficits horn cells can be damaged during dissection
postoperatively, the MEP disappeared in one of intraspinal tumors or by ischemia caused
and was reduced in amplitude by 60–85% by compression or traction. Radiculopathies
in three. Kitagawa et al.21 reported a similar caused by local compression or traction of a
experience with 20 patients undergoing cer- root are an occasional complication of scolio-
vical spine surgery. The three patients with sis surgery. In cases where roots or nerves are
new postoperative deficits all had at least a at risk, monitoring with a combination of nerve
50% reduction in the averaged spinal MEP conduction studies (NCS) and EMG recording
recorded from the epidural space below the for the presence of neurotonic discharges in
level of surgery. limb muscles innervated by the affected motor
Owen et al. have recorded MEP follow- neurons or axons can warn of potential damage
ing spinal stimulation in 300 patients under- caused by manipulation, traction, or ischemia
going a variety of neurosurgical, orthopedic, of nerve roots and can minimize risk to these
and vascular surgical procedures.22 Ketamine structures (Fig. 44–10B).23, 24
or narcotic-NO anesthesia were used in the In the intraoperative setting, the primary
majority. NO concentrations greater than 60% potentials of interest with EMG recordings
or the use of halogenated anesthetics greatly are neurotonic discharges and motor unit
reduced the amplitude of the MEP. Using potentials. Neurotonic discharges are distinc-
a drop in amplitude of 60% or greater as tive discharges of a motor unit that appear
a significant change in the MEP, there were as rapid, irregular bursts lasting several mil-
18 patients early in their experience that had liseconds or prolonged trains lasting up to
false-positive MEP. These were attributed to 1 minute. Neurotonic potentials occur in
technical difficulties that with experience could response to mechanical, thermal, or metabolic
be easily recognized and corrected. All seven irritation of the nerve that innervates a mus-
of the patients with new postoperative deficits cle whereas motor unit potentials reflect reflex
had significant changes in the MEP amplitude activity of anterior horn cells.25 Each dis-
intraoperatively. SEP were said to be preserved charge may contain 1–10 individual motor unit
in all seven patients with motor deficits. Two potentials which discharge at frequencies of
additional patients had new sensory deficits 50–200 Hz.26, 27 They are distinguished from
with changes in the SEP but not in the MEP motor unit potentials by this burst pattern as
recorded intraoperatively. well as the relationship to thermal, mechanical,
or metabolic irritation of the nerve membrane.
By contrast, motor unit potential will have a
Key Points semi-rhythmic pattern, which may arise with
• MEP can be elicited with either tran- either irritation to the motor axon or voluntary
scranial (magnetic or electrical) or spinal activity due to incomplete muscle relaxation.
electrical stimulation.
◦ In the operative setting electrical stimu- Key Points
lation is more effective and reliable than • Peripheral nerve and muscle monitoring
magnetic. uses EMG and NCS.
Figure 44–10. Three examples of monitoring with EMG and NCS during lumbosacral surgery. A, CMAPs evoked in the
anal sphincter by direct stimulation of tissue in the surgical field identified the tissue as axons in the L2–L4 nerve roots
(lipomeningocele resection). B, Neurotonic discharges in the anal sphincter during lipomeningocele dissection warned
the surgeon of irritation of the L2–L4 axons. C, Motor unit potential firing during lumbar fusion warned the surgeon of
irritation of dorsal root axons. D, CMAPs evoked in L5-innervated muscles by a stimulating electrode in a pedicle screw
hole with less than 10 mA current warned the surgeon that the pedicle screw was close enough to the dorsal root to irritate
it or damage it.
• EMG recordings look for two patterns: and anesthesiologist. Recordings can be made
◦ MEPs indicating inadequate anesthesia. from any somatic muscle and the selection of
◦ Neurotonic discharges indicating irrita- muscles depends on the structures at most
tion or damage to axons. risk. For example, monitoring of L3–S3 inner-
• NCS have two purposes during surgery: vated muscles in the leg, including the anal
◦ Quantitatively define the severity of sphincter, would be helpful in operations for
preexisting and new damage. myelomeningocele (Fig. 44–10A and 44–10B).
◦ Localize the site of the damage. Also, intramuscular recordings of motor unit
potential firing caused by a reflex response can
detect irritation of the sensory axons in the
dorsal root (Fig. 44–10C).
Recording Techniques The EMG recordings are made with stan-
dard gains of 100–500 μV, LFF of 20–30 Hz,
EMG RECORDING HFF of 20 kHz, and sweep speed of
EMG activity can be recorded with a variety of 10–200 ms/division. EMG recordings from
electrodes. Recording with surface electrodes multiple muscles can be presented simultane-
is inadequate since they cannot record activity ously over a loudspeaker as well as on a digital
deep in a muscle nor can they clearly identify or analog display and the activity of interest can
the specific responsible muscle. EEG needles be printed or stored for later review.
placed subcutaneously are commonly used and
in those situations where a deeper muscle is
required, fine nichrome wires can be inserted CMAP RECORDING
with a hollow bore needle. Standard monopo-
lar or concentric needle electrodes can reli- Direct stimulation of nerves in the surgical
ably record EMG activity but have limitation field can provide information about the loca-
related to being bulky and difficult to keep tion and integrity of the nerves.28 If the
in place and out of the way of the surgeon normal anatomy is distorted, recording the
Before clamp During clamp After clamp
CMAP 5 mV
H-reflex 2 mV
MEP 2 mV
SEP
25 μV
20 ms
Figure 44–11. The relative value of different monitoring methods (see text) in detecting spinal cord ischemia is shown by
the marked loss of the MEP with much less change in the other modalities that occurs when the aorta is clamped for TAA.
muscle response to stimulation can help distin- inhalation agents or narcotic anesthesia is
guish among nerve roots and differentiate the preferred although neurotonic discharges can
roots from nonneural structures. NCS record be recorded if short-acting, nondepolarizing
CMAPs produced by local stimulation of indi- neuromuscular blocking agents are titrated
vidual nerves (Fig. 44–10A). Recordings of the to produce a 50% reduction of the baseline
CMAPs can be made with surface, subcuta- motor action potentials. Although such a level
neous, or intramuscular electrodes. of muscle relaxation increases the possibility
Stimulation of the nerves or roots may be of unwanted movement during the operation,
applied with a variety of stimulators, includ- movements of the patient can be prevented
ing monopolar, bipolar, or forceps electrodes. with adequate levels of narcotic or inhalation
Pedicle screw or drill stimulation is also utilized anesthesia. At times additional agents such as
to assure proper placement. This technique fentanyl or midazolam must be administered
involves direct stimulation of the pedicle screw to reduce background muscle contractions and
or screw hole with recording of CMAPs in associated motor unit potentials. If partial neu-
the appropriate limb muscles with either sub- romuscular blockade is used, a continuous
cutaneous or intramuscular needle electrodes. monitor of the degree of blockage should be
If there has been broaching of the pedi- used.
cle wall, CMAPs will be recorded with rela-
tively low levels of stimulation (Fig. 44–10D).
Various studies have determined this risk of Applications and Interpretation of
root trauma secondary to misplaced pedicle Findings
screws to be present with threshold stimulus
intensities less than 6–10 mA.29, 30 This tech- Recording of EMG activity is a simple tech-
nique has also been utilized with placement nique that provides a means for rapid feed-
of iliosacral screws during pelvic ring fracture back to the surgeon to warn of potential
repair.31 nerve trauma. The presence of these dis-
charges warns the surgeon that a nerve is
being affected by surgical activity, and absence
Physiologic and Technical Effects of these discharges can usually reassure the
surgeon that the nerve remains unaffected.
Neuromuscular blockade and Anesthesia. Neu- This technique has use in spinal surgeries
romuscular blockade will significantly attenu- when the roots are felt to be at risk dur-
ate the EMG and CMAP activity and should ing decompression procedures. Cervical or
be avoided as much as possible. Alternative lumbar decompression in those with severe
spondylosis, tumors, or infectious processes important factors to consider are the age of the
that place the roots at significant risk, espe- patient, the surgical risk, and the spinal level of
cially in those myotomes with preexisting neu- surgery.
rologic deficit, are typical indications for this Three major factors must be considered
monitoring strategy. when monitoring is performed in patients
Additionally, these techniques can be used younger than 21 years; each of these factors
for nerve localization when the tissue is not presents a unique challenge to the clinical
clearly identified due to altered anatomy from neurophysiologist. First, infants and small chil-
a tumor that displaces or encases the nerve. dren usually require different stimulating and
recording electrodes and great care in elec-
Key Points trode placement. Second, scalp SEP averag-
ing is more difficult in children, especially
• Neurotonic discharges represent mechan-
younger ones, because the amplitude of slow-
ical irritation of the nerve. wave activity is much higher when the child
◦ Their recognition can minimize nerve
is anesthetized. Because of this, slower rates
root damage during exploration, decom- of stimulation, a larger number of averaged
pression, or pedicle screw placement. stimuli, or a lower level of anesthesia (or a
◦ Intramuscular nichrome wires may be
combination of these) are often required. In
used for deeper muscles but short all children, recordings from the cervical cord
monopolar or intramuscular EEG nee- are ideal to demonstrate that the spinal cord is
dles are generally adequate. intact, even if the scalp response is not clearly
◦ These responses are not affected by
recognizable. Third, in some children and ado-
inhalation agents but neuromuscular lescents, the scalp response is lost early during
blockage should be minimized. anesthesia, presumably an idiosyncratic reac-
tion to the anesthetic agent. In these cases,
a cervical cord recording from a nasopharyn-
TYPES OF SPINAL SURGERIES geal, esophageal, or laminar needle electrode
is necessary to monitor spinal cord function.
Electrophysiologic monitoring can be benefi- Surgical risk varies according to the amount
cial in many surgical procedures in infants, of spinal cord deformity, the severity of preop-
children, and adults by reducing the extent erative deficit, the size and type of lesion to be
and duration of damage. The surgeon should excised, bony stability, previous operations, and
decide whether monitoring is needed because other medical disorders. The surgical risk may
he or she can best judge the risk of neu- be low enough that intraoperative monitoring
ral damage and the structures at risk. The of neural function is not necessary. Also, moni-
anesthesiologist selects the optimal anesthe- toring may not provide any benefit because the
sia, and, after discussion with the surgeon deficit is complete and cannot be made worse
and anesthesiologist, the clinical neurophysiol- or the structure has to be sacrificed to com-
ogist selects the optimal monitoring methods. plete the operation. In many cases SEPs alone
The spine surgery procedures commonly mon- may be adequate but increasingly the addition
itored include scoliosis, kyphoscoliosis, cervical of MEPs has been indicated given more com-
spondylosis and stenosis, lumbar spondylosis plex spine surgeries and intraspinal processes.
and stenosis, spine trauma, rheumatoid arthri- For each level of spine surgery, the risk to each
tis, spine tumor, and herniated disk. neural structure should be assessed separately.
If the risk is primarily to nerve roots rather
than the spinal cord, spinal cord monitoring
Primary Spine Disease may not be needed. In those settings EMG
alone may be indicated. For example, in most
The optimal methods of monitoring differ from lumbar decompression surgeries there is little
patient to patient depending on the age of the benefit to monitoring SEP as these procedures
patient, preoperative deficit, type of surgery, occur below the level of the spinal cord and
spinal level of surgery, anesthetic agents used, carry the largest risk to the nerve roots. Needle
and other individual patient factors regardless or fine wire EMG recording of multiple bilat-
of the spinal level of the operation. The most eral myotomes would be more important in
Figure 44–12. Gradual improvement of SEP amplitude and latency during spinal cord decompression for rheumatoid
arthritis. The patient’s neurologic deficit improved postoperatively. (From Daube, J. R., C. M. Harper, W. J. Litchy, and
F. W. Sharbrough. 1990. Intraoperative monitoring. In Current practice of clinical electroencephalography, ed. D. D. Daly,
and T. A. Pedley, 2nd ed., 766. New York: Raven Press. By permission of Lippincott Williams & Wilkins.)
this setting. This is particularly true when there include low-level inhalation agents as well as
is a fusion procedure that may involve screw neuromuscular blockade.
placement. If there is encroachment upon the Although the major purpose of SEP moni-
root, neurotonic discharges will generally be toring is to help recognize subclinical changes
detected and the surgeon will assess to position that could herald new postoperative neural
and move or remove the screw or the device. deficit, SEP occasionally show improvement
Stimulation of the screw hole or pedicle screw when the procedure reduces spinal cord com-
may also be beneficial in this situation to assure pression or ischemia (or both). This is often
the pedicle wall has not been breached. seen with cervical stabilization in patients with
At the thoracic and cervical level, adjacent cervical rheumatoid arthritis (Fig. 44–12).
vertebrae commonly are wired together to
obtain the stabilization needed for bone heal- Key Points
ing and fusion. The risk to neural structures can
occur at any time during the procedure, but the • The specific type of spine surgery and
risk is particularly high during fixation. In this structures at risk define the modality and
setting both SEP and MEP monitoring is gen- location of neurophysiologic monitoring.
erally indicated. SEPs are recorded throughout ◦ Preservation of spinal cord function
these procedures and MEPs at critical times is optimized by monitoring TCES leg
when the spinal cord is at risk. As there is no MEP along with arm and leg SEP.
or little neuromuscular blockade in this setting, ◦ Recordings both proximal and distal to
the surgeon needs to be informed as there may the segment of spinal cord at risk will
be some movement of the patient. For stan- minimize false-negative recording.
dard decompression surgeries of the cervical or • Continuous assessment of potential sour-
thoracic spine when there is no significant neu- ces of error, as well as consideration of
rologic deficit, the risk of a catastrophic event is possible false-positive and false-negative
low and in this setting SEPs may be adequate. response alteration, is necessary during
In this setting the anesthetic regimen can monitoring of spine cases.
A B C
43.6
200 μV
5 μv
Figure 44–13. Simultaneous recording of SEPs from the scalp and arm and EMG recordings during an operation on a
cervical cord tumor with syrinx. A, Magnetic resonance image of the syrinx (oval). B, Tibial SEP were absent from the onset
of recording. The left median SEP was lost at 1:00 pm. The only postoperative deficit was loss of proprioceptive sensation
in the left arm. C, Neurotonic discharges warned the surgeon of irritation of the local root or anterior horn cells.
Cervical Spine Disease one showed substantial change in the SEP

responses.32 Although this study reported sen-
Monitoring during cervical spine surgery sitivity of SEP recording at only 25%, larger
requires initially determining which neural studies performing only SEP monitoring have
structures are at risk and the level of risk. reported sensitivity of 77% and specificity of
The major concern for most patients having 100% in cervical spine corpectomy.33 If the
an operation at the C1–C4 level is myelopa- cervical spine is unstable, monitoring should
thy. Preoperative neurologic deficit, a multi- begin before anesthesia, because positioning
level operation, upper cervical surgery, and the head after the patient has been anes-
instrumentation increase the risk. Monitoring thetized may compromise the spinal cord.
of tibial and median SEP is generally suffi- Monitoring of nerve roots with EMG should
cient unless the spinal cord is already com- be considered for patients having spine surgery
promised, in which case MEP monitoring may at the C5–T1 level if there is significant preop-
also be needed. The combination of MEP erative radiculopathy or if the roots or spinal
and SEP monitoring will again provide criti- nerves may be compromised during the opera-
cal information about the sensory and motor tion (Fig. 44–13).
pathways and in cases of either spinal cord
tumors or severe spondylotic myelopathy, the
combined method may be preferred. This is
especially true during an anterior approach Key Points
where the motor pathways may be at greater • Surgery for cervical spine pathology is
risk. In a study of 427 cervical procedures Hili- optimized to detect both central and
brand et al. found sensitivity and specificity of peripheral damage by
100% when monitoring myogenic MEP with ◦ EMG recordings for neurotonic dis-
a threshold value of 60% loss. Of 12 patients charges in myotomes at risk particularly
that showed a significant reduction or loss of if there is preexisting radiculopathy.
MEP, 10 were reported correctable with return ◦ TCES MEP through the cord segment
to normal values and no postoperative clinical at risk to both arm and leg muscles if
deficit whereas 2 remained absent and with there is clinical evidence of cord com-
clinical deficit. Of the two with deficit, only promise.
◦ Arm and leg SEP provide the greatest by several investigators. Pelosi et al.39 reported
coverage: ulnar SEP if the cord damage on 126 spinal operation where both SEP and
extends below C7; median SEP if above MEP were successfully recorded in 82% of
that level. cases. In this series they report four patients
• Bilateral tibial SEP with stimulation at the who had either no change in SEP or full recov-
knee, hip, or cauda equina may be needed ery of SEP who postoperatively had significant
for patients with preexisting neurologic lower limb weakness. In contrast, no new post-
deficit. operative motor deficits occurred in patients
who had either no change in MEP or MEP
that returned to baseline at completion of the
surgery. In seven patients only MEP changes
Thoracic Spine Disease and occurred and in three others the MEP changes
Scoliosis Surgery preceded SEP changes by several minutes.
Figure 44–14 shows a case where during scol-
Somatosensory evoked potential monitoring iosis repair there was relatively minimal change
is sufficient for most patients undergoing in SEP but loss of MEP with postoperative
thoracic spine surgery.34 However, because motor deficit.
motor function may be lost despite intact SEP,
a combination of MEP and SEP monitoring
has been recommended as the “standard of
care” for scoliosis surgery.35 In some academic Key Points
medical centers, MEP monitoring can easily
be applied and is used routinely.36 However, • Monitoring modalities should be selected
anesthetic and technical considerations make based on the risk of spinal cord damage in
MEP monitoring more difficult to apply in individual patients.
many settings. In these cases, MEP monitor- ◦ When the greater complexity of MEP
ing should be considered if there is marked monitoring make them more difficult to
deformity, preoperative deficit, anterior verte- monitor, SEP can be quite satisfactory.
brectomy, or other evidence of considerable ◦ Marked spine deformity, structural
risk to the spinal cord.37 compression of the spinal cord, and
The possible loss of sensory function alone clinical evidence of myelopathy all war-
is sufficient reason to monitor SEP even if rant MEP monitoring.
MEP monitoring is available.38 The utility of ◦ Sensory deficits warrant SEP moni-
combined monitoring of MEP and SEP dur- toring, even if MEP monitoring is
ing spinal correction surgery has been reported available.
Figure 44–14. A, Right and left tibial SEPs recorded over the scalp remained relatively stable despite loss of MEPs
recorded over the soleus. B, Loss of soleus MEP with electric stimulation of the cervical spinal cord after rod placement.
Temporary return with rod removal, but lost again with replacement. When rods were permanently removed, the patient
awoke with incomplete paraplegia which gradually cleared over a few months.
Lumbosacral Spine Disease et al.46 reported electrophysiologic monitoring

in these patients. Two to five fascicles are dis-
Often with surgery at the lumbosacral level, sected apart in each dorsal root and stimulated
spinal nerves are at greater risk than the spinal with single stimuli and trains of stimuli to elicit
cord itself. Because SEP are less effective than a reflex. The character of the responses bilat-
EMG in identifying nerve root damage, EMG erally in L2- to S2-innervated limb muscles is
monitoring is often an important part of mon- assessed to determine the contribution of each
itoring during procedures at the lumbosacral fascicle to the spasticity. It is critical to iden-
level.40 As these procedures often involve tify the motor root to avoid increased weakness
fusion, compound muscle action recordings from sectioning motor axons. The motor and
with drill or pedicle screw stimulation are often sensory roots are distinguished readily by mea-
used to assure proper placement. suring the threshold to single stimuli. Despite
many reports on the benefits of dorsal rhizo-
Key Points tomy, no controlled study has demonstrated
that electrophysiologic testing provides a bet-
• EMG provides optimal monitoring for the ter outcome than blindly sectioning 30%–50%
spinal nerves that are at greater risk. of the roots that innervate spastic muscles.
• SEP and MEP provide additional value
only if there is evidence of preexisting
lumbosacral cord damage. Key Points
• Stimulation of pedicle screws for hard-
ware placement provides warning of prox- • Section of lumbosacral dorsal rootlets has
imity to nerve roots that may be damaged. been reported to decrease spasticity.
• Monitoring the reflex motor response to
stimulation of individual sensory root fas-
cicles could theoretically allow section of
Primary Neural Disease those with the greatest contribution to
spasticity.
Operations on tumors of the spinal cord, par- ◦ High-frequency stimulation is applied
ticularly intramedullary tumors, are high-risk briefly.
procedures that can easily lead to further spinal ◦ Identification of more abnormal rootlets
cord damage. Intraoperative monitoring can depends on identifying spread of the
warn the surgeon of this damage which may reflex response beyond the root stimu-
lead to modification of the procedure. MEP lated, increasing response with contin-
and SEP monitoring maximize the possibil- ued stimulation, and continuation of the
ity of recognizing significant change.41 Some reflex after the stimulus stops.
groups have found that monitoring during • Success of the stimulation is highly depen-
intramedullary cord surgery decreases the fre- dent on anesthetic levels that do not sup-
quency of major complications.42–44 press all reflex responses.
Key Points
• Direct spinal cord surgery warrants both
Cauda Equina and Tethered Cord
SEP and MEP monitoring.
Surgical procedures below the spine of L1
pose a risk to the cauda equina rather than to
Dorsal Rhizotomy the spinal cord. The overlap of multiple roots
innervating individual dermatomes makes SEP
In patients with spasticity, especially chil- monitoring insensitive to nerve root damage in
dren with cerebral palsy, improvement in the the cauda equina. Therefore, monitoring cauda
spasticity and function reportedly has been equina function relies heavily on a combination
obtained by cutting a proportion of the fibers of EMG and NCS. The presence and distri-
in the L2–S1 dorsal roots, but the reports are bution of evoked responses in limb and anal
anecdotal.45 Although this procedure had been sphincter muscles with direct stimulation of tis-
known for many years, only recently has Staudt sue in the area of the cauda equina allow neural
tissue and specific roots to be identified. Con- ablation of arteriovenous malformations can
tinuous EMG monitoring of the same muscles be monitored with SEP or MEP if the sur-
can warn the surgeon of impending damage. geon deems the risk sufficient. Monitoring with
For example, neurotonic discharges indicate SEP or MEP during temporary occlusion of
that a ventral root is mechanically irritated and the major feeder vessels of an arteriovenous
motor unit potential firing indicates that a dor- malformation can indicate the risk of ablat-
sal root is irritated.28 The most difficult oper- ing the vessel by embolization. The difficulty
ations to perform in the region of the cauda of performing MEP monitoring has precluded
equina are those for congenital abnormalities. its use during embolization of an arteriovenous
Several congenital abnormalities of the lum- malformation.
bosacral cord and cauda equina can result in Monitoring is important in thoracoabdomi-
progressive neurologic deficit referred to as nal aortic aneurysm surgery because the risk of
tethered cord syndrome.47 The primary pur- paraplegia is as high as 15%. To decrease this
pose of electrophysiologic monitoring is to rate, the surgical procedure has been modified,
preserve neural tissue by identifying it and including spinal cord cooling, cerebrospinal
distinguishing it from other tissue that will fluid drainage, premedication, cross-clamping
be dissected or sectioned. Continuous EMG at short distances to minimize the segment
monitoring from multiple limb and sphinc- of spinal cord exposed to ischemia, femoral
ter muscles is the most effective method for bypass, and measurement of spinal cord blood
doing this by identifying mechanical irritation flow.
of neural tissue during dissection and immedi- For each of these, the combined func-
ately warning the surgeon of possible damage tional measures of SEP, MEP, and H reflexes
(Fig. 44–10). Direct stimulation of unidenti- (Fig. 44–11) allow distinguishing the effects
fied tissue elements will quickly identify it as of ischemia at the cortical (blood pressure or
neural tissue if a CMAP is evoked in a limb carotid occlusion), peripheral nerve (femoral
or sphincter muscle. In a recent study utilizing artery clamping), entire spinal cord (aorta
SEP and EMG monitoring of 44 consecutive clamping), and segmental spinal cord (seg-
adults undergoing surgery of tethered cord, mental spinal artery occlusion) levels.49 Ini-
SEP had a high specificity but low sensitivity tial studies of the benefit of revascularization
whereas EMG had a sensitivity of 100% but a using SEP as a guide were less successful than
low specificity of 19%.48 more recent studies using MEP.50, 51 The rapid
alterations that can occur with occlusion and
Key Points revascularization are shown in Figure 44–15.
MEPs and H reflexes recorded from periph-
• Cauda equina and sacral cord surgical
eral muscles are better indicators of danger-
monitoring relies on ous ischemia, because the motor neurons and
◦ EMG of L3- to S3-innervated muscles
synapses in the anterior horn are more sensi-
to record neurotonic discharges. tive to ischemia than the motor pathways in the
◦ Direct stimulation to identify individual
spinal cord.52
nerve roots.
◦ SEP to identify low spinal cord comp-
romise. Key Points
• The spinal cord is at risk of ischemia
during vascular malformation surgery
Vascular Diseases and for thoracoabdominal aneurysm
surgery.
In addition to its well-known use during • Monitoring can rapidly identify the occur-
cerebral aneurysm, carotid artery, and other rence of dangerous ischemia that will
cerebral vascular operations, electrophysio- result in permanent cord damage if not
logic monitoring is used during two proce- repaired quickly.
dures: thoracoabdominal aortic aneurysm and ◦ MEP and H-reflex monitoring showed
vascular malformation operations. These surg- loss of function within minutes that if
eries put the spinal cord at risk for loss of not reversed resulted in postoperative
blood supply and paraplegia. Direct surgical paralysis.
Figure 44–15. A loss of left and right anterior tibial MEPs, with electric stimulation of the cerebral cortex (transcranial
[Tc]-MEP), immediately after the aorta was clamped at level Th12 during thoracoabdominal aneurysm. (From Jacobs, M. J.
H. M., P. de Haan, S. A. Meylaerts, B. A. de Mol, and C. J. Kalkman. 2000. Benefits of monitoring motor-evoked potentials
during thoracoabdominal aortic aneurysm repair: Technique of choice to assess spinal cord ischemia? In Perspectives in
vascular surgery, ed. P. Gloviczki, and J. Goldstone, Vol. 12, 1–16. New York: Thieme Medical Publishers. By permission
of the publisher.)
◦ SEP was less effective in identifying cord injury is caused by a vascular insult, with
these changes. purely motor damage, SEP monitoring can
◦ Leg ischemia from femoral occlusion identify the damage early enough to alert the
results in a much slower loss of both surgeon. The addition of MEP monitoring fur-
MEP and SEP peripherally. ther protects the motor pathways that may be
at risk during some spinal procedure. Neu-
rotonic discharges recorded from peripheral
muscle are sensitive to nerve root irritation
SUMMARY and, thus, can help surgeons recognize when
and where damage may be occurring. These
Continuous electrophysiologic monitoring of techniques appear reliable and with experi-
spinal cord or spinal nerve (or both) function ence the neurophysiologist can acquire the
intraoperatively can minimize potential dam- skills to perform and correctly interpret these
age that may occur during spine surgery. SEPs studies thus enhancing the neurologic and
are easiest to use for monitoring function and functional outcomes during the often complex
have had the widest application. Unless spinal procedures.
REFERENCES 17. Chen, X., D. Sterio, X. Ming, et al. 2007. Success

rate of motor evoked potentials for intraoperative neu-
1. Daube, J. R. 1999. Intraoperative monitoring reduces rophysiologic monitoring: Effects of age lesion loca-
complications and is therefore useful. Muscle & Nerve tion and preoperative neurological deficits. Journal of
22:1151–3. Clinical Neurophysiology 24(3):281–5.
2. Guerit, J. M. 1998. Neuromonitoring in the operat- 18. Levy, W., D. York, M. McCaffrey, and F. Tanzer. 1984.
ing room: Why, when, and how to monitor? Elec- Motor evoked potentials from transcranial stimula-
troencephalography and Clinical Neurophysiology tion of the motor cortex in humans. Neurosurgery
106:1–21. 15:214–27.
3. Herdmann, J., V. Deletis, H. L. Edmonds Jr., and N. 19. Levy, W. 1987. Clinical experience with motor and
Morota. 1996. Spinal cord and nerve root monitor- cerebellar evoked potential monitoring. Neurosurgery
ing in spine surgery and related procedures. Spine 20:169–82.
21:879–85. 20. Zentner, J. 1989. Noninvasive motor evoked poten-
4. MacEwen, G. D., W. P. Bunnell, and K. Sriram. 1975. tial monitoring during neurosurgical operations on the
Acute neurological complications in the treatment of spinal cord. Neurosurgery 24:709–12.
scoliosis. A report of the Scoliosis Research Society. 21. Kitagawa, H., T. Itog, H. Takano, et al. 1989. Motor
The Journal of Bone and Joint Surgery (American) evoked potentials monitoring during upper cervical
57:404–8. spine surgery. Spine 14:1078–83.
5. Burke, D., M. R. Nuwer, J. Daube, et al. 1999. Intra- 22. Owen, J., K. Bridwell, R. Grubb, et al. 1991. The clin-
operative monitoring. The International Federation ical application of neurogenic motor evoked potentials
of Clinical Neurophysiology. Electroencephalography to monitor spinal cord function during surgery. Spine
and Clinical Neurophysiology 52(Suppl):133–48. 16:S384–90.
6. Nuwer, M. R., J. R. Daube, C. Fischer, J. Schramm, 23. Holland, N. R. 1998. Intraoperative electromyog-
and C. D. Yingling. 1993. Neuromonitoring during raphy during thoracolumbar spinal surgery. Spine
surgery. Electroencephalography and Clinical Neuro- 23:1915–22.
physiology 87:263–76. 24. Harper, C. M. Jr., J. R. Daube, W. J. Litchy, and R.
7. Nuwer, M. R., E. G. Dawson, L. G. Carlson, L. E. A. Klassen. 1988. Lumbar radiculopathy after spinal
Kanim, and J. E. Sherman. 1995. Somatosensory fusion for scoliosis. Muscle & Nerve 11:386–91.
evoked potential spinal cord monitoring reduces neu- 25. Obi, T., M. Mochizuki, K. Isobe, K. Mizoguchi,
rologic deficits after scoliosis surgery: Results of a M. Takatsu, and Y. Nishimura. 1999. Mechani-
large multicenter survey. Electroencephalography and cally elicited nerve root discharge: Mechanical irri-
Clinical Neurophysiology 96:6–11. tation and waveform. Acta Neurologica Scandinavica
8. Sloan, T. 1998. Anesthetic effects on electrophysio- 100:185–8.
logic recordings. Journal of Clinical Neurophysiology 26. Daube, J., and C. Harper. 1989. Surgical monitoring of
15(3):217–26. cranial and peripheral nerves. In Neuromonitoring in
9. Lauer, K., C. Munshi., and S. Larson. 1994. The effect surgery, ed. J. E. Desmedt, Vol. 1, 115–38. New York:
of midazolam on median nerve somatosensory evoked Elsevier.
potentials. Journal of Clinical Monitoring 10(3):181–4. 27. Harper, C., and J. Daube. 1998. Facial nerve elec-
10. Laureau, E., B. Marciniak, A. Hebrard, B. Herbaux., tromyography and other cranial nerve monitoring.
and J. D. Guieu. 2004. Comparative study of propo- Journal of Clinical Neurophysiology 15(3):206–16.
fol and midazolam effects on somatosensory evoked 28. Kothbauer, K., U. D. Schmid, R. W. Seiler, and W.
potentials during surgical treatment of scoliosis. Neu- Eisner. 1994. Intraoperative motor and sensory moni-
rosurgery 45(1):69–74. toring of the cauda equina. Neurosurgery 34:702–7.
11. Banoub, M., J. E. Tetzlaff, and A. Schubert. 29. Calancie, B., P. Madsen, and N. Lebwohl. 1994.
2003. Pharmacologic and physiologic influences Stimulus-evoked EMG monitoring during transpedic-
affecting sensory evoked potentials. Anesthesiology ular lumbosacral spine instrumentation. Initial clinical
99(3):716–37. results. Spine 19(24):2780–6.
12. Nuwer, M. R. 1999. Spinal cord monitoring. Muscle & 30. Maguire, J., S. Wallace, R. Madiga, R. Leppanen, and
Nerve 22:1620–30. V. Draper. 1995. Evaluation of intrapedicular screw
13. Lesser, R. P., P. Raudzens, H. Lüders, et al. 1986. position using intraoperative evoked electromyogra-
Postoperative neurological deficits may occur despite phy. Spine 20(9):1068–74.
unchanged intraoperative somatosensory evoked 31. Moed, B. R., B. K. Ahmad, J. G. Craig, G. P. Jacob-
potentials. Annals of Neurology 19:22–5. son, and M. J. Anders. 1998. Intraoperative monitoring
14. Burke, D., and R. G. Hicks. 1998. Surgical monitoring with stimulus-evoked electromyography during place-
of motor pathways. Journal of Clinical Neurophysiol- ment of iliosacral screws. An initial clinical study.
ogy 15:194–205. The Journal of Bone and Joint Surgery American
15. Sloan, T. B., and E. J. Heyer. 2002. Anesthe- 80:537–46.
sia for intraoperative neurophysiologic monitoring of 32. Hilibrand, A. S., D. M. Schwartz, V. Sethuraman,
the spinal cord. Journal of Clinical Neurophysiology A. R. Vaccaro, and T. J. Albert. 2004. Comparison of
19(5):430–43. transcranial electric motor and somatosensory evoked
16. Langeloo, D., M. A. Lelivelt, H. L. Journee, R. Slap- potential monitoring during cervical spine surgery.
pendel, and M. de Kleuver. 2003. Transcranial electri- The Journal of Bone and Joint Surgery (American)
cal motor-evoked potential monitoring during surgery 86(6):1248–53.
for spinal deformity. A study of 145 patients. Spine 33. Khan, M. H., P. N. Smith, J. R. Balzer, et al.
28(10):1043–50. 2006. Intraoperative somatosensory evoked potential
monitoring during cervical spine corpectomy surgery. monitoring: Basic principles, regeneration, patho-
Experience with 508 cases. Spine 31(4):E105–13. physiology, and clinical aspects, ed. E. Stålberg,
34. Nuwer, M. R. 1998. Spinal cord monitoring with H. S. Sharma, and Y. Olsson, 509–20. Wien/New York:
somatosensory techniques. Journal of Clinical Neuro- Springer-Verlag.
physiology 15:183–93. 45. Gul, S. M., P. Steinbok, and K. McLeod. 1999. Long-
35. Padberg, A. M., T. J. Wilson-Holden, L. G. Lenke, term outcome after selective posterior rhizotomy in
and K. H. Bridwell. 1998. Somatosensory- and motor- children with spastic cerebral palsy. Pediatric Neuro-
evoked potential monitoring without a wake-up test surgery 31:84–95.
during idiopathic scoliosis surgery. An accepted stan- 46. Staudt, L. A., M. R. Nuwer, and W. J. Peacock.
dard of care. Spine 23:1392–400. 1995. Intraoperative monitoring during selective pos-
36. Burke, D., R. Hicks, J. Stephen, I. Woodforth, and terior rhizotomy: Technique and patient outcome.
M. Crawford. 1992. Assessment of corticospinal and Electroencephalography and Clinical Neurophysiol-
somatosensory conduction simultaneously during sco- ogy 97:296–309.
liosis surgery. Electroencephalography and Clinical 47. McQuillan, P. M., and N. Newberg. 1995. Intraop-
Neurophysiology 85:388–96. erative electromyography. In Primer of intraopera-
37. Deutsch, H., M. Arginteanu, K. Manhart, et al. 2000. tive neurophysiologic monitoring, ed. G. B. Russell,
Somatosensory evoked potential monitoring in ante- and L. D. Rodichok, 171–87. Boston: Butterworth-
rior thoracic vertebrectomy. Journal of Neurosurgery Heinemann.
92(Suppl):155–61. 48. Paradiso, G., G. Y. Lee, R. Sarjeant, L. Hoang,
38. Lorenzini, N. A., and J. H. Schneider. 1996. Tem- E. M. Massicotte, and M. G. Fehlings. 2006. Mul-
porary loss of intraoperative motor-evoked poten- timodal intraoperative neurophysiologic monitoring
tial and permanent loss of somatosensory-evoked findings during surgery for adult tethered cord syn-
potentials associated with a postoperative sensory drome: Analysis of a series of 44 patients with long-
deficit. Journal of Neurosurgery and Anesthesiology 8: term follow-up. Spine 31(18):2095–102.
142–7. 49. Sueda, T., K. Okada, M. Watari, K. Orihashi,
39. Pelosi, L., J. Lamb, M. Grevitt, S. M. H. Mehdian, H. Shikata, and Y. Matsuura. 2000. Evaluation of
J. K. Webb, and L. D. Blumhardt. 2002. Combined motor- and sensory-evoked potentials for spinal cord
monitoring of motor and somatosensory evoked poten- monitoring during thoracoabdominal aortic aneurysm
tials in orthopaedic spine surgery. Clinical Neurophys- surgery. The Japanese Journal of Thoracic and Cardio-
iology 113:1082–91. vascular Surgery 48:60–5.
40. Weiss, D. S. 2001. Spinal cord and nerve root mon- 50. Galla, J. D., M. A. Ergin, S. L. Lansman, et al. 1999.
itoring during surgical treatment of lumbar stenosis. Use of somatosensory evoked potentials for thoracic
Clinical Orthopaedics 384:82–100. and thoracoabdominal aortic resections. The Annals of
41. Nagle, K. J., R. G. Emerson, D. C. Adams, et al. Thoracic Surgery 67:1947–52.
1996. Intraoperative monitoring of motor evoked 51. Jacobs, M. J. H. M., P. de Haan, S. A. Meylaerts,
potentials: A review of 116 cases. Neurology 47: B. A. de Mol, and C. J. Kalkman. 2000. Benefits
999–1004. of monitoring motor-evoked potentials during tho-
42. Jones, S. J., R. Harrison, K. F. Koh, N. Mendoza, and racoabdominal aortic aneurysm repair: Technique of
H. A. Crockard. 1996. Motor evoked potential moni- choice to assess spinal cord ischemia? In Perspectives
toring during spinal surgery: Responses of distal limb in vascular surgery, ed. P. Gloviczki, and J. Gold-
muscles to transcranial cortical stimulation with pulse stone, Vol. 12, 1–16. New York: Thieme Medical
trains. Electroencephalography and Clinical Neuro- Publishers.
physiology 100:375–83. 52. Shine, T. S., B. A. Harrison, M. L. De Ruyter,
43. Kothbauer, K., V. Deletis, and F. J. Epstein. et al. 2008. Motor and somatosensory evoked
1997. Intraoperative spinal cord monitoring for potentials: Their role in predicting spinal cord
intramedullary surgery: An essential adjunct. Pediatric ischemia in patients undergoing thoracoab-
Neurosurgery 26:247–54. dominal aortic aneurysm repair with regional
44. Tamaki, T. 1998. Intraoperative spinal cord lumbar epidural cooling. Anesthesiology 108(4):
monitoring—Clinical overview. In Spinal cord 580–7.
Chapter 45
Peripheral Nervous System

Monitoring
INTRODUCTION APPLICATIONS
METHODS Entrapment Neuropathies
Nerve Conduction Studies Repair of Traumatic Peripheral
Electromyography Nerve Injury
Somatosensory Evoked Potentials Prevention of Injury during
and Motor Evoked Potentials Peripheral Nerve Surgery
SUMMARY
grafting, or neurotization of nerves. With

minor modifications, standard techniques of
INTRODUCTION electrodiagnosis such as nerve conduction
studies (NCS), electromyography (EMG), and
Electrophysiological peripheral nervous sys-
somatosensory evoked potentials (SEPs) are
tem monitoring is performed during surgery
used to monitor the peripheral nerves dur-
for entrapment neuropathies, nerve trauma,
ing surgery. Appropriate monitoring protocols
and primary or metastatic neoplasms that
can be designed for each patient after the
involve peripheral nerves. Monitoring helps
findings of the preoperative neurological exam-
locate and preserve peripheral nerve func-
ination, NCS, EMG, and surgical goals are
tion, particularly in situations where nor-
reviewed.
mal anatomy is distorted by pathology. In
cases of nerve repair, information obtained
from intraoperative electrophysiological stud- Purpose and Role of Peripheral Nerve
ies supplements preoperative studies providing Monitoring
more precise data regarding the location
and severity of lesions as well as the status • Supplements information gained from
of natural repair mechanisms. This informa- preoperative electrodiagnostic studies.
tion helps guide therapeutic decision-making • Helps locate nerves when pathology dis-
with regard to decompression, neurolysis, torts normal anatomy.
777
• Assesses the number, location, severity of

lesions, and status of natural repair.
• Guides surgical decision-making.
METHODS
Nerve Conduction Studies

Figure 45–2. Hooked electrode that can be used for
Intraoperative NCS are similar in principle stimulation or recording from nerves exposed at surgery.
and technique to those used for standard elec- The recording surface is exposed along the concavity of the
trodiagnostic studies.1 Mixed motor and sen- hook. The electrodes are set 10 mm apart.
sory nerves or sensory cutaneous nerves are
stimulated with a hand-held bipolar electri-
cal stimulator placed directly onto the nerve and conduction velocity). Stimulus threshold
within the surgical field (Fig. 45–1). The size of should be monitored closely as diseased nerves
the stimulator is matched to the size of nerve. have higher thresholds than normal nerves.3
Larger stimulators similar to those used in rou- Orientation of the stimulator is also impor-
tine NCS are used to stimulate large nerve tant. Bipolar stimulation produces a more
trunks that are isolated from other nerves. focal distribution of current than monopolar
Hooked stimulating electrodes can be used stimulation and is therefore preferred for
to elevate nerves from the surrounding tis- intraoperative stimulation of most periph-
sue when better stimulus isolation is required eral nerves. The cathode and anode should
(Fig. 45–2). Small electrodes are used when be aligned with the long axis of the nerve
individual or small groups of fascicles are and the cathode should be proximal to the
stimulated. Occasionally, monopolar stimula- desired direction of the current. Peripheral
tion is used if there is insufficient exposure nerve ischemia produces transient conduction
to apply a bipolar stimulator to the nerve. block in affected nerves. To prevent this from
Intraoperative stimulation requires careful affecting electrophysiologic recordings, tourni-
consideration of the location and strength of quet should be avoided or removed at least
stimulus. The normal nerve is activated with 30 minutes before the anticipated time of
as little current as 1–5 mA of 0.5 ms dura- recording.
tion.2 Over-stimulation produces unintended Compound muscle action potentials
stimulation of the surrounding nerves or stim- (CMAPs) are recorded from surface or sub-
ulus spread along the course of the nerve cutaneous needle electrodes placed over distal
(leading to inaccurate calculation of latency muscles. Compound nerve action potentials
(NAPs) are recorded from large mixed nerves
with a hand-held bipolar recorder placed
directly on the nerve or from small cutaneous
nerves with needle electrodes placed next
to the nerve.2 CMAP recordings are larger,
less contaminated with artifact, and exclusively
monitor the function of motor axons.
NAP recordings are technically more diffi-
cult to perform and typically require averag-
ing of 8–10 potentials.2, 4 A distance of 5–8 cm
between stimulating and recording electrodes
is typically required to record a reliable NAP
free from excessive stimulus artifact. Tempera-
Figure 45–1. Bipolar stimulating electrode used for ture of exposed peripheral nerves is typically
direct stimulation of nerves exposed at surgery. The anode lower than normal but cannot be controlled
and cathode are separated by a distance of 10 mm. during intraoperative studies. Although low
Peripheral Nervous System Monitoring 779
temperatures prolong absolute latencies and potentials (neurotonic discharge) that can be
increase durations and amplitudes of NAPs easily distinguished from artifact and other
recorded during surgery, this rarely alters the motor unit potential activity in EMG record-
reliability of these techniques. NAPs are more ings.6 Neurotonic discharges are used to locate
sensitive than CMAPs in localizing abnormali- nerves within the surgical field and to warn
ties along nerves and better able to detect early of the potential for nerve injury should the
axonal regeneration through an area of nerve irritation continue.9
injury long before regeneration is reflected in
CMAP recordings.5 The presence of CMAPs
or NAPs indicates that some axons are in con-
tinuity; amplitude and area of the response Somatosensory Evoked Potentials
are proportional to the number of function- and Motor Evoked Potentials
ing axons. Focal slowing of conduction velocity,
conduction block, or increased threshold of SEPs and MEPs are typically used to study
stimulation localizes pathology along the nerve conduction in central nervous system and are
with 1–2 cm accuracy.6–8 also used in peripheral nerve monitoring dur-
ing surgery. The intraoperative SEP techniques
are virtually identical to those used in the
Electromyography electrodiagnostic laboratory. Averaged poten-
tials are recorded from surface electrodes over
EMG activity can be recorded during surgery the cervical spine and scalp following direct
with small monopolar needles or small intra- electrical stimulation of peripheral nerve ele-
muscular wire electrodes (Fig. 45–3).2, 7 The ments in the surgical field. The presence
wires are introduced percutaneously with a of a response indicates continuity of sensory
hollow needle, which is then withdrawn leav- axons between the spinal cord and the site
ing the wire in place. When intramuscu- of peripheral nerve stimulation. For MEP,
lar electrodes are used, electrical stimulation transcranial electrical stimulation is used to
of the nerve produces a polyphasic EMG activate the motor cortex. This is best per-
response that, although difficult to quanti- formed with a specially designed, commer-
tate, is less likely than surface electrodes to cially available constant voltage stimulator that
record nonspecific activity from adjacent mus- delivers a short-duration stimulus with rapid
cles. Mechanical irritation of the nerve pro- rise time in a train of 3–4 stimuli at 2–3 ms
duces a high-frequency discharge of motor unit intervals.2, 7 Stimulus intensities range from
500–800 V. MEP recordings are made from
hook electrodes applied directly to surgically
exposed nerves or from muscles. Stimulus arti-
fact is common but can be eliminated by
averaging 4–6 stimuli of alternating polarity.
When NAPs are recorded directly from the
nerve, neuromuscular blockade is required to
eliminate the possibility of recording artifact
from muscles in the region of the recording
electrodes.
Key Points
• With minor modifications, standard elec-
trodiagnostic techniques (NCS, EMG,
SEP, MEP) are easily applied to operative
Figure 45–3. Nichrome intramuscular wires used for
intramuscular EMG recordings. The wire sits inside the setting.
core of a small hypodermic needle. The needle is used to • Intraoperative NCS require specially
insert the wire into the muscle. The wires remain in places designed sterile stimulating and recording
after the needle is withdrawn. electrodes.
• MEP recordings are performed with is unclear, NCS performed during surgery can
a specially designed constant voltage frequently add valuable localizing information.
transcortical stimulator. This typically occurs when the site of entrap-
• Stimulus artifact can be managed with ment is deep or in an unusual location (e.g.,
good technique, averaging, and alternat- median and radial nerves in forearm and sciatic
ing stimulus polarity. nerve in buttock) or when the lesion is predom-
inately axonal (e.g., ulnar neuropathy at the
elbow and peroneal neuropathy at the knee).
In the absence of conduction block, there is
no change in CMAP amplitude between prox-
APPLICATIONS imal and distal stimulation sites. Despite the
lack of obvious conduction block on CMAP
Entrapment Neuropathies recordings, a segmental change in NAP ampli-
tude or a focal slowing of conduction is fre-
NCS are utilized in selected cases to improve quently observed even in purely axonal lesions
localization during surgery for various entrap- (Fig. 45–4). To demonstrate this, a bipolar
ment neuropathies,6–8 including ulnar neu- hooked nerve electrode is placed proximal to
ropathy at the elbow and wrist, median the lesion for NAP recording and stimulat-
neuropathy in the forearm, radial neuropathy ing electrodes are applied distal to the seg-
in the arm or forearm, peroneal neuropathy at ment being tested. Proximal placement of the
the knee, and sciatic neuropathy in the but- recording electrode eliminates movement arti-
tock. Little additional information is added by fact and CMAPs caused by contraction of adja-
doing intraoperative NCS in cases where pre- cent muscles. The stimulator is moved from
operative studies demonstrate focal conduction distal to proximal in successive 1–2 cm seg-
block or slowing over a 2–3 cm segment of ments. The lesion is identified by a change
nerve. However, when the site of entrapment in amplitude and slowed conduction of the
Figure 45–4. Intraoperative recording of CMAPs and NAPs during ulnar nerve exploration and stimulation at 1-cm
intervals. The “0” point indicates the location of the medial epicondyle. The greatest change in latency and amplitude
occurred over a 3-cm segment spanning the origin of the cubital tunnel. FCU, flexor carpi ulnaris; Hypo, hypothenar.
(From Daube, J. R., and C. M. Harper. 1989. Surgical monitoring of cranial and peripheral nerves. In Neuromonitoring in
surgery, ed. J. E. Desmedt, 115–51. Amsterdam: Elsevier. By permission of Elsevier Science Publishers.)
NAP over the short nerve segment. NAPs may status of axonal continuity across the injured
provide useful information before regeneration area.
has had a chance to reach the muscle, a time The utility of these techniques is best
when CMAPs are absent. illustrated by examining their role in the sur-
Localization by clinical signs and preopera- gical repair of traumatic injuries to the brachial
tive NCS and EMG is adequate in the major- plexus.2, 7, 13 The complexity of brachial plexus
ity of cases of entrapment neuropathy.1, 2, 6, 8 anatomy, multiplicity and severity of injury to
Carpal tunnel release and most cases of ulnar its elements, and frequent occurrence of nerve
transposition do not benefit from intraopera- root avulsion make lesions of this structure par-
tive NCS. However, complicated cases of ulnar ticularly difficult to evaluate and treat.2, 7, 13, 14
or median neuropathy, or radial, femoral, sci- The presence or absence of nerve root avul-
atic, tibial, and peroneal neuropathies are often sion is one of the most important factors in
monitored because of inherent difficulties in determining prognosis and the need for surgi-
defining the number and location of lesions cal intervention in brachial plexus injuries. If
with preoperative studies in these cases.8, 10–12 root avulsion is present, then repair of post-
Conduction block and focal slowing are eas- ganglionic elements innervated by the avulsed
ier to detect and localize accurately when the root will be of no benefit. The clinical examina-
nerve is exposed. In addition, over-stimulation tion, NCS, needle EMG, and myelography are
is easier to detect and correct intraoperatively used preoperatively to assess the integrity of
and areas of increased threshold help iden- the cervical nerve roots.2, 7 The combination of
tify damaged nerve segments. The sensitiv- Horner’s syndrome, denervation of paraspinal
ity and accuracy of intraoperative NCS may and other proximal muscles, preserved sen-
help the surgeon choose the most appropri- sory nerve action potentials (SNAPs), and the
ate treatment. These principles are illustrated presence of a meningocele on myelography
in Figure 45–4, which summarizes the findings in the setting of a paralyzed anesthetic limb
of intraoperative NCS during a case of ulnar strongly suggests the presence of multiple root
nerve exploration. Preoperative NCS revealed avulsions. However, any one of these find-
localized slowing with increased CMAP disper- ings in isolation is less predictive. Examples
sion, approximately 3 cm distal to the medial of false-positive and false-negative myelograms
epicondyle. CMAPs were recorded intraoper- have been reported.13, 14 Because the posterior
atively over the abductor digiti minimi and primary ramus of a given nerve root inner-
flexor carpi ulnaris muscles and NAPs were vates paraspinal muscles at multiple levels, the
recorded from the proximal ulnar nerve. There distribution of fibrillation potentials may over-
were changes in amplitude and latency of both estimate the number of roots involved. In addi-
CMAPs and NAPs over a 3-cm segment at the tion, the presence of a postganglionic lesion
origin of the cubital tunnel. There were no with diminished SNAP may mask an associated
areas of slowing proximal or distal to this point, lesion involving preganglionic segments. The
so a cubital tunnel release was performed. predictive value of preoperative SEP in detect-
A similar strategy for improved localization is ing root continuity as well as the presence of
applicable to any nerve that can be exposed at mixed pre- and postganglionic lesions has been
suspected sites of entrapment. disappointing.15
These uncertainties can usually be resolved
by performing NCS, and SEP and MEP
recordings during surgery.15–19 Partial lesions
Repair of Traumatic Peripheral are incomplete and associated with residual
Nerve Injury motor and/or sensory axonal continuity. Intra-
operative recordings help determine the loca-
Intraoperative monitoring is particularly use- tion and severity of partial lesions in brachial
ful when multiple, deep, proximal nerves are plexopathy. Using direct stimulation of plexus
injured and when multiple potential mecha- elements exposed at surgery, the fascicles of
nisms of injury are involved (traction, con- interest are isolated allowing more precise seg-
tusion, ischemia, etc.). The primary purpose mental localization of partial lesions. In partial
of monitoring in this setting is to local- lesions, CMAPs, NAPs, and SEPs all produce
ize the injured segment(s) and assess the recordable responses across and distal to the
Figure 45–5. Electrophysiologic techniques for monitoring brachial plexopathy. (Upper right) SEPs recorded over scalp
during root stimulation. (Center) NAPs recorded directly (DNAP) form short segments of the plexus. (Lower left)
CMAPs recorded from distal muscles during selective stimulation of plexus elements. S, stimulating electrodes; R, record-
ing electrodes. (From Daube, J. R., and C. M. Harper. 1989. Surgical monitoring of cranial and peripheral nerves.
In Neuromonitoring in surgery, ed. J. E. Desmedt, 115–51. Amsterdam: Elsevier. By permission of Elsevier Science
Publishers.)
injured segment (Figs. 45–5, 45–6, and 45–7). plexus elements following transcranial elec-
The size of the response is proportional to trical stimulation (i.e., MEP) provides more
the number of functioning axons. Most partial direct evidence of ventral root continuity
lesions are left alone or treated with external (Fig. 45–8). NAPs recorded after peripheral
neurolysis (removal of scar tissue). Complete nerve stimulation can also be used to assess
lesions are associated with total axonal inter- root integrity indirectly but the proper inter-
ruption at the time of injury. When brachial pretation of NAP in the setting of root avulsion
plexus surgery is carried out several months requires experience and is not always defini-
postinjury, complete lesions can be subdi- tive. Total absence of the NAP while stimulat-
vided into those that show signs of regener- ing the root and recording it 4 cm distal along
ation across the injured segment and those the spinal nerve or proximal trunk suggests root
that do not. avulsion, but a very proximal postganglionic
First and most importantly, SEPs16–19 and lesion could also produce this finding. Also, in
MEPs2, 7, 15 are used to determine the presence root avulsion, if the postganglionic element is
or absence of nerve root avulsion. A well- intact, a well-defined NAP will be recorded
defined SEP recorded from the brain or spinal from preserved postganglionic sensory fibers.
cord after direct stimulation of the exposed Once nerve root continuity is confirmed,
nerve root indicates central continuity of the intraoperative NCS and SEPs can be used to
dorsal root while the absence of a response assess the integrity of more distal elements
confirms the presence of root avulsion at that of the brachial plexus. Detecting conduction
level (Fig. 45–6). In most cases, this indicates block or focal slowing of conduction veloc-
a high likelihood of ventral root continuity ity over short segments in either CMAP or
as well. However, recording of an NAP from NAP recordings sometimes localizes lesions.
typically has yet to reach a target muscle

(Fig. 45–9). When regeneration is detected,
the segment is left alone or neurolysis is
performed. If the NAP and/or SEP are
absent, then grafting is performed, as long
as proximal nerve root continuity is con-
firmed. Recording the NAP from individ-
ual nerve fascicles has been reported to
help guide partial fascicular repair in the
setting of incomplete lesions.12, 13 Technical
difficulties related to the size of the elec-
trodes, occurrence of shock artifact, and avoid-
ance of current spread to adjacent fascicles
have limited the widespread use of fascicular
recordings.
Figure 45–6. Intraoperative recording of SEPs over
scalp with stimulation of cervical nerve roots directly in
surgical field. Well-defined responses were seen with stim-
ulation of the C-7 and C-8 roots. No response was obtained Prevention of Injury during
with stimulation of C-5 or C-6 roots, indicating avul-
sion of the root at these levels. (From Daube, J. R., and
Peripheral Nerve Surgery
C. M. Harper. 1989. Surgical monitoring of cranial and
peripheral nerves. In Neuromonitoring in surgery, ed. As with the cranial nerves, the peripheral
J. E. Desmedt, 115–51. Amsterdam: Elsevier. By permis- nerves that innervate the trunk and extrem-
sion of Elsevier Science Publishers.) ity muscles are susceptible to mechanical or
ischemic injury during a variety of surgical
NAP and SEP recordings are better suited procedures. EMG and SEP monitoring have
than CMAP recordings in detecting early been utilized in an attempt to prevent injury to
regeneration since reinnervation at this stage the phrenic nerve20 and brachial plexus during
Figure 45–7. Intraoperative NCS showing conduction block along medial cord of brachial plexus. (From Daube, J. R.,
and C. M. Harper. 1989. Surgical monitoring of cranial and peripheral nerves. In Neuromonitoring in surgery, ed.
J. E. Desmedt, 115–51. Amsterdam: Elsevier. By permission of Elsevier Science Publishers.)
stimulated mechanically or electrically while

monitoring EMG activity in distal muscles. An
attempt is made to preserve fascicles that pro-
duce a distal EMG response while those that
do not are sacrificed.
Key Points
Figure 45–8. Transcranial motor evoked potential • Intraoperative NCS improve localization
(MEP) recorded directly from the C5 nerve root during during surgery for selected entrapment
surgical repair of traumatic brachial plexopathy. The
NAP is recorded with hook electrodes applied to the C5
neuropathies (axonal lesions or entrap-
root. It represents an averaged response to six trains of ment in deep or proximal locations).
transcranial electrical stimuli (4 pulses per train at 3 ms • In brachial plexus repair, SEPs and MEPs
intervals). The well-defined MEP indicates that the C5 accurately assess the presence of partial or
motor root is intact. complete nerve root avulsion.
• NAP recordings over short segments
cardiac and other major surgical procedures.21 assess the presence of axonal continu-
During these procedures, EMG monitoring for ity and early regeneration across lesions
neurotonic discharges helps localize and warn affecting peripheral nerve elements.
of potential injury to the nerve, while direct • Intraoperative NCS help identify nerves
electrical stimulation with a handheld stimula- and functioning fascicles within nerves
tor is used to identify viable nerves or fascicles. during tumor resection or other pathologi-
SEPs can detect early conduction block asso- cal conditions that distort normal anatomy.
ciated with compression or ischemia of limb
nerves caused by improper positioning of limbs
during surgery. Orthopedic procedures that SUMMARY
involve disarticulation or extensive manipula-
tion of the limbs are often monitored. EMG Electrophysiological monitoring of the periph-
and/or SEP monitoring have been reported as a eral nervous system provides useful infor-
means to help detect and prevent injury to the mation that supplements and complements
axillary and musculocutaneous nerves during preoperative assessment. Monitoring improves
shoulder surgery, and to the femoral, obtu- localization and understanding of the underly-
rator, and sciatic nerves during high-risk hip ing pathophysiology of peripheral nerve lesions
surgery.22 leading to more rational treatment decisions
EMG monitoring may also be useful during and potentially improved outcomes. Monitor-
the resection of primary or metastatic periph- ing is accomplished by adaptation of routine
eral nerve neoplasms.14 In this setting, the goal electrodiagnostic techniques (i.e., nerve con-
is to resect the tumor with as little damage duction studies, evoked potentials, and elec-
to normal nerve fascicles as possible. During tromyography) with special attention to tech-
dissection, individual or groups of fascicles are nical factors including electrical and move-
ment artifact. These techniques have been
successfully applied during surgery for entrap-
ment neuropathies, traumatic nerve injury and
repair, and during procedures that risk periph-
eral nerve injury.
REFERENCES
1. Brown, W. F., and J. Veitch. 1994. AAEM minimono-
Figure 45–9. NAP recorded directly from the middle graph #42: Intraoperative monitoring of peripheral
trunk of the brachial plexus during surgical repair of trau- and cranial nerves. Muscle & Nerve 17:371–7.
matic brachial plexopathy. A well-defined NAP (arrow) 2. Crum, B. A., and J. A. Strommen. 2007. Peripheral
indicates axonal continuity across the suspected area of nerve stimulation and monitoring during operative
injury. procedures. Muscle & Nerve 35:159–710.
3. Mandpe, A. H., A. Mikulec, and R. K. Jackler. 1998. 15. Turkof, E., H. Millesi, R. Turkof, et al. 1997.
Comparison of response amplitude versus stimula- Intraoperative electroneurodiagnostics (tran-
tion threshold in predicting early postoperative facial scranial electrical motor evoked potentials) to
nerve function after acoustic neuroma resection. The evaluate the functional status of anterior spinal
American Journal of Otology 19:112–17. roots and spinal nerves during brachial plexus
4. Tiel, R. L., L. T. Happel Jr., and D. G. Kline. 1996. surgery. Plastic and Reconstructive Surgery 99:
Nerve action potential recording method and equip- 1632–41.
ment. Neurosurgery 39:103–8. 16. Yiannikas, C., B. T. Shahani, and R. R. Young.
5. Oberle, J. W., G. Antoniadis, S. A. Rath, and 1983. The investigation of traumatic lesions of the
H. P. Richter. 1997. Value of nerve action potentials brachial plexus by electromyography and short latency
in the surgical management of traumatic nerve lesions. somatosensory potentials evoked by stimulation of
Neurosurgery 41:1337–42. multiple peripheral nerves. Journal of Neurology, Neu-
6. Daube, J. R., and C. M. Harper. 1989. Surgical rosurgery, and Psychiatry 46:1014–21.
monitoring of cranial and peripheral nerves. In Neu- 17. Landi, A., S. A. Copeland, C. D. Wynn-Parry,
romonitoring in surgery, ed. J. E. Desmedt, 115–51. et al. 1980. The role of somatosensory evoked
Amsterdam: Elsevier. potentials and nerve conduction studies in the
7. Harper, C. M. 2005. Preoperative and intraopera- surgical management of brachial plexus injuries.
tive electrophysiologic assessment of brachial plexus The Journal of Bone and Joint Surgery (British)
injuries. Hand Clinics 21:39–46. 62:492–7.
8. Campbell, W. W., S. K. Sahni, R. M. Pridgeon, G. Riaz, 18. Sugioka, H., N. Tsuyana, T. Hara, et al. 1982. Inves-
and R. T. Leshner. 1988. Intraoperative electroneurog- tigation of brachial plexus injuries by intra-operative
raphy: Management of ulnar neuropathy at the elbow. cortical somatosensory evoked potentials. Archives of
Muscle & Nerve 11:75–81. Orthopaedic and Trauma Surgery 99:143.
9. Harper, C. M. 2004. Intraoperative cranial nerve mon- 19. Mahla, M. E., D. M. Long, J. McKennett, C. Green,
itoring. Muscle & Nerve 29:339–51. and R. W. McPherson. 1984. Detection of brachial
10. Kline, D. G., D. Kim, R. Midha, et al. 1999. Manage- plexus dysfunction by somatosensory evoked poten-
ment and results of sciatic nerve injuries: A 24-year tial monitoring–a report of two cases. Anesthesiology
experience. Journal of Neurosurgery 90:806–7. 60:248–52.
11. Kim, D. H., and D. G. Kline. 1996. Management 20. Mazzoni, M., C. Solinas, E. Sisillo, et al. 1996. Intra-
and results of peroneal nerve lesions. Neurosurgery operative phrenic nerve monitoring in cardiac surgery.
39(2):312–9. Chest 109:1455–60.
12. Gruen, J. P., W. Mitchell, and D. G. Kline. 1998. 21. Seal, D., J. Balaton, S. G. Coupland, et al. 1998.
Resection and graft repair for localized hypertrophic Somatosensory evoked potential monitoring during
neuropathy. Neurosurgery 43:78–83. cardiac surgery: An examination of brachial plexus
13. Kline, D. G., E. R. Hackett, and L. H. Happel. 1986. dysfunction. Journal of Cardiothoracic and Vascular
Surgery for lesions of the brachial plexus. Archives of Anesthesia 12:129–30.
Neurology 43:170–81. 22. Helfet, D. L., N. Anand, A. L. Malkani, et al. 1997.
14. Spinner, R. J., and D. G. Kline. 2000. Surgery for Intraoperative monitoring of motor pathways during
peripheral nerve and brachial plexus injuries or other operative fixation of acute acetabular fractures. Jour-
nerve lesions. Muscle & Nerve 23:680–95. nal of Orthopaedic Trauma 11:2–6.
SECTION 3
APPLICATIONS OF CLINICAL
NEUROPHYSIOLOGY:
ASSESSING SYMPTOM
COMPLEXES AND DISEASE
ENTITIES
The many electrophysiologic assessment and examination must also be considered in select-
techniques of clinical neurophysiology descri- ing individual components of the techniques,
bed in earlier chapters outline the wide range for example, which nerves to test with nerve
of measurements that can be made in patients conduction studies in a patient with suspected
with suspected disease of the central or periph- peripheral nerve disease. Neurologic symp-
eral nervous system. Each of the techniques toms will suggest whether one or more neu-
are applied either to assist clinicians in assess- ral systems are involved. Confirmatory signs
ing disease of the central or peripheral ner- on examination provide more certainty about
vous system or, less commonly, in monitoring the involvement of a particular neural system.
changes in neural function. These techniques Clinical neurophysiologic testing can often
can be used to monitor neural function in provide further confirmation if needed. The
observing progression of disease or improve- symptoms that suggest disorders of specific sys-
ment in a patient’s condition with specific treat- tems include paralysis, weakness, tremor, other
ment. They are also used in the intensive care extraneous movements, and posture abnormal-
unit and operating room to identify progressive ities (motor system); sensory loss, paresthesia,
neural damage. Each approach has advantages pain, and impairment of vision, hearing, or bal-
and shortcomings. The clinical neurophysio- ance (sensory system); or perspiration abnor-
logic testing technique that is most appropriate malities, fatigue, vascular changes, pain, and
for a patient depends on the clinical problem. emotional disorders (internal control system
Often, some combination of techniques best and autonomic disorders).
provides the necessary data. The focus of the Symptoms or signs involving movement are
following chapters is the application of clin- strong evidence that disease affects the motor
ical neurophysiologic techniques in assessing system at some level of the nervous system. If
clinical problems. there is atrophy, loss of power, jerking, shak-
The selection of a technique first requires ing, stiffness, or any of the many manifestations
taking a medical history and examining the of disease of the motor system, clinical neu-
patient and then formulating a differen- rophysiologic testing with one of the follow-
tial diagnosis. This differential diagnosis may ing modalities that assesses the motor system
include many possible disorders, but for prac- should be considered. Motor nerve conduction
tical purposes, disorders that are most likely studies can be critically important in identify-
are considered first in selecting a diagnostic ing disease of the peripheral motor pathways
technique. Findings of the clinical history and in the plexus, peripheral nerve, neuromuscular
787
junction, or muscle. Amplitudes of evoked decrease in size in cortical disease and be

responses can help define the severity of the reduced or delayed in subcortical processes
disease process and the presence of conduc- such as multiple sclerosis. Visual evoked poten-
tion block. Conduction slowing can define the tials are able to distinguish disease in the optic
precise location of the damage. Repetitive stim- nerves and tracts from that in the cerebral cor-
ulation can separate functional fatigue from tex. Multichannel movement recordings can
defects of neuromuscular transmission caused assist in determining the origin of disorders
by disease at the neuromuscular junction. Nee- with tremor, jerking, and twitching and may
dle EMG can identify and characterize periph- be identified as cortical or basal ganglionic
eral nerve and muscle diseases. Multichannel by the character of the pattern of firing and
surface electromyographic (EMG) recordings distribution on these recordings.
from widespread muscle groups can often char- Localization at the posterior fossa level can
acterize the presence or absence of a motor be assisted with several techniques. AEPs
system disorder in patients with symptoms of and auditory testing can specifically identify
disorders of movement such as tremors, jerks, involvement of the peripheral auditory path-
and twitches. These symptoms cannot always ways and distinguish it from involvement of
be classified clearly by clinical observation, auditory pathways at the level of the pons or
but quantitative measurements can sometimes midbrain. AEPs also can distinguish the nature
define them more precisely. of these disorders. Posturography can identify
Complaints of numbness, tingling, localized the presence of a peripheral vestibular disorder
pain, loss of vision, hearing impairment, and and distinguish it from central disease of the
poor balance suggest involvement of sensory vestibular pathways. Tibial and median SEPs
systems. Clinical neurophysiologic testing can recorded from the scalp with recordings at the
assist in identifying and characterizing the dis- neck can identify involvement at the poste-
ease involving the sensory system at any level rior fossa level. Blink reflexes and facial nerve
of the nervous system. The following tests conduction studies can identify involvement of
should be considered for patients with sen- cranial nerves V and VII and distinguish their
sory symptoms. Somatosensory evoked poten- involvement from diseases of the midbrain or
tials (SEPs) can help determine the presence pons. Needle EMG of cranial muscles can pro-
of impairment along the sensory pathways, par- vide evidence of damage to brain stem motor
ticularly if there is an identifiable sensory loss. neurons or peripheral motor pathways in pri-
Sensory nerve conduction studies can localize mary neurogenic processes, myasthenia gravis,
disease in peripheral nerves. Slowing of sen- or primary myopathies. Patients with bulbar
sory nerve action potentials may distinguish symptoms can often be separated into those
between primarily axonal and demyelinating with an upper motor neuron–pseudobulbar
disorders. Visual evoked potentials and audi- disorder and those with lower motor neuron
tory evoked potentials (AEPs) can help localize involvement in amyotrophic lateral sclerosis or
the site of abnormality and characterize the myasthenia gravis.
type of abnormality, even without visual or The spinal cord level can be assessed with
hearing impairment. SEPs, nerve conduction studies and needle
Electrophysiologic techniques can help in EMG, and autonomic testing. SEPs recorded
the localization of a disease to a specific level with median, ulnar, or tibial nerve stimulation
of the nervous system. A variety of techniques can distinguish involvement of the peripheral
can localize disease to and within the supra- and spinal nerves from direct spinal cord dam-
tentorial level. Electroencephalography is one age and separate lumbar and thoracic cord
of the most helpful techniques for distinguish- lesions from cervical cord lesions. Occasionally,
ing among supratentorial diseases because of SEPs are able to determine the nature of the
its ability to identify localized areas of cerebral disorder, especially demyelinating disorders.
involvement caused by either an epileptogenic On nerve conduction studies, F waves, patterns
or a destructive process. Electroencephalogra- of amplitude changes with motor and sensory
phy can also help distinguish between subcor- conduction studies, and H reflexes can some-
tical and cortical diseases. Evoked potentials times identify involvement of specific levels of
can be used to distinguish subcortical from the spinal cord. Paraspinal fibrillation poten-
cortical involvement. SEPs may increase or tials on needle EMG provide evidence of lower
Assessing Symptom Complexes and Disease Entities 789
motor neuron involvement at the spinal cord and recording. EMG can also assist in dis-
level; they can also help define distribution of tinguishing primary neurogenic disease from
lower motor neuron loss along the spinal cord. neuromuscular junction and muscle diseases.
For example, in amyotrophic lateral sclerosis, Autonomic function tests separate periph-
EMG may demonstrate evidence of subclinical eral sympathetic and parasympathetic disor-
involvement at the thoracic level. Autonomic ders from spinal cord disease and involvement
function testing in primary spinal cord dis- of central autonomic pathways. Patterns of dis-
ease, particularly localized disease with trauma, tribution of temperature change, alteration in
inflammatory disease, or ischemia, will show sweating, and vascular reflexes are combined to
localized changes at a specific segmental level. provide this information.
Localization at the peripheral level is assisted The approach to particular groups of clini-
by nerve conduction studies, needle EMG, and cal problems, with suggestions on the approach
sometimes autonomic testing. Motor and sen- to patients, is reviewed in Chapters 46 and
sory nerve conduction studies identify localized 47, focusing on the use of electroencephalogra-
areas of damage to individual nerves. Repet- phy, electromyography, and nerve conduction
itive stimulation identifies and characterizes studies. These suggestions will not be entirely
disorders of the neuromuscular junction. Nee- correct for any individual patient because
dle EMG localizes lesions in cases in which assessment depends on the unique history and
nerve conduction studies are not successful, physical examination findings of each patient,
either because the damage is primarily axonal but provide a guideline for consideration of
or the nerves are not accessible for stimulation specific studies.
Chapter 46
Assessing Central Nervous System

Symptoms
Elson L. So
INTRODUCTION Electroencephalographic
ASSESSMENT OF MOTOR Evaluation of Impaired
SYMPTOMS OF CENTRAL ORIGIN Consciousness or Delirium
ASSESSMENT OF SENSORY Electroencephalographic
SYMPTOMS OF CENTRAL Evaluation of Cognitive Dysfunction
ORIGIN Electroencephalographic
ASSESSING IMPAIRMENT OF Evaluation of Seizures and Other
CONSCIOUSNESS AND Paroxysmal Disorders
COGNITION EEG in the Intensive Care Unit
ASSESSING IMPAIRMENT OF Intraoperative EEG
VISCERAL FUNCTION AND SLEEP EEG in the Newborn
IDENTIFYING DISEASE TYPES EEG in the Epilepsy Monitoring
PROGNOSIS Unit
ASSESSING CLINICAL
SUMMARY
DISORDERS WITH EEG
than imaging tests in monitoring the ongoing

INTRODUCTION effects of surgical procedures.
This chapter will provide an overview of
Clinical history and neurological examina- the application of clinical neurophysiology in
tions form the foundation of assessing cen- central nervous system disorders.
tral nervous system (CNS) disorders. In many
instances, laboratory tests are needed to affirm
diagnosis and guide treatment. Imaging tests
alone are often inadequate for these pur- ASSESSMENT OF MOTOR
poses, and neurophysiologic tests are needed to SYMPTOMS OF CENTRAL
understand the mechanisms underlying CNS
symptoms and to establish the neurological
ORIGIN
diagnosis. Moreover, neurophysiological pro-
Symptoms or signs involving movement are
cedures are more widely available and useful
strong evidence that disease affects the motor
791
system at some level of the nervous sys- ASSESSING IMPAIRMENT OF

tem. Electroencephalography (EEG) should CONSCIOUSNESS AND
be considered if the motor symptom or deficit COGNITION
may be caused by disease at the cortical level,
either a seizure or a destructive process (e.g., Episodic or continuous confusion, uncon-
tumor or infarct). EEG is of limited value in sciousness, and disorders of sleep can be eval-
assessing movement disorders, except when an uated and characterized with the following
epileptic mechanism is under consideration. electrophysiologic tests:
Transcranial motor evoked potentials (MEPs)
test the entire motor pathway from the cere- • EEG, often critical in defining the
bral cortex (where the motor system is acti- nature of an impairment of consciousness,
vated by a magnetic or electric pulse) to muscle can distinguish seizure disorders from
(where the response is recorded). MEPs are metabolic disturbances, focal lesions with
most useful in distinguishing primary motor increased pressure, hysterical disorders,
system disease from functional or psychiatric and sleep disorders.
disease. For example, a patient with hysterical • Polysomnography provides a precise assess-
paralysis of the legs can be evaluated with this ment and characterization of sleep dis-
technique. orders, often assisting in defining their
specific nature and, thereby, appropriate
treatment modalities.
• SEPs and brain stem auditory evoked
potentials (BAEPs) can determine whether
ASSESSMENT OF SENSORY central sensory pathways from the level
of the brain stem to the cerebral cortex
SYMPTOMS OF CENTRAL are intact and functioning normally. This
ORIGIN helps define the status of patients in coma.
Severity of impairment of SEPs after head
Complaints of numbness, tingling, localized trauma is helpful in judging prognosis.
pain, loss of vision, hearing impairment, and
poor balance suggest involvement of sensory EEG is the most precise laboratory tech-
systems. Clinical neurophysiologic testing can nique available for assessing cortical function.
assist in identifying and characterizing the dis- Disorders of cognition with impairment of
ease involving the sensory system at any level of mentation, language, and memory are caused
the nervous system. The following tests should primarily by disorders at the level of the cere-
be considered for patients with sensory symp- bral cortex. Only clinical testing of thinking
toms: and memory can provide more information
than EEG. Characteristic EEG abnormali-
ties can help define the nature of the disor-
• EEG can help identify sensory symptoms ders that produce impaired mentation, includ-
arising at the cortical level. For example, ing Alzheimer’s disease, chronic infection,
unilateral paresthesia caused by a seizure ischemic vascular disease, subdural hematoma,
discharge or a local destructive lesion may and frontal lobe mass lesions.
be associated with focal spikes or slowing.
• Somatosensory evoked potentials (SEPs)
can help determine the presence of ASSESSING IMPAIRMENT OF
impairment along the sensory pathways, VISCERAL FUNCTION AND
particularly if there is an identifiable sen- SLEEP
sory loss.
• Visual evoked potentials and auditory Impairment of visceral function, including
evoked potentials (AEPs) can help local- autonomic disorders, syncope, and disorders of
ize the site of abnormality and character- the reproductive organs, can be assessed with
ize the type of abnormality, even without two groups of electrophysiologic tests: auto-
visual or hearing impairment. nomic function testing and polysomnography.
Assessing Central Nervous System Symptoms 793
• Autonomic function testing can assess PROGNOSIS

peripheral sympathetic function in vascu-
lar disease or peripheral nerve disease. Although several clinical neurophysiologic pro-
It can also assess central vascular con- cedures can define the severity of a disease
trol mechanisms that may be altered in process, the procedures vary in their abil-
autonomic diseases such as multisystem ity to characterize the stage of evolution of
atrophy. the disease or to provide prognostic infor-
• Polysomnography assesses sleep disor- mation. EEG is the most established and
ders by measuring both EEG activity widely used laboratory procedure that pro-
and associated autonomic function. Dis- vides nonstructural—but objective—evidence
orders of the autonomic nervous system of the severity and progression of cerebral
may be manifested on polysomnography. disorders. However, except for a few disor-
Polysomnography is particularly valuable ders, the association between an abnormal
in central disorders of the internal regu- EEG finding and disease severity varies among
latory system, but it can also be helpful in patients, partly because of the variability in the
assessing impotence and sleep apnea. manner and the severity with which the cere-
bral symptoms present for similar stages of
the disorder. A key principle in using EEG is
that serial EEG recordings must be obtained
IDENTIFYING DISEASE TYPES to optimize the value of this test in assess-
ing disease progression and prognosis in each
Clinical neurophysiology can facilitate identifi- person.
cation of specific diseases. Electrophysiologic The physiologic and pathologic mechanisms
testing can sometimes supplement the initial that underlie normal and abnormal EEG are
classification of a disease as vascular, inflamma- not as well understood as those in peripheral
tory, degenerative, or neoplastic, but it does so nerve and muscle disorders. One reason is that
in different ways for different tests. In many brain specimens of living patients are not as
instances, only a broad category of disease can readily available for study and correlation with
be suggested, but in others, specific diseases EEG findings. Nonetheless, EEG is used to
can be identified. These are described in detail evaluate nearly all diseases affecting the brain.
in the chapters on each of the techniques. The It remains the best tool for assessing severity
following are some examples. and prognosticating the outcome of many cate-
gories of cerebral dysfunction, including toxic,
• EEG can distinguish epileptogenic from metabolic, infectious, vascular, degenerative,
destructive lesions, but rarely can it cate- traumatic, and seizure disorders.
gorize the changes as neoplastic, inflam-
matory, infectious, or degenerative. At
times, specific clinical entities can be
suggested, such as the Lennox–Gastaut
syndrome, hepatic encephalopathy, and ASSESSING CLINICAL
hypsarrhythmia of infancy. DISORDERS WITH EEG
• Evoked potentials rarely provide evi-
dence of a specific clinical entity. Marked As indicated by the discussion above, clini-
increase in latency is usually evidence of cal neurophysiologic testing cannot be applied
a demyelinating process in patients with in a routine fashion. For each patient, test-
multiple sclerosis. Evoked potentials are ing should be designed to answer the clinical
useful particularly in identifying areas of question posed by the patient’s problem. This
subclinical involvement to confirm the requires giving careful thought to the selection
presence of multiple lesions in multiple of the testing procedures. The ensuing sections
sclerosis. discuss the practical approach in using EEG to
• Polysomnography testing can identify assess central nervous system disorders.
sleep apnea and some forms of sleep Unlike other clinical neurophysiologic tests,
impairment such as periodic movements EEG is used frequently and regularly for
of sleep. patients of all age groups and with various
conditions. Therefore, the following informa- • Compressing the timescale (reducing the
tion about the patient and recording condi- draw speed) emphasizes the display of
tions must be obtained and documented with alpha and beta activities and may mask
each EEG procedure: (1) age (including ges- underlying delta and theta waveforms.
tational age in neonates), (2) clinical history, Reducing the setting of the low-pass filter
(3) reason for the procedure, (4) medications, (i.e., from high linear frequency filter set-
(5) time of last meal, (6) number of hours of ting of 70 Hz to 15 Hz) could alleviate the
previous night’s sleep, (7) time of last occur- masking of slower waveforms and reveal
rence of symptom, (8) current level of con- generalized or focal slowing.
sciousness and alertness, and (9) previous EEG • Review waveforms of interest in different
records. montages, at least in one bipolar montage
The following are the general technical con- and one referential or Laplacian montage.
siderations when recording or reviewing an • Review waveforms of interest with a mon-
EEG: tage that covers sufficiently all regions
of both hemispheres, before focusing
on montages that cover selected brain
• Recording environment must be comfort- regions.
able for the patient and without any dis- • Display the electrocardiographic monitor
traction if possible. at proper amplitude so that complexes will
◦ Document the patient’s behavior, espe- be easily visualized yet not interfere with
cially when there is an EEG discharge. EEG tracings.
• Document potential or known source of • Increasing the time base (increasing draw
artifacts. speed) unmasks high-frequency wave-
◦ Monitor respiratory movements and forms that are hidden by digital under-
oximeter readings if necessary. sampling of the EEG data.
◦ Stimulate the patient to elicit the most • Reducing the low-pass filter (high lin-
alert state and its corresponding EEG ear frequency filter) corrects the digital
appearance. aliasing effect of distorting high-frequency
• Record effect of eye opening and eye waves into slower waveforms.
closure on the EEG.
• Use additional electrodes when appropri- A segment of the recording should show EEG
ate to enhance detection and localization activity when the patient is most alert. This
of abnormal activities. is important because many patients become
◦ Adjust sensitivity or gain settings and drowsy or sleepy during the EEG procedure.
frequency filters to display properly The EEG activities associated with drowsiness
waveforms of different amplitude range. are difficult to differentiate from abnormal
(With analog recording, pages of differ- background slowing of a mild degree. Drowsy
ent settings may be necessary.) and sleep EEG activities can also mask back-
◦ Avoid setting low linear frequency filter ground abnormality of generalized or focal
at greater than 1.0 Hz. (Analog record- slowing. The patient should be stimulated ver-
ing should include pages with recording bally or physically during the recording to
at 0.5 Hz.) determine the highest level of arousal that the
◦ Minimize distortion of both benign patient is capable of achieving, especially when
sharp transients and abnormal sharp the EEG is performed to evaluate impaired
waves by minimizing use of low-pass fil- consciousness.
ter (high linear frequency filter) below The best approach to discussing the use
70 Hz. of EEG in evaluating neurologic disorders is
• Expanding the timescale (increasing draw a symptom-oriented approach. The discussion
speed) enhances the appearance of fast should also include the recording of EEGs
frequency waveforms. in special environments such as the operat-
• Expanding the timescale (increasing draw ing room and intensive care units. Preceding
speed) facilitates the appreciation of phase chapters have already presented the abnormal
relationship between waveforms in differ- EEG features or patterns of specific cerebral
ent channels. disorders.
Electroencephalographic Beware that stimulation may not be advis-

Evaluation of Impaired able in patients with increased intracra-
Consciousness or Delirium nial pressure or unstable cardiovascular
status.
• When there is spontaneous or unexpected
Impairment of consciousness is one of the most
frequent clinical manifestations of acute neu- change in the baseline EEG, look for
rologic illnesses. Electrophysiologic recording potential stimulus of the change in the
is the only laboratory method that objec- recording environment, such as suctioning
tively assesses the severity of disturbance in or repositioning the patient.
◦ Observe the patient continually for
cerebral function. This property of electro-
physiologic recording makes it valuable as a abnormal movements.
◦ Monitor respiratory movements and
method for objectively following the progress
of the patient’s condition to supplement clini- oximeter readings if necessary.
• Record effect of eye opening and eye
cal observation. In general, the severity of EEG
abnormality parallels the observed depth of closure on the EEG.
◦ When paroxysmal or periodic EEG
mental obtundation.
Several abnormal EEG patterns indicate the patterns occur or the background
severity and the prognosis of the patient’s con- of the EEG changes spontaneously,
dition (e.g., burst-suppression pattern, spindle observe and document any correspond-
coma pattern, alpha coma pattern). Moreover, ing change in motor or behavioral
a number of abnormal EEG patterns sug- activity.
◦ Use extraocular eye leads to help
gest strongly the probable cause or mecha-
nism underlying the mental obtundation (e.g., resolve the nature of frontal waveforms
triphasic waves, seizure discharges, periodic that are difficult to distinguish from eye
lateralized epileptiform discharges [PLEDs]). movement artifacts.
• Reposition the head if ambiguous wave-
Not uncommonly, an EEG shows objective
signs of cerebral dysfunction before other test forms occur in the cranial region that rests
results are positive or available (e.g., slow- against the examination table.
• Consider recording sleep activity to acti-
ing or PLEDs in infectious, ischemic, trau-
matic, metabolic, or toxic disorders). The vate spikes and sharp waves in patients
following should be considered when perform- with mild to moderate mental obtun-
ing or reviewing EEGs for evaluation of mental dation, unless the recording already
obtundation: shows abnormalities highly suggestive of
the underlying cause or mechanism of
delirium.
• When prolonged recording is not con- • Consider serial EEGs to help monitor the
templated, use of electrode cap may be course of the patient’s disorder.
appropriate in a patient who is not restless. • In making the diagnosis of electrocerebral
• Use compressed timescale (slow draw inactivity, adhere to the recommendations
speed) and increase the time constant if of the American Clinical Neurophysiology
subtle focal slowing or asymmetries are Society.
suspected.
• Compressing the timescale emphasizes
the display of alpha and beta activities
and may mask underlying delta and theta
waveforms. Reducing the setting of the Electroencephalographic
low-pass filter (i.e., from high linear fre- Evaluation of Cognitive
quency filter setting of 70 Hz to 15 Hz) Dysfunction
alleviates this masking of slower wave-
forms so that generalized or focal slow- Cognitive dysfunction is another frequent neu-
ing becomes apparent. Verbally stimu- rologic complaint and presenting symptom.
late the patient, and observe and docu- Whereas neuropsychologic testing directly
ment any behavioral or EEG response. measures the symptom of cognitive deficit,
Use physical stimulation if necessary. EEG indirectly evaluates the severity of the
cognitive dysfunction. However, the infor- Electroencephalographic

mation provided by EEG is independent Evaluation of Seizures and Other
of the patient’s cognitive effort and perfor- Paroxysmal Disorders
mance at the time of the recording. Thus,
EEG findings can complement the clinical Epileptic seizure episodes result from the
assessment of cognitive disorders by detect- abnormal and excessive discharge of neurons.
ing objective evidence of cerebral dysfunc- Thus, EEG is commonly used to detect
tion. Furthermore, memory impairment or for- abnormal interictal and ictal electric discharges
getfulness is frequently a symptom of anx- that are highly associated with epileptic seizure
iety and depression. The public’s awareness disorders. Interictal abnormalities that have
of Alzheimer’s disease has resulted in more the best association with epileptic seizure dis-
patients becoming concerned about the sig- orders are spikes, sharp wave discharges, and
nificance of their own symptoms of forget- temporal intermittent rhythmic delta activity
fulness. Abnormal EEG findings help deter- (TIRDA). These interictal epileptiform dis-
mine the organic nature of the complaint. charges (IEDs) occur in only 40% of the ini-
However, normal EEG findings in combina- tial EEGs of patients with seizures and in
tion with normal clinical and laboratory results approximately 1% of persons without epilepsy.
can be reassuring. Although normal EEG find- Nonetheless, their presence strongly supports
ings do not completely exclude an organic a diagnosis of seizure disorder when other
cause of cognitive disorder, an abnormal wake appropriate clinical or laboratory data are
EEG background eliminates psychologic dis- present. Furthermore, focal or generalized
order as the only explanation of cognitive EEG slowing may reveal an underlying struc-
dysfunction. tural or functional derangement associated
The correlation between the severities of with the seizure disorder.
EEG abnormality and cognitive dysfunc- EEG helps in determining the specific
tion is not as good as that between EEG seizure type or epilepsy syndrome. The clas-
and mental obtundation. This is particu- sifications of epileptic seizures and epilepsy
larly true for subcortical dementias. Despite syndromes are based on both the clinical infor-
this, a few cognitive disorders have char- mation and the type of EEG abnormality. The
acteristic EEG patterns. Conditions charac- location and distribution of IEDs or ictal dis-
terized by rapid cognitive decline in adults charges help determine the seizure type and
often require consideration of Creutzfeldt– the epilepsy syndrome. At the minimum, the
Jakob disease. Because brain biopsy is often distinction between focal and generalized dis-
avoided in these patients, the characteris- charges contributes importantly to the initial
tic pattern of periodic sharp waves may step of seizure diagnosis, that is, establishing
be the only supportive laboratory evidence whether the disorder is focal or primary gen-
available. eralized. This is an essential step in seizure
Some specific causes of childhood dementia management, because selecting the antiepilep-
or progressive encephalopathy can be tic drug appropriate for treatment depends
suggested by characteristic abnormal EEG largely on distinguishing between these two
patterns. Examples are the patterns of peri- main seizure types. Although some antiepilep-
odic waveforms in patients with metabolic tic drugs are effective for both seizure types,
encephalopathy, subacute sclerosing panen- many are effective in controlling one type and
cephalitis (SSPE), or abnormal storage dis- not the other. Knowledge of the location and
eases. Epileptic encephalopathy can also be distribution of IEDs and ictal discharges is also
suggested by the detection of frequent clinical essential in localizing the surgical focus and
or subclinical seizures or widespread epilepti- selecting candidates for epilepsy surgery.
form discharges. Because formal neuropsycho- Certain interictal EEG discharge patterns
logic testing may be difficult to perform during are characteristic of specific epilepsy syn-
early childhood, EEG may be especially useful dromes. In the appropriate clinical setting, the
as an objective tool in the serial assessment of hypsarrhythmia pattern is specific for infantile
cognitive function. spasms, whereas centrotemporal discharges
induced by sleep are highly suggestive of • Use precipitating measures for patients
the syndrome of benign rolandic epilepsy whose spells have known precipitants.
(benign epilepsy with centrotemporal spikes). • Consider supplementing the “routine”
The diagnosis of the epilepsy syndrome deter- EEG recording with simultaneous video
mines the clinical management and progno- recording if the patient is experiencing
sis of many seizure disorders. Many epilepsy daily spells.
syndromes are age-dependent in onset and • Minimize distortion of sharp transients
remission, and the likelihood of spontaneous and paroxysmal activities by minimizing
remission is good. In comparison, some syn- the use of low-pass filter (high linear fre-
dromes typically are intractable to drug treat- quency filter) of less than 70 Hz.
ment. Also, certain epilepsy syndromes are • Use of the timescale or draw speed of
highly associated with an underlying structural 30 mm/second is generally suitable for
abnormality. detecting sharp transients and paroxysmal
The presence or absence of IEDs can serve activities. Lower speed may interfere with
as a prognostic factor in assessing the risk of the visual detection of low or medium
seizure recurrence. Following the first unpro- amplitude sharp transients and paroxys-
voked seizure, the presence of IEDs is associ- mal activities.
ated with a higher risk of seizure recurrence. • Expanding the timescale (increasing draw
Many studies also support the finding that if speed) helps in assessing phase relation-
the current EEG shows IEDs in seizure-free ship between sharp waves in different
patients who discontinue taking antiepilep- channels. Use additional electrodes when
tic medication, seizures are more likely to appropriate to enhance detection and
recur. localization of sharp transients and parox-
The following should be considered when ysmal activities.
evaluating patients for epileptic seizure disor- • Use one channel for the electrocardio-
ders and other paroxysmal events: graphic monitor and another channel for
oximeter monitoring.
• When the capture of paroxysmal events
• Obtain a sleep recording as well as a wake is still indicated despite having performed
recording, unless the wake recording has the EEG with the above measures, con-
already disclosed IED activity that is suffi- sider ambulatory EEG recording or pro-
cient for clinical management. longed monitoring in the epilepsy moni-
• Consider partial sleep deprivation before toring unit.
the EEG procedure, especially if a
previous EEG did not show epileptiform
abnormalities. EEG in the Intensive Care Unit
• Schedule sleep-deprived patients for EEG
to be performed the following morning Recording EEGs in the intensive care unit
and not the afternoon. presents special challenges. Several devices
• If the patient still is unable to fall and pieces of equipment in the intensive care
asleep during the procedure despite sleep unit can introduce artifacts into the EEG
deprivation, consider administering chlo- recording and make EEG recording difficult,
ral hydrate to promote sleep. (Precau- such as electrocardiographic and blood pres-
tions of conscious sedation should be exer- sure monitors, indwelling catheters, respira-
cised, particularly for children. Instruct tors, intravenous pumps, surgical drains, and
the patient not to drive for the rest of the positive leg pressure devices. Placing elec-
day if a sedative is given.) trodes on the patient and recording the EEG
• Use anterior temporal electrodes to may interfere with nursing care and vice versa.
enhance the probability of recording tem- Generally, the patients in intensive care who
poral IEDs. have an EEG study have altered mentation
• Perform photic stimulation and hyperven- or are experiencing seizures and other parox-
tilation unless contraindicated medically. ysmal events. Thus, recommendations made
above for EEG recording for specific clinical recording must be interpreted immediately to
situations should be followed when applicable provide the information necessary to guide
(i.e., recording the EEG of a patient in coma the surgical procedure. Because intraoperative
or with seizures). Additional recommendations recordings are essentially prolonged monitor-
for making EEG recordings in the intensive ing that extends over hours, digital EEG should
care unit are the following: be used. Digital recording allows prompt
retrieval of segments of the recording for side-
• Ensure electrical safety (see Chapter 2). by-side comparison to assess the course of the
Avoid introducing the patient into the patient and the effect of surgical intervention.
path of a ground loop or double ground,
especially a patient with an indwelling car- EEG in the Newborn
diovascular catheter.
• Apply electrodes cautiously, and adjust
Frequently, EEG is performed in a full-
the application accordingly when there term or premature newborn for evaluation of
are traumatic or surgical wounds at the suspected abnormal movements and apneic
craniocervial regions, or in patients with episodes. Clinical manifestations of seizures in
cervical spine injuries. the newborn differ from those in older children
• Observe infection control measures.
and adults. Many of the seizure behaviors in
• Observe closely for artifacts. Determine
the newborn are subtle, and many also mimic
and document their origin. normal physiologic events. In the newborn,
• Document the time and the amount of
apnea is much more frequently an epileptic
administration of medications that have manifestation than it is in older patients. How-
CNS effects. ever, apnea in the newborn is also commonly
• If artifacts from other equipment inter-
a manifestation of cerebral injury or severe
fere excessively with the EEG recording, prematurity. For these reasons, EEG is fre-
inquire whether the equipment responsi- quently used to detect objective abnormali-
ble for the artifacts can be turned off or ties that help in determining the mechanism
removed temporarily. or the nature of the clinical manifestations
• Modify and document electrode place-
in the newborn. Recording EEG in the new-
ments if head dressings or wounds inter- born presents unique challenges and requires
fere with standard electrode placements. special skills. Considerable skill is needed in
• Consider prolonged recording or inter-
applying electrodes on a small head, especially
mittent recording to monitor the clinical in premature neonates. The scalp of the new-
course of the patient. born is more delicate than that of an older
• For prolonged monitoring, conduct daily
child or adult. Extracerebral monitors such
safety and integrity checks of the record- as those for eye movements, respiration, and
ing system. muscle activity are needed to help define the
• Consider simultaneous video-EEG record-
wake and sleep states in the newborn. Many
ing for evaluating the nature of behav- EEGs of premature newborns are performed
ioral or EEG events, and for recognizing in the neonatal intensive care unit. Thus, the
sources of EEG artifacts. requirements and constraints discussed in the
preceding section about recording in the inten-
sive care unit setting also apply. In addition, the
Intraoperative EEG following should be considered:
Electroencephalographic monitoring of the • Assure the parents or caregiver about the
cerebral cortex is performed most often during nature of the study.
carotid endarterectomy and epilepsy surgery. • If possible, perform the recording during
Recording in the operating room setting or right after feeding.
presents challenges similar to those of record- • Use miniature cup electrodes for record-
ing in the intensive care unit. Intraopera- ing the EEG, surface EMG, eye
tive EEG has additional constraints, such as movements, and electrocardiogram; use
anesthetic agents and limited ability to physi- piezoelectric transducers or impedance
cally adjust the patient or the equipment. The pneumographs for recording respiration.
• In the newborn, oximeter recordings can occupational therapists, psychologists or psy-

identify seizures that result in oxygen chiatrists, and social work personnel. The need
desaturation. for their services should be individualized
• Use a nasal thermistor for recording according to the medical, psychologic, and
airflow and simultaneous respiratory piezo- social conditions of each patient. The follow-
electric transducers or impedance pneu- ing should also be considered when conducting
mographs if apnea is suspected. epilepsy monitoring:
• Make certain that the setting is well ven-
tilated if using collodion or acetone. Do
not use either of these substances inside • Counsel patients and guardians about the
an isolette. nature and the requirements of the proce-
• Use a heat lamp, and monitor body tem- dure, including the need to push the event
perature if the newborn is removed from button when symptoms occur.
the isolette for applying electrodes. • Ensure a safe monitoring environment.
• Note if there is scalp edema, which may Minimize or strategically locate equip-
affect the EEG recording. ment and furniture. Protect the patient
• Use the newborn montage, with fewer from hard surfaces or protruding fixtures.
electrodes. Use padding if necessary.
• Use the appropriate frequency filter set- • The patient should be accompanied when
tings and gain to optimize recordings. ambulating if spontaneous or activated
• Perform part of the recording with a mon- seizures are likely to occur. Have selected
tage that includes the midline or central patients wear helmets.
regions. • Make sure the electrode connections are
• Adjust or prop the patient’s head to mini- stable and the patient is in view through-
mize electrocardiographic and ballistocar- out the monitoring. Use a high-resolution
diographic or other movement artifacts. color camera during the daytime and an
infrared camera at night to obtain the best
video image.
• Facilitate the occurrence of seizures or
EEG in the Epilepsy Monitoring symptoms by activating procedures if nec-
Unit essary (e.g., withdrawal of antiepileptic
drug therapy, sleep deprivation, photic
Video-EEG monitoring has become a major stimulation, hyperventilation, or psycho-
procedure in clinical neurophysiology (see logic suggestion).
Chapter 11). Currently, epilepsy monitoring is • Provide continuous visual monitoring
conducted in many hospitals and clinics. The unless the type of spell or seizure is
monitoring can be done in a dedicated facil- not likely to cause injury. Be aware that
ity with fixed equipment or in other locations patients with epilepsy are at risk for
with mobile recording equipment. The advent the development of prolonged convulsive
of digital EEG and video has made it eas- seizures after antiepileptic drug treatment
ier to store and access data for review. The has been withdrawn.
correlation between EEG activity and clini- • For each patient, provide a plan for pre-
cal behavior of the patient is enhanced by the venting or interrupting prolonged seizures
simultaneous display of video and EEG data. or spells. This may include having venous
The recording can be retrieved and reviewed access with a heparin lock. Certified
at remote locations as needed if the equipment equipment and qualified personnel must
used for recording, storing, and reviewing data be immediately available for cardiorespi-
are linked in a network. ratory resuscitation.
Successful use of video-EEG monitoring • Except for brief seizures such as absence
depends on both technical and nontechni- or myoclonic seizures, evaluate the patient
cal factors. Epilepsy monitoring is performed when each seizure occurs, noting the time
best as part of a comprehensive program of of occurrence.
patient evaluation and management. Essen- • Ictal and postictal neurologic evalua-
tial participants in the program include nurses, tions should include tests for alertness,
orientation, comprehension, language, and SUMMARY

motor function.
◦ Document the time and the amount The major value and primary application of
of administration of antiseizure medi- clinical neurophysiology is in the assessment
cations or other medications that have and characterization of neurologic disease.
CNS effects. Selection of appropriate studies for the prob-
◦ Conduct daily safety and integrity lem of an individual patient requires a careful
checks of the recording system. clinical evaluation to determine possible causes
• Encourage assisted physical activity to of the patient’s symptoms. The nature of the
minimize complications of long-term bed symptoms and the conclusions of the clinical
rest. evaluation are the best guides to appropriate
• Adjust settings of spike or seizure detec- use of clinical neurophysiologic testing.
tion programs according to patient’s The approach to testing can be assisted
sleep–wake state and physical activity to by deciding which structures are likely to be
optimize detection of events. involved. EEG, autonomic function testing,
◦ Secure the leads of intracranial elec- and polysomnography provide distinct assess-
trodes to minimize pulling or movement of disturbances of consciousness, cog-
ment. nition, visceral function, and sleep. The level
◦ Verify that the connection of intracra- of the nervous system that is likely to be
nial electrode leads to the jackbox is involved by the disease process can also guide
correct. selection of the neurophysiologic methods that
will be most helpful in sorting out the clini-
At our institution, the connections are verified cal problem. Disorders of the cerebral hemi-
by two technologists each time it is made, and sphere are best characterized electrically by
they document their verification by signing on EEG, SEPs, polysomnography, and movement
the daily technologist worksheet. recordings.
Chapter 47
Application of Clinical
Neurophysiology: Assessing
Peripheral Neuromuscular
Symptom Complexes
Devon I. Rubin and Jasper R. Daube
CLINICAL NEUROPHYSIOLOGY PERIPHERAL NEUROPATHY

IN THE ASSESSMENT OF Defining Pathophysiology of
PERIPHERAL NERVOUS Peripheral Neuropathies
SYSTEM DISORDERS (Axonal vs. Demyelinating)
Confirming a Clinical Diagnosis Axonal Neuropathies
Excluding Alternate Diagnoses Demyelinating Neuropathies
Localizing the Disease BRACHIAL PLEXOPATHY
Identifying Disease in Patients Who Trunk Lesions
Are Difficult to Examine Clinically Cord Lesions
Identifying Subclinical Disease
Characterizing Disease GENERALIZED WEAKNESS
Pathophysiology MYOPATHY
Defining Disease Severity MYALGIAS, MUSCLE STIFFNESS,
Defining Stage and Evolution of AND EPISODIC MUSCLE
Disease WEAKNESS
Determining Prognosis NMJ DISORDERS
Identifying Disease Types POLYRADICULOPATHY
MOTOR NEURON DISEASE
ASSESSING CLINICAL FACIAL WEAKNESS
DISORDERS: ASSESSMENT ANOMALOUS INNERVATION
WITH EMG AND NCS Martin-Gruber Anastomosis
RADICULOPATHIES Riche–Cannieu Anastomosis—“All
Cervical and Lumbosacral Radiculopathies Ulnar Hand”
Thoracic Radiculopathies Accessory Peroneal Nerve
COMMON FOCAL MONONEUROPATHIES UNEXPECTED FINDINGS ON
Carpal Tunnel Syndrome NERVE CONDUCTION
Ulnar Neuropathy STUDIES: CAUSE AND ACTION
Peroneal Neuropathy SUMMARY
801
CLINICAL NEUROPHYSIOLOGY Excluding Alternate Diagnoses

IN THE ASSESSMENT OF
PERIPHERAL NERVOUS SYSTEM Clinical neurophysiology can be of value when
a specific disease is suspected, but other dis-
DISORDERS eases with similar findings have to be excluded.
For example, a patient with hand numbness
When a patient comes to a physician with spe- and arm pain may have clinical features sugges-
cific complaints, the assessment of neurologic tive of carpal tunnel syndrome (CTS) although
disease begins with generating a hypothesis a C6 radiculopathy might also be a possibil-
about the location and type of disease based on ity; nerve conduction studies (NCS) and needle
the patient’s symptoms, history, and the results electromyography (EMG) can be used to help
of the neurologic examination. A hypothesis is exclude the latter.
formulated on the neural system(s) involved,
the level of involvement, the type of dis-
ease, and the prognosis. If there is sufficient
certainty that the hypothesis is correct, electro- Localizing the Disease
physiologic testing may not be needed. How-
ever, in many instances, precise anatomic local- Clinical neurophysiology may help localize the
ization may be difficult and more than one sys- disease with a precision not possible clinically.
tem may be involved. Neurophysiologic testing Localization may be general, such as iden-
can assist in confirming the suspected localiza- tifying a disorder at the NMJ in a patient
tion. In the peripheral nervous system, a dis- with generalized weakness. Additionally, local-
ease can be localized to the spinal cord, brain ization may be focal and precise along a nerve,
stem, spinal root, plexus, peripheral nerve, such as when short segmental stimulation stud-
neuromuscular junction (NMJ), or muscle, and ies (“inching”) on NCS identify compression
may involve any combination of the motor at the aponeurosis of the flexor carpi ulnaris
system, sensory system, or autonomic ner- muscle as the cause of an ulnar neuropathy.
vous system. The neurophysiologic tests used
to confirm the localization and the systems
involved are selected based on the patients’ Identifying Disease in Patients
symptoms. In addition to localization, clini- Who Are Difficult to Examine
cal neurophysiology can help define a num- Clinically
ber of important features of the underlying
problem. In situations in which the physician cannot
obtain an adequate clinical history or perform
an adequate neurologic examination, clinical
Confirming a Clinical Diagnosis neurophysiology may provide the information
needed to make a diagnosis. These situations
The most common application of clinical neu- include patients who are in the intensive care
rophysiology is to confirm a suspected clinical unit, have dementia or psychiatric disease, or
diagnosis. Uncertainty about the diagnosis usu- may not be able to cooperate. A language
ally reflects an atypical or incomplete symp- barrier may interfere with taking a medical
tom complex, incomplete or mixed findings history and performing a neurologic examina-
that do not all fit with the suspected disor- tion. When traumatic injuries such as fractures
der, a relatively mild stage of the disease with or postoperative immobilization preclude thor-
a minimum of symptoms and signs, or unex- ough neurologic examination, clinical neuro-
pected findings that are not consistent with physiology may be able to assess function and
the diagnosis. For example, a patient with a provide essential information.
recent history of mild, symmetric proximal arm
and leg weakness may be clinically suspected
of having a myopathy, such as polymyosi- Identifying Subclinical Disease
tis. The findings on electrodiagnostic testing
can help to confirm an underlying muscle Several electrophysiologic techniques can be
disease. used to identify subclinical disease by detecting
Application of Clinical Neurophysiology 803
an abnormality that either is below thresh- to occur if the underlying cause can be
old for clinical identification or has no clinical eliminated. In contrast to conduction block,
accompaniments. Examples include slowing of slowing of conduction alone may be associated
conduction in a hereditary neuropathy with no with little or no clinical deficit.
deficit, fibrillation potentials in a radiculopathy Axonal degeneration. Axonal disruption or
with no clinical deficit and widespread fibril- degeneration is associated with a loss of axons
lation potentials from ALS with deficit in only through Wallerian degeneration. Therefore,
one arm. recovery of function depends on reinnerva-
tion. Reinnervation can occur rapidly, within
days to weeks, if the number of axons lost is
Characterizing Disease not great and the remaining axons can pro-
Pathophysiology vide reinnervation by local collateral sprouting.
Reinnervation is much slower, over months to
In clinical situations in which the physician is years, if it requires sprouting and growth of the
able to localize a disorder to the peripheral ner- damaged axons.
vous system, clinical neurophysiology may be
Electrophysiologic changes in muscle, as iden-
needed to characterize the disease. In disor-
tified during needle EMG, associated with
ders of the peripheral nerve, the combination
nerve damage depend primarily on whether
of findings on NCS and needle examination can
the degeneration is Wallerian. Slowing of con-
help to classify the problem with major impli-
duction is not associated with measurable
cations for identifying the underlying etiology
changes on needle EMG or in estimates of the
and determining prognosis.
number of motor units. Conduction block is
In disorders of the peripheral nerve, sev-
associated with reduced recruitment of motor
eral mechanisms may define the underlying
unit potentials (MUPs) on needle EMG and
pathophysiology:
reduced estimates of the number of motor
units proximal to the site of damage. Wallerian
Conduction block. In a neuropraxic nerve degeneration produces reduced recruitment of
lesion due to focal demyelination or a focal MUPs, reduced estimates of the number of
region of inexcitability of the nerve, a localized motor units, and muscle changes associated
block of conduction of the action potentials with denervation and reinnervation.
occurs along some or all of the axons. The pro- Denervation of muscle results in a loss of
portion of fibers that are blocked in a nerve the trophic factors that maintain normal mem-
directly correlates with the amount of clini- brane function. With the loss of innervation, a
cal deficit. With complete block, there is no muscle fiber discharges spontaneously and the
voluntary movement of the muscle innervated associated discharges are fibrillation potentials.
by the affected nerve, whereas with partial Fibrillation potentials develop 1–3 weeks after
block, there is some degree of weakness. Nerve acute denervation. Delay in their appearance
function proximal and distal to the conduc- varies with species and muscle characteristics.
tion block can be entirely normal. Conduction In humans, the delay depends most on the
block may be caused by either metabolic or length of axon attached to the muscle fiber.
structural changes. Conduction block caused If axonal destruction occurs close to a mus-
by local anesthetics, anoxia, and some tox- cle fiber, fibrillation potential develops more
ins may improve over minutes to hours and quickly than if the damage is more proximal
may not be associated with histologic alter- (i.e., the shorter the segment of axon attached
ation. Conduction block caused by distortion to the muscle, the more quickly Wallerian
or loss of myelin persists for days to weeks, degeneration occurs). The corollary is that
because it requires remodeling of the histo- muscles closer to the lesion show fibrillation
logic abnormality. Conduction block caused by potentials sooner than muscles more distant to
either of these mechanisms may not improve if the damage.
the offending mechanism is not eliminated. The reinnervation of muscle is associated
Partial demyelination. Slowing of conduction with a defined sequence of changes in the
is usually caused by myelin changes and, thus, estimate of the number of motor units and
requires weeks to months for improvement in MUPs. If reinnervation is by collateral
sprouting, the estimate of the number of motor Defining Disease Severity

units remains low and increases only in propor-
tion to the number of axons that regenerate and Clinical assessment usually can define the
reach the muscle. With initial reinnervation, severity of a disease as mild, moderate, severe,
MUPs consist of activity in only a small number or with total loss of function. Clinical neu-
of muscle fibers, with poor synchrony of firing rophysiology can quantitate the severity with
and unstable NMJs. Therefore, potentials are reproducible measures of the extent of abnor-
low amplitude, polyphasic, and unstable. These mality that can then be followed over time.
have been called nascent MUPs. As reinnerva- Quantitation can assist in making decisions
tion proceeds to include more muscle fibers about the best treatment and prognosis, espe-
with better synchrony of firing, MUPs become cially the likelihood of improvement. As an
higher in amplitude, longer in duration, less example, NCS can define a 30% block, rather
polyphasic, and more stable. Therefore, late than axonal loss, as the cause of weakness in
after reinnervation, MUPs are of long duration a peroneal neuropathy. Such quantitation pro-
and high amplitude. Reinnervation is usually vides evidence that the process is relatively
completed by less than the normal number of mild and a good recovery can be expected.
axons; thus, recruitment is decreased because
there are fewer motor units. These changes are
summarized in Tables 47–1, 47–2, and 47–3.
Both conduction block and axonal disruption Defining Stage and Evolution
can have different time courses, depending on of Disease
the underlying mechanism and the number
of axons involved. The changes over time in Selecting a treatment for the cause of
compound muscle action potentials (CMAPs) symptoms requires identifying the underly-
and the results of needle EMG examination ing disease and its stage. It is not suffi-
after a localized nerve injury are shown in cient to know that a patient’s symptoms are
Tables 47–1, 47–2, and 47–3. caused by a localized lesion in the midthoracic
Table 47–1 Compound Muscle Action Potential Amplitude after

Peripheral Nerve Injury∗
Amplitude
0–5 Days After 5 days During recovery
Conduction block
Distal stimulation Normal Normal Normal
Axonal disruption
Distal stimulation Normal Low Increases
∗ Supramaximal stimulation.
Table 47–2 Results of Needle Examination after Peripheral Nerve Injury

0–15 Days After 15 days During recovery
Conduction block
Fibrillation potentials None None None
Motor unit potentials ↓ number ↓ number ↑ number
Axonal disruption
Fibrillation potentials None Present Reduced
Motor unit potentials ↓ number ↓ number Nascent
↓, decrease; ↑, increase.
Table 47–3 Interpretation of Electromyographic Findings after

Peripheral Nerve Injury
Finding Interpretation
0–5 Days
Motor unit potentials present Nerve intact, functional axons
Fibrillation potentials present Old lesion
Low-amplitude compound action potential Old lesion
5–15 Days
Compound action potential, distal only Conduction block
Low-amplitude compound action potential Axonal disruption
After 15 days
Compound action potential, distal only Conduction block
Fibrillation potentials Axonal disruption
Recovery
Increasing compound action potential Block clearing
Decreasing number of fibrillation potentials Reinnervation
Nascent motor unit potentials Reinnervation
spinal cord without knowing the nature of the process, the procedures vary in their ability
lesion. Thoracic cord disease caused by a her- to characterize the stage of evolution of the
niated disk, intraspinal tumor, arteriovenous disease or to provide prognostic information.
malformation, multiple sclerosis, or vitamin Compared with other neurophysiologic tech-
B12 deficiency is treated differently. Clinical niques, EMG and NCS are more consistent
neurophysiology can help define prognosis by for detecting abnormalities that typify the stage
classifying changes as acute, subacute, chronic, of development of an underlying neurologic
or residual from an old process. Whether the disorder. Clinically valuable prognostic infor-
disease is rapid, intermediate, slow, stable, or mation can be gained from EMG and NCS in
improving produces different clinical neuro- many peripheral nerve disorders. Identification
physiologic findings. An acute process develops of disease type with EMG is well known, but
within seconds to a few days. A subacute dis- changes in EMG findings with time are less
order evolves over a few days to weeks and familiar. Recognition of different stages in the
a chronic disorder, over months to years. In evolution of a disease depends on understand-
progressive diseases, there is increasing dam- ing the pathophysiologic changes that occur in
age and impairment of function. Improve- nerve and muscle. The three types of nerve
ment occurs when the disease process subsides damage—conduction block, slowing of con-
and neural mechanisms of repair can begin duction, and axonal destruction—evolve over
to reduce the severity of damage. In a sta- very different time courses. Secondary changes
ble process, damage has occurred but remains in muscle with each of these evolve over time
unchanged, because either the rate of neu- courses that vary with severity of disease.
ral reparative processes is able to keep pace
with the rate of neural damage in a chronic
continuous disorder or the disease process has Identifying Disease Types
subsided entirely but the damage cannot be
repaired. This usually is referred to as a resid- Clinical neurophysiology can facilitate identifi-
ual of the disease. cation of specific diseases. Electrophysiologic
testing can sometimes supplement the initial
classification of a disease as vascular, inflamma-
Determining Prognosis tory, degenerative, or neoplastic, but it does so
in different ways for different tests. In many
Although several clinical neurophysiologic pro- instances, only a broad category of disease can
cedures can define the severity of a disease be suggested, but in others, specific diseases
can be identified. These are described in detail cases. The following sections outline the util-
in the chapters on each of the techniques. The ity of electrophysiologic tests, discuss typical
following are some examples. findings, and review general guidelines and
approaches for different types of neuromuscu-
• NCS help distinguish demyelinating dis- lar problems. Protocols and algorithms to guide
ease from axonal disease. Marked slowing the assessment of these problems are included
and dispersion are evidence of a demyeli- in the accompanying CD.
nating neuropathy, which may be caused
by either an inherited or an acquired dis-
order. RADICULOPATHIES
• Repetitive nerve stimulation (RNS) can
demonstrate specific patterns of abnor- The diagnostic value of EMG in assess-
malities seen in myasthenia gravis (MG) ing patients with a suspected radiculopathy
and distinguish them from Lambert– includes answering the questions: (1) Is there
Eaton myasthenic syndrome. evidence of radiculopathy? (2) Which nerve
• Occasionally, EMG can assist in identi- root is involved in the radiculopathy? (3) How
fying specific disorders by characteristic severe is the neural damage caused by the
findings such as fibrillation potentials and radiculopathy? (4) Is the radiculopathy of
short duration MUP with polymyositis or recent onset, is it ongoing, or is it a resid-
radiation damage with myokymia. ual of an old lesion? (5) Is there evidence of
• Autonomic function testing can provide other peripheral nerve disease? EMG does not
evidence of specific disorders such as define the cause of the radiculopathy, and the
multisystem atrophy or reflex sympathetic findings on testing may be similar whether the
dystrophy. radiculopathy is caused by a disk, tumor, or dia-
betes mellitus. Because disorders of the nerve
Physicians and clinical neurophysiologists must roots produce changes only if the nerve fibers
be aware of the potential applications of clini- are damaged, EMG can never exclude the
cal neurophysiology and make full use of them. presence of a radiculopathy and EMG findings
The focus of this section is on considerations may be normal even when the radiculopathy
important in deciding whether one or more causes severe pain.
clinical neurophysiologic techniques are war-
ranted for a particular clinical problem and
on applications of testing and interpretation of Cervical and Lumbosacral
findings in different types of neuromuscular Radiculopathies
disorders.
NCS may be of assistance in the evaluation
of cervical and lumbosacral radiculopathies,
but have limitations. They are most useful
ASSESSING CLINICAL for identifying or excluding other peripheral
DISORDERS: ASSESSMENT WITH nerve disorders, such as mononeuropathies or
EMG AND NCS plexopathies, that may clinically mimic radicu-
lopathies. The most sensitive and important
Since the types and locations of neuromuscu- method used to evaluate radiculopathies is
lar disorders are relatively well defined, algo- needle EMG. Identifying abnormalities on the
rithms and approaches to testing a patient with needle examination confined to muscles inner-
a suspected specific disorder can be developed. vated by a common nerve root, and particularly
The algorithms must take into account findings involving the paraspinal muscles, helps to con-
obtained during the test in order to determine firm a radiculopathy. Furthermore, the specific
the amount and types of testing that should be spontaneous waveforms and MUP changes
performed. They are suggestions for EMG and assist in identifying the temporal profile of
NCS in neuromuscular disorders and nearly the radiculopathy and in the determination of
always need to be modified according to the whether there is ongoing denervation (“active”
particular problem and findings in individual radiculopathy). The utility and limitations of
each type of electrophysiologic tests are dis- marked loss of axons and can be residual of an
cussed below. old radiculopathy.
Sensory NCS. Sensory nerve action potentials
Motor NCS. The most commonly performed (SNAPs) should be normal in nerve root dis-
motor NCS are the median and ulnar in the ease, since root diseases typically involve the
arm and the peroneal and tibial in the leg. root proximal to the dorsal root ganglion. As
The most common abnormality that may be a result, SNAPs help in differentiating radicu-
seen with radiculopathies in motor NCS is lopathies from more peripheral diseases. The
a reduction in CMAP amplitudes, generally choice of sensory NCS should reflect the dis-
with mild or no slowing in the conduction tribution of sensory symptoms. Commonly per-
velocities depending on the severity of axonal formed sensory studies include the median and
loss. The degree of amplitude reduction corre- ulnar antidromic, radial, sural, or superficial
lates with the degree of axonal loss from Wal- peroneal.
lerian degeneration. However, these routine
motor NCS are limited in that they only assess Needle EMG. The most useful method for
the C8–T1 roots and L5–S1 roots. Therefore, identifying radiculopathy is needle EMG.
in patients with suspected radiculopathies in Since EMG changes evolve over time, the
other roots, routine NCS will be normal. Motor age of the lesion can be judged from both
NCS of other nerves can be performed, such the distribution and the type of abnormality.
as femoral (rectus femoris) for an L3–4 radicu- The severity of damage can be estimated by
lopathy, radial (extensor digitorum communis the amount of motor fiber degeneration (fib-
[EDC] recording) for C7–8 radiculopathy, and rillation potentials and MUP changes). Well-
musculocutaneous (biceps) for C5–6 radicu- defined fibrillation potentials are not seen until
lopathy. However, given the large size of the 3 weeks after nerve damage. Examination of
muscle being recorded in these NCS, abnor- the paraspinals is particularly important, since
malities will generally not be seen unless the changes in paraspinal muscles localize the pro-
radiculopathy is very severe. Furthermore, in cess proximal to the plexus. Proximal and
radiculopathies that primarily involve the dor- paraspinal muscles are the earliest to show fib-
sal root and cause pain and sensory loss but rillation potentials (evidence of axonal destruc-
no motor involvement, motor NCS will not tion) and also the earliest to show improve-
demonstrate any abnormalities. ment. Unfortunately, persistent abnormalities
in paraspinal muscles due to local muscle
F waves and H reflexes. Late responses, such as trauma following neck or back surgery pre-
F waves and H reflexes, can measure proximal clude postoperative testing and interpretation
conduction through the nerve roots. Similar of abnormalities in these muscles. The periph-
to standard motor NCS, ulnar and median F eral distribution of an abnormality defines the
waves assess conduction through the C8–T1 root involved. EMG is particularly valuable in
root; peroneal and tibial F waves through the differentiating relatively recent nerve damage
L5–S1 roots. The H reflex is most commonly with abundant fibrillations (especially in prox-
recorded from the soleus muscle and assesses imal muscles) from the residual of an old dis-
the S1 root. Occasionally, the C6–7 root can ease with scanty fibrillation potentials (mainly
be assessed if the median H reflex is recorded in distal muscles).
from the flexor carpi radialis (FCR). Prolon-
gation of the latencies of F waves indicates
proximal slowing in the motor fibers, while pro-
longation of the H-reflex latencies indicates Thoracic Radiculopathies
proximal slowing in the motor or sensory fibers.
In a small proportion of patients with lesions There are limited electrophysiologic studies
of the C7, C8, L5, or S1 nerve root, particu- that can be reliably used to evaluate for a
larly those with recent damage, the F waves thoracic radiculopathy. NCS are generally not
or H reflexes may be abnormal when other helpful, since there are no reliable techniques
measurements are normal. It must be recalled used to stimulate or record from the intercostal
that F waves and H-reflex latencies are gen- nerves. The exception is T1 radiculopathies,
erally most prolonged in a patient who has a in which the median motor amplitude may be
low if there is sufficient axonal loss. Needle mononeuropathies. While needle EMG cannot
examination of the thoracic paraspinal muscles precisely localize the problem to the region of
as well as the rectus abdominis or external the carpal tunnel, it is also important to help
oblique abdominal muscles may demonstrate define the degree of denervation in the thenar
findings in thoracic radiculopathies. muscles, as well as to exclude other super-
imposed disorders, such as proximal median
neuropathies or cervical radiculopathies.
COMMON FOCAL
MONONEUROPATHIES Motor NCS. The median motor NCS will often
be normal in patients with mild CTS; how-
Electrophysiologic changes on NCS in monon- ever, with more severe or long-standing nerve
europathies vary with the rapidity of devel- compression, the motor fibers may be affected.
opment, the duration and severity of dam- In most instances, the median nerve recorded
age, and the underlying pathologic condition. from the abductor pollicis brevis (APB) will
Localized narrowing of axons or paranodal or demonstrate prolongation of the latency with
internodal demyelination caused by a chronic stimulation at the wrist. Occasionally, slow-
compressive lesion produces localized slowing ing of the median motor conduction velocity
of conduction. Narrowing of axons distal to (CV) in the forearm may occur. With severe
chronic compression results in slowing of con- CTS in which axonal loss has occurred, the
duction along the entire length of the nerve. CMAP amplitude will be reduced. In cases
Telescoping of axons with intussusception of where no response can be recorded from the
one internode into another produces distortion APB, median and ulnar nerve stimulation while
and obliteration of the nodes of Ranvier and, recording from the second lumbrical and pal-
thus, results in a conduction block. Moderate mar interosseus may demonstrate prolongation
segmental demyelination and local metabolic of the median-lumbrical latency.
alterations are often associated with conduc- Sensory NCS. Sensory NCS are the most
tion block. With stimulation proximal to the sensitive and earliest affected in CTS. Pro-
site of damage, the conduction block is man- longation of the distal latency (DL) followed
ifested as lower amplitude evoked responses. by reduction in the SNAP amplitudes occurs
In an acute lesion with disruption of the axons, with increased severity of disease. Stimula-
the segment of nerve distal to the lesion may tion and recording over the shortest segment
continue to function normally for up to 5 days; of nerve, such as with orthodromic palmar
then, as the axons undergo Wallerian degener- studies, increases the sensitivity of the study
ation, they cease to conduct and the amplitude compared to stimulation and recording over
of the evoked response diminishes and finally longer nerve segments. Additionally, in very
disappears. One week after an acute injury, the mild CTS, comparison of the median nerve
amplitude of the evoked response is a rough latency with either the ulnar or the radial
gauge of the number of intact viable axons. nerve latencies recorded over the same dis-
Interpretations of the duration and severity tance of nerve segment may be the only
of nerve injury after focal neuropathies and abnormality.
radiculopathies can be made on the basis of an
analysis of the combination of these changes. Needle EMG. Needle examination of the APB
Examples of these interpretations are given in or opponens pollicis provides evidence of
Tables 47–1, 47–2, and 47–3. axonal damage when conduction studies are
still normal. Fibrillation potentials correlate
with the degree of axonal loss and long-
Carpal Tunnel Syndrome duration MUPs indicate chronicity of the syn-
drome. Examination of proximal median and
In patients with hand numbness and sus- other nonmedian nerve innervated muscles is
pected CTS, electrodiagnostic testing is useful useful to exclude other nerve disorders.
in confirming the localization to the median
nerve at the wrist and defining the severity The abnormalities on NCS and needle EMG
of the lesion. Sensory and motor NCS are may improve following treatment, such as sur-
the most useful studies to assess for median gical decompression. However, the findings
may not always resolve completely and some to exclude a C8–T1 radiculopathy or lower
degree of abnormalities may persist trunk plexopathy. Mild slowing along the entire
indefinitely. length of the ulnar nerve occurs with a moder-
ately severe C8 or lower trunk damage and loss
of faster conducting axons.
Ulnar Neuropathy Sensory NCS. Similar to the ulnar motor NCS,
the ulnar sensory NCS may demonstrate slow-
Similar to CTS, NCS provide important infor- ing of CV or reduction in amplitude. In fact,
mation in localizing a disorder to the ulnar the sensory NCS are often affected earlier
nerve. In patients with hand weakness and and more severely than the motor conduc-
sensory loss in the 4th or 5th digit, electrodi- tions. Conduction block is difficult to identify
agnostic testing is utilized to confirm an ulnar on sensory conduction studies, but may occa-
neuropathy and exclude C8 radiculopathies sionally be identified with stimulation above
or lower trunk/medial cord plexopathies. The and below the elbow when compared to the
confirmation of an ulnar neuropathy at the unaffected side. In ulnar neuropathies at the
elbow often relies on the identification of con- wrist, the dorsal ulnar cutaneous NCS should
duction slowing across the elbow (compared be spared even when the ulnar antidromic
to forearm nerve segments), conduction block study is involved. If a lower trunk plexopa-
across the elbow, or needle examination find- thy is considered and the ulnar sensory NCS
ings limited to muscles supplied by the ulnar is abnormal, medial antibrachial studies may
nerve. In most instances, the ulnar neuropa- be useful and will be abnormal in lower trunk
thy will be localized to the region of the medial plexopathies.
epicondyle or the cubital tunnel. Rarely, ulnar
neuropathies will occur at the wrist or in the Needle EMG. Needle EMG is used to assist
hand. in localizing the lesion and determine the
chronicity and degree of axonal loss. In ulnar
Motor NCS. The ulnar motor nerve conduction neuropathies at the elbow, abnormal find-
study is the most useful test to precisely localize ings are most severe in the FDI and ADM,
the site of ulnar nerve compression. Record- but forearm muscle abnormalities will con-
ing is most often performed from the abductor firm proximal damage if conduction studies
digiti minimi (ADM); however, in patients with are non-localizing. Examination of nonulnar
greater weakness in the first dorsal interosseus innervated muscles is important to exclude
(FDI), recording should also be made from alternative localizations, such as low cervical
this muscle. Well-defined criteria to identify radiculopathies or plexopathies.
an ulnar neuropathy at the elbow have been
published,1 and include slowing of CV across
the elbow of >10 m/second compared to the
forearm CV, reduction in the CMAP of >20% Peroneal Neuropathy
with elbow stimulation compared to the wrist
stimulation (in the absence of a median–ulnar The electrodiagnostic confirmation of per-
anastomosis), or focal conduction block identi- oneal neuropathy relies on identification of
fied with short segment stimulation across the conduction block or slowing along only the
elbow. In more severe ulnar neuropathies, the peroneal nerve. This usually involves the com-
CMAP amplitude may be reduced. In long- mon peroneal nerve at the fibular head. Iso-
standing ulnar neuropathies, slowing will be lated involvement of the superficial or deep
present along the entire nerve distal to the site peroneal nerve can occur. NCS will demon-
of the damage. In patients with only intrinsic strate focal slowing or conduction block in
ulnar hand muscle weakness, and with a nor- lesions characterized by focal demyelination,
mal ulnar motor response recorded from the and amplitude reduction when axonal loss
ADM, recording from the FDI should be per- occurs. When a focal demyelinating lesion is
formed and compared to the unaffected side to identified with routine studies, short segment
assess for a lesion involving the deep branch of (“inching”) studies can be performed, as in
the ulnar nerve in the hand. In addition, per- the ulnar nerve, to attempt to more precisely
formance of the median motor NCS is useful localize the lesion.
Motor NCS. The peroneal motor NCS, record- peripheral neuropathy from a polyradiculopa-
ing from either extensor digitorum brevis or thy, multiple lumbosacral radiculopathies, or
anterior tibialis muscles, are used to iden- from spinal cord or nonorganic disease. It also
tify focal demyelination. When >20% ampli- has utility in providing information regard-
tude reduction is present with knee stimula- ing underlying pathology and classification of
tion compared to ankle stimulation, stimulation neuropathies, such as differentiation of a pre-
at the fibular head should be performed. In dominant demyelinating vs. axonal neuropathy,
peroneal neuropathies characterized by axonal identifying involvement of motor and sensory
loss, CMAP amplitude reduction is typically fibers, and determining chronicity of the dis-
seen. Tibial NCS should also be performed to ease. Distinguishing these characteristics may
assess for a sciatic nerve lesion or lumbosacral assist in identifying the potential etiologies
plexopathy. With a long-standing peroneal neu- of neuropathy. Most neuropathies are char-
ropathy at the knee, slowing will be present acterized by a “length-dependent” pattern of
along the entire length of the nerve distal to nerve involvement, and the findings are most
the knee. evident in the most distal nerves and mus-
cles. In axonal neuropathies, motor and sen-
Sensory NCS. The superficial sensory nerve sory NCS amplitudes are often decreased
is the only peroneal sensory nerve that can or absent, with relative sparing of the con-
be studied. In most peroneal neuropathies, duction velocities and distal latencies. How-
this sensory nerve is affected early and ever, with sufficient loss of large, fast con-
more severely than motor fibers. However, ducting axons, a mild degree of conduction
in lesions characterized by focal demyelina- slowing may occur. In contrast, in predomi-
tion, the superficial peroneal sensory response nantly demyelinating neuropathies, significant
may be preserved since the conduction block CV slowing and prolongation of the latencies,
occurs proximal to the segment of the nerve out of proportion to amplitude reduction, are
studied. Also, in patients with isolated involve- prominent features. The findings of increased
ment of the deep peroneal nerve, the super- temporal dispersion or conduction block
ficial peroneal sensory response will also be are characteristic of segmental demyelination
normal. and suggest an autoimmune, or inflammatory
Needle EMG. Needle EMG is important to etiology.
confirm a peroneal neuropathy, localize the In most cases, the etiology of the neuropathy
lesion along the peroneal nerve when pre- cannot be determined by electrophysiologic
cise localization cannot be made with conduc- testing. Most toxic, metabolic, or nutritional
tion studies, and define the severity. Muscles neuropathies are predominantly axonal neu-
supplied by both the deep and the superfi- ropathies. Occasionally, clues to the underlying
cial branches of the peroneal nerve should etiology are noted by the pattern of findings,
be examined. When distal peroneal muscles such as the finding of a mixed axonal and
are abnomal, examination of the short head demyelinating, length-dependent neuropathy
of the biceps (the most proximal muscle along with paraspinal fibrillation potentials in
innerved by the peroneal nerve) is important diabetes or bilateral CTS superimposed on a
to identify the proximal extent of the lesion. largely axonal neuropathy in amyloidosis. With
Additionally, other L5 muscles not innervated many nerves and muscles available to test, the
by the peroneal nerve should be studied most appropriate ones must be selected. For
to exclude L5 radiculopathy or lumbosacral example, plantar nerves show earlier abnor-
plexopathy. malities in neuropathy than more proximal
sensory nerves.
NCS are the most direct measure of the
functionality of the nerve, and abnormali-
PERIPHERAL NEUROPATHY ties reflect the severity of the neuropathy,
the underlying pathology, and the duration
Electrodiagnostic testing is important to con- of the symptoms. In most “length-dependent”
firm the presence of a disorder involving the sensorimotor peripheral neuropathies, certain
peripheral nerves diffusely, and distinguish typical patterns of findings are commonly
a “length-dependent” or distal predominant seen.
• Sensory NCS are often affected earlier be useful to determine the underlying patho-
and more severely than motor NCS. physiology. The degree of amplitude reduc-
• Lower extremity NCS are affected earlier tion correlates with the degree of axonal loss.
and more severely than upper extremity Careful observation of the waveforms to assess
NCS. for abnormal dispersion or conduction block
• NCS abnormalities affect nerves of similar in demyelinating neuropathies is important.
length to a similar degree (e.g., peroneal = Since motor NCS reflect the integrity of the
tibial, median = ulnar). motor nerve fibers as well as the neuromuscu-
lar junction (NMJ) and muscle, interpretation
of abnormalities on motor NCS must be made
Although not all peripheral neuropathies
in the context of other findings on NCS and
demonstrate these features, caution should be
needle examination. Several general points are
taken in making the diagnosis of a “length-
important to consider in the interpretation of
dependent peripheral neuropathy” when vari-
motor NCS in the evaluation of peripheral
ations of these findings are seen. For exam-
neuropathy:
ple, with some exceptions, patients who have
predominantly upper extremity symptoms with • The amplitude and area of the recorded
abnormal NCS in the arms and normal lower
CMAP reflect the number and integrity of
extremity NCS are more likely to have another
the axons, while the DL and CV reflect
process, such as median or ulnar neuropathies.
both the integrity of myelin sheath and the
Patients who demonstrate significant asymme-
number of large conduction axons.
try in findings are more likely to have a radicu- • Loss of substantial number of large diam-
lopathy, polyradiculopathy, mononeuritis mul-
eter axons may lead to a slowing of the CV
tiplex, or focal mononeuropathy. Patients who
(and mimic “demyelination”). Therefore,
have unequal involvement of similar length
electrophysiologic criteria for demyeli-
nerves may have another process such as a
nation should be met (discussed later)
radiculopathy or mononeuropathy. For exam-
before a “demyelinating” neuropathy is
ple, a peroneal CMAP amplitude of 0.2 mV
diagnosed.
with a tibial CMAP amplitude of 12 mV would • The waveform morphology is as impor-
raise the possibility of a peroneal neuropathy or
tant as the absolute numerical data of
L5 radiculopathy.
each motor conduction study since tem-
Normal findings on electrophysiologic test-
poral dispersion and focal conduction
ing do not exclude peripheral neuropathy. The
block, which are indicators of segmental
results of NCS and EMG are often normal with
demyelination, cannot be identified with-
only small-fiber involvement in diabetic, amy-
out scrutiny of the waveforms.
loid, or some hereditary sensory neuropathies. • Standard motor NCS measure the con-
Alternatively, EMG may provide evidence of
duction in distal segments of the nerves,
nerve damage before it is evident clinically,
recorded from distal muscles. Since
such as in patients with complaints of vague or
in most sensorimotor peripheral neu-
nonspecific pain or those with diabetes.
ropathies the distal nerves are affected
The evaluation of patients with a suspected
more than the proximal nerves, abnor-
generalized neuropathy includes motor and
malities in routine motor NCS would be
sensory NCS and needle EMG. Each of these
expected. The proximal segment of the
studies provides important information related
nerve is less frequently affected and can
to peripheral neuropathy.
be assessed with proximal nerve stimu-
lation (e.g., at Erb’s point) or with late
Motor NCS. The most commonly performed reflexes, such as the F waves. When the
motor NCS are those that are the most distal proximal nerve segments are abnormal,
and likely to be involved early in a neuropathy, a disorder such as a polyradiculopathy is
such as the peroneal (EDB) and tibial (abduc- more likely than a peripheral neuropathy.
tor hallucis [AH]). If the responses to both of
these are absent or markedly low, a peroneal Disorders that produce abnormalities only on
motor NCS recording from the anterior tib- motor NCS while sparing sensory conductions
ialis or a median or ulnar motor NCS may are less common than sensorimotor peripheral
Table 47–4 Disorders Producing Low Motor CMAP Amplitudes without

Sensory Involvement
Diffuse, symmetric motor neuropathies Acute inflammatory demyelinating
polyradiculopathy (AIDP)
Chronic inflammatory demyelinating
polyradiculopathy (CIDP)
Paraproteinemia (e.g., monoclonal gammopathy
of undetermined significance)
Spinal muscular atrophy
GM1-associated motor neuropathy
Lead intoxication
Diffuse or focal, asymmetric motor Amyotrophic lateral sclerosis
neuropathies
Multifocal motor neuropathy with conduction
block
Monomelic amyotrophy (Sobue, Hirayama
syndrome)
Nonneurogenic disorders with low CMAP Distal myopathies (Inclusion body myositis,
amplitudes distal dystrophies)
neuropathies. Etiologies of motor predominant what is seen clinically with patients’ complaints
neuropathies or other disorders involving only of sensory symptoms and objective sensory
motor NCS to be considered are listed in loss prior to development of weakness. In the
Table 47–4. American Association of Neuromuscular and
Electrodiagnostic Medicine (AANEM) con-
sensus statement on identification of periph-
F waves. F-wave latencies provide a mea- eral neuropathy, an abnormal sural sensory
sure of proximal conduction and, in some dis- response along with an abnormality in at least
orders, may show early abnormality. F esti- one other nerve is considered the minimal
mates provide a simple, convenient method degree of abnormality necessary to support
to determine the proximal–distal distribution an electrodiagnosis of a distal sensorimotor
of conduction slowing. The F estimate is peripheral neuropathy.2 In more severe cases,
calculated from the limb length, peripheral upper extremity sensory NCS are also affected.
CV, and DL, thus predicting the F latency Several important issues are necessary to con-
on the assumption that the CV is the same sider with respect to sensory NCS in the evalu-
along the entire length of the nerve. If the ation of peripheral neuropathy.
F latency is the same as the F estimate, con-
duction is the same along the entire length,
• Normal sensory responses do not nec-
be it normal or slow (the latter suggests a
polyradiculoneuropathy). If the F latency is essarily exclude a peripheral neu-
less than the F estimate, then conduction is ropathy. A relatively preserved sural sen-
faster proximally as would be expected with sory response may not be “normal.” The
a peripheral neuropathy. If the F latency is amplitude may be significantly lower than
longer than the F estimate, then conduction is the patient’s normal value (e.g., a sural
slower proximally, as would be expected with a amplitude of 10 μV may fall within the
polyradiculopathy. normal laboratory range although it would
be 50% reduced in a patient who had a
Sensory NCS. Sensory nerves are more sus- sural amplitude of 22 μV two years prior).
ceptible to metabolic or toxic insults than Alternatively, the underlying process may
motor nerves and are therefore affected prior be affecting predominantly small sensory
to and more severely than motor nerves in fibers, which are not reliably assessed
most types of neuropathy; this is similar to with routine NCS. Finally, preservation
of the sensory responses in the con- are most severe in the distal lower extrem-
text of true sensory loss may indicate ity muscles, such as the abductor hallucis
that the pathology is proximal to the or peroneus tertius. When proximal muscles
dorsal root ganglion, such as diffuse are affected, other processes such as motor
involvement of the nerve roots in a neuron disease, polyradiculopathy, or multiple
polyradiculopathy. mononeuropathies would be considered. Fib-
• Diffusely abnormal sensory NCS do rillation potentials are often seen in axonal neu-
not necessarily confirm a general- ropathies, although in very slowly progressive
ized peripheral neuropathy. Techni- neuropathies, slow, continuous reinnervation
cal factors, such as low skin temperature, may reduce the degree of fibrillation poten-
can cause diffuse slowing of conduc- tials. The chronicity of the neuropathy can be
tion velocities or prolongation of the judged from the types of abnormalities in vol-
distal latencies, mimicking a peripheral untary MUPs. In primarily demyelinating neu-
neuropathy. ropathies, the abnormalities on needle EMG
• Lower extremity sensory NCS are may be minimal.
less reliable in the older population.
There are a limited number of sensory
NCS that can be reliably performed in Defining Pathophysiology of
the lower extremity (sural, superficial per-
oneal, medial plantar). A percentage of
Peripheral Neuropathies (Axonal
“normal” individuals over 55–60 years vs. Demyelinating)
do not have recordable lower extremity
SNAPs. When these are absent in older One of the most important roles of electro-
patients, the electromyographer cannot diagnostic testing is to provide information
reliably determine whether sensory fibers about the pathophysiology of the neuropathy,
are affected. which may be helpful in narrowing the list
of potential etiologies and assisting the refer-
Several disorders may produce a severe ring physician in the subsequent evaluation for
sensory neuropathy or neuronopathy the cause. Although predominantly demyeli-
(Table 47–5). In these cases, diffusely abnor- nating neuropathies are less common, identi-
mal or absent sensory responses on NCS occur fication of demyelination can lead to a more
with normal motor NCS and normal needle focused differential diagnosis, a higher poten-
examination. tial for response to treatment, and an overall
better prognosis for recovery. Specific findings
Needle EMG. Needle EMG is used to and criteria are used to determine whether the
assess the degree of axonal loss as well process is predominantly axonal or demyelinat-
as the chronicity of the neuropathy. In ing. Consideration of these criteria is important
length-dependent neuropathies, abnormalities for appropriate interpretation of the study. In
many instances, the findings may suggest a
combination of axonal loss and demyelination.
Table 47–5 Sensory Neuronopathies
or Ganglionopathies
Sjogren’s syndrome Axonal Neuropathies
Vitamin B6 toxicity
Vitamin B12 deficiency The majority of patients evaluated for sus-
Vitamin E deficiency pected peripheral neuropathy will demonstrate
Cis-platinum features of axonal loss. The primary pathol-
HIV ogy of axonal loss or degeneration is man-
HTLV-1 ifest on NCS as a reduction in the SNAP
Paraneoplastic sensory neuronopathy and CMAP amplitudes. In addition to ampli-
Syphilis tude reduction, a mild degree of motor and
Lyme
Spinocerebellar ataxias
sensory CV slowing and prolongation of the
Friedrich’s ataxia distal latencies is often present as a result of
either secondary demyelination or loss of the
larger, faster conducting axons. However, the of myelin. This can occur in a uniform and dif-
degree of slowing is mild and generally less fuse manner such as in Charcot–Marie–Tooth
than 30% of normal. When there is a signif- type I, or in a segmental and patchy manner
icant loss of axons, slowing of up to 50% of such as in chronic inflammatory demyelinating
the lower limit of normal may be seen solely polyradiculopathy (CIDP) (Table 47–7).
due to axonal loss. In contrast to demyelinating A number of different electrophysiologic cri-
neuropathies, conduction block and increased teria for demyelination have been proposed
temporal dispersion are not seen in axonal over the years utilizing different combina-
neuropathies. tions and degrees of CV slowing, prolonga-
The electrophysiologic features of axonal tion of DL, prolongation of F-wave latencies,
neuropathies are as follows: the presence of conduction block or temporal
dispersion, and the number of nerves that are
• CMAP or SNAP amplitude less than 70% required to demonstrate the abnormalities.3–8
of normal (with CV greater than 70% of These criteria should be used when making
lower limit of normal) or amplitudes less a diagnosis of a “demyelinating neuropathy,”
than 50% with any degree of CV slowing. rather than interpreting the study as a demyeli-
• Distal latencies less than 130% of normal. nating neuropathy when only slightly slowed
• No dispersion or block. CV or prolonged distal latencies are present.
• Normal F waves and blink reflexes. Characteristic features of demyelinating neu-
• Needle EMG—Fibrillation potentials in ropathies include the following:
distal greater than proximal muscles
variable). CV slowing—The degree of CV slowing
• Needle EMG—long-duration, high- required to indicate a predominantly demyeli-
amplitude MUP (in chronic disease). nating neuropathy varies among the published
criteria, but is generally taken as less than
70%–80% of the lower limit of normal. Several
Axonal neuropathies are due to a vast num-
criteria take the CMAP amplitude into consid-
ber of etiologies, some of which are listed in
eration when using the criteria. As noted previ-
Table 47–6.
ously, loss of large, faster conducting axons will
produce a mild degree of CV slowing. There-
fore, if there is evidence of axonal loss (as iden-
Demyelinating Neuropathies tified by low CMAP amplitudes), the degree
of CV slowing to indicate concomitant pri-
Demyelinating neuropathies are pathologically mary demyelination should be less than 50% of
characterized by primary or predominant loss LLN. Using these criteria, according to Mayo
Table 47–6 Examples of Etiologies of Axonal Sensorimotor

Peripheral Neuropathies
Metabolic Diabetes, hypothyroid, chronic liver disease,
chronic renal failure, critical illness
Vitamin deficiency B12, thiamine, copper, zinc
Vasculitis (connective tissue disease) Periarteritis nodosa, SLE, rheumatoid arthritis,
cryoglobulinemia
Toxins Arsenic, thallium, lead
Medications Metronidazole, phenytoin, hydroxychloroquine,
colchicine, vincristine, taxol, simvistatin,
cyclosporine
Inherited Hereditary motor sensory neuropathy type II
Paraprotein associated MGUS, myeloma, amyloid
Infectious Syphilis, CMV, HIV
Table 47–7 Predominantly Demyelinating Neuropathies

Acquired (segmental demyelination)∗
Acute inflammatory demyelinating polyradiculopathy (AIDP)
Chronic inflammatory demyelinating polyradiculopathy (CIDP)
Paraprotein-associated neuropathy
Monoclonal gammopathy
Osteosclerotic myeloma
Waldenstrom’s macroglobulinemia
Diphtheria
Medications and toxins (perhexiline, amiodarone, toluene)
Hereditary (uniform demyelination)
Hereditary motor sensory neuropathy type I (CMT)
Hereditary motor sensory neuropathy type III (Dejerine Sottas)
Congenital hypomyelinating neuropathy
Leukodystrophies
Metachromatic
Globoid cell (Krabbe’s)
Adrenomyeloneuropathy∗
Cerebrotendinous xanthomatosis
Congenital hypomyelinating neuropathy
Pelizeus–Merzbacher disease∗
Refsum’s disease∗
∗ May demonstrate multifocal conduction slowing.
EMG laboratory normative data, evidence for

demyelination in commonly performed motor
NCS requires conduction velocities in the fol-
lowing ranges:
Normal CMAP Low CMAP

amplitude amplitude
Peroneal <29–33 m/second <21 m/second Figure 47–1. Abnormal temporal dispersion in the
Tibial <28–32 m/second <20 m/second median nerve in CIDP.
Median <34–38 m/second <24 m/second
Ulnar <36–41 m/second <25 m/second finding suggests an acquired etiology when
it occurs in more than one nerve. This is
often visualized as a “ratchety” or “serrated”
DL prolongation—The DL must be prolonged waveform, rather than a smooth negative peak
to ≥125%–150% of upper limit of normal. (Fig. 47–1).
Abnormal temporal dispersion—Abnormal tem- Conduction block—Conduction block is defi-
poral dispersion is defined as a reduction in ned as failure of an axon to conduct an action
CMAP amplitude at a proximal site of stim- potential and is another characteristic feature
ulation compared to a distal site, along with of focal demyelination. In a normal motor
an increase in the duration of the CMAP by nerve, all of the axons within the nerve conduct
>15%–30%. The finding of abnormal tem- an action potential along the nerve at a rela-
poral dispersion is characteristic of demyeli- tively equal rate. When the action potentials
nating neuropathies and indicates an increase of all of the axons within a nerve are blocked
in the range and variability in the conduc- (complete conduction block), stimulation prox-
tion times of axons within a nerve due to imal to the block will produce no response
segmental or multifocal demyelination. This when recorded over a distal muscle, whereas
and area than with distal stimulation. The cri-

teria for defining conduction block vary from
between 20% and 60% reduction in CMAP
amplitude and area from distal to proximal
sites of stimulation9 (Table 47–8). These per-
centages depend on the nerve being studied
and the distance between the two stimulation
sites—the shorter the distance between the
two sites of stimulation, the lower the required
reduction in amplitude or area. For exam-
ple, stimulation at the ankle and knee sites
Figure 47–2. Radial motor nerve “inching” conduction along the peroneal nerve may demonstrate a
study demonstrating focal conduction block. 15% reduction in amplitude and area and still
be considered normal. However, in the same
stimulation distal to the block will produce a nerve, a 15% reduction in amplitude and area
normal response since the distal portion of the over a 2-cm segment across the fibular head
nerve remains intact (Fig. 47–2). If only some would be considered abnormal. Careful atten-
of the axons within a nerve are blocked (partial tion to technical factors is important when
conduction block), the CMAP to proximal stim- interpreting conduction block. Submaximal
ulation will be significantly lower in amplitude stimulation at the proximal site where the
Table 47–8 Electrodiagnostic Criteria for Partial Conduction Block and Percent
Reduction in Amplitude or Area
Defining criteria Percent reduction in amplitude or area
Definite CMAP area difference Median, ulnar >50% or 40%

(proximal to distal) greater than
50%, regardless of distance or
CMAP duration (except for
tibial nerve study)
CMAP area difference Peroneal, tibial >60% or 50%
30%, less than 15% difference
in CMAP duration, regardless
of distance
CMAP area difference
20% over distances of 10 cm or
less, regardless of CMAP
duration
CMAP area difference
10% over a distance of 2 cm,
regardless of CMAP duration
Probable (with no temporal Median, ulnar 40%–50% or
dispersion) 30%–40%
Probable (with temporal Median, ulnar >50% or 40%
dispersion)
Peroneal, tibial >60% or 50%
Possible CMAP area difference
30%, regardless of distance or
CMAP duration
nerve may be deeper (e.g., the tibial nerve at horn cells in the spinal cord, the appropri-
the knee or the brachial plexus at Erb’s point) ate CMAP amplitudes may be reduced with
may mimic conduction block. injuries at each of these sites, and abnormal-
ities in motor amplitudes alone do not con-
Prolonged F-wave latencies or blink reflex firm a plexopathy. In some instances, segmen-
latencies—Prolongation of the F-wave laten- tal conduction slowing or conduction block
cies is an indicator of proximal demyelina- through the plexus may occur. “Routine” upper
tion. In some immune-mediated neuropathies, limb NCS (median and ulnar motor) evalu-
such as acute inflammatory demyelinating ate only the lower trunk or medial cord of the
polyradiculopathy (AIDP) and CIDP, proximal plexus, and therefore other, less commonly per-
segments of the nerves and nerve roots may be formed, motor NCS may be necessary when
affected early in the course of the disease. The involvement of other segments of the plexus is
criteria of degree of prolongation above nor- suspected.
mal vary from prolongation of >120%–150% Sensory NCS. SNAP abnormalities occur early
of upper limit of normal. in the course of plexopathies, usually prior
to abnormalities on motor NCS. Abnormali-
ties in sensory nerve responses indicate a pro-
BRACHIAL PLEXOPATHY cess distal to the dorsal root ganglia, whereas
normal sensory responses raise the likelihood
In patients with suspected brachial plexopathy, of a preganglionic process such as a cervi-
electrodiagnostic studies play an important role cal radiculopathy. In plexopathies, a reduction
in helping to confirm the clinical suspicion of or absence of SNAP amplitude is most com-
brachial plexus involvement, localize the site(s) monly seen. Comparison of the findings in
of involvement within the plexus, exclude alter- the affected sensory nerve to the same nerve
native sites of localization (e.g., lower trunk vs. on the unaffected side, even in studies where
C8 root vs. ulnar nerve), identify subclinical the results are within the absolute normal lab-
plexus involvement (e.g., in neuralgic amyotro- oratory limits, is important to identify rela-
phy), define the severity and determine the tive reductions in amplitude and, generally,
degree of reinnervation (e.g., traumatic plex- a 50% side-to-side reduction in amplitude is
opathy), and occasionally define the etiology abnormal.
(e.g., neoplastic vs. radiation plexopathy). An
organized approach in the use of NCS and Needle EMG. Needle EMG is the most
needle examination is crucial to diagnose and efficient method of evaluating the different
localize lesions of the brachial plexus. Exten- components of the brachial plexus. The
sive and uncommon NCS and a liberal needle decision on muscles to be examined requires
examination may be required to assess each the use of localization from the clinical exami-
major component of the plexus. nation and findings on NCS. The optimal strat-
NCS are commonly performed as a first step egy in performing the needle examination is to
in the electrodiagnostic evaluation. The selec- test muscles supplied by each major periph-
tion of studies to be performed depends on the eral nerve and each major spinal root, and
index of suspicion of a plexus lesion as well as then to “narrow” the localization by examin-
the presumed localization based on the clinical ing two or more muscles supplied by each cord
history and examination. and trunk, finding the “common link.” If all
of the muscles involved can be localized to a
Motor NCS. Motor NCS are useful in their single nerve or single root, caution should be
ability to assess the integrity of the motor taken in making the diagnosis of “brachial plex-
axons coursing through different segments of opathy.” The clinical examination is important
the plexus and by their ability to stimulate the in defining which muscles should be exam-
trunks at the supraclavicular fossa (Erb’s point). ined, as muscles that are clinically involved
In most plexopathies, CMAP amplitude reduc- should be tested during the needle exami-
tion is a more common finding than slowing nation. The proximal extent of the damage
of the CV. Since lesions of the root, plexus, must be defined by the examination of proxi-
or individual nerves all may produce Walle- mal muscles (infraspinatus, rhomboids, serra-
rian degeneration of axons distal to the anterior tus anterior, and occasionally the diaphragm)
and cervical paraspinals. Examination of the antebrachial cutaneous, median antidromic

unaffected side may be helpful in identifying (thumb), dorsal radial sensory, and median
subclinical involvement, such as often occurs antidromic (index). The most common muscles
in neuralgic amyotrophy. The findings on nee- affected on the needle examination include
dle examination reflect the temporal profile deltoid, teres major (axillary nerve); brachio-
and the underlying pathophysiology of plexus radialis (BR), extensor carpi radialis (ECR)
injury. In patients with acute or early involve- (radial nerve); biceps, brachialis (musculocu-
ment, or plexopathies characterized predom- taneous nerve); and PT, FCR (median nerve).
inantly by neuropractic lesions, the only or Abnormalities may also occur in the infraspina-
predominant findings may be reduced recruit- tus and supraspinatus (suprascapular nerve)
ment. In patients in whom substantial axonal depending on whether the injury occurred
loss has developed and who are examined proximal to the takeoff of the suprascapular
more than 3 weeks following the insult, fibril- nerve from the upper trunk (Table 47–9).
lation potentials and MUP changes (increased Middle trunk. NCS assessing the middle
polyphasia and long-duration MUPs) are seen. trunk are limited. The radial (EDC) motor
In some circumstances, unusual spontaneous studies may be affected in some instances.
discharges may support an underlying eti- The most reliable median sensory NCS is the
ology, such as the presence of myokymic median antidromic (middle finger), although
discharges in patients with radiation-induced the median antidromic (index) is affected in
plexopathy. 80% of lesions and the radial sensory is affected
in 40%. Needle examination may demon-
strate abnormalities in the triceps, anconeus,
Trunk Lesions and EDC (radial nerve); and PT and FCR
(median nerve). Suggested studies to evalu-
Upper trunk. Routine motor NCS do not assess ate for middle trunk plexopathies are listed in
the upper trunk segments. Motor conduction Table 47–10.
studies that may demonstrate abnormalities
when significant axonal loss occurs include the Lower trunk. Routine NCS are most use-
axillary (deltoid), musculocutaneous (biceps), ful in evaluating the lower trunk lesions.
and suprascapular (infraspinatus). In rare These include median motor (APB), ulnar
instances, or when studied early in the course motor (ADM or FDI), ulnar antidromic (5th),
of the onset of symptoms, conduction block and medial antebrachial cutaneous. Needle
between proximal and distal stimulation points examination demonstrates abnormalities in all
along these nerves may be seen. Sensory NCS ulnar-innervated muscles: APB, flexor pollicis
that may show abnormalities include the lateral longus (FPL), and pronator quadratus (PQ)
Table 47–9 Upper Trunk Assessment

Nerve conduction studies Needle examination
Sensory Supraspinatus
Median—Thumb Infraspinatus
Median—Index Deltoid
Radial (dorsum) (Teres minor)
Lateral antebrachial cutaneous Biceps
Ulnar—5th∗ Brachioradialis
Supinator
Motor Pronator teres
Axillary
Musculocutaneous Rhomboid major∗
Suprascapular Triceps∗
Median (APB)∗ FDI∗
Ulnar (ADM)∗ Upper cervical PSP∗
∗ Expected to be normal.
Table 47–10 Middle Trunk Assessment

Sensory Pronator teres

Median—Index Flexor carpi radialis
Median—Middle Triceps
Radial
Median—Thumb∗ Deltoid∗
Ulnar—5th∗ Biceps∗
Brachioradialis∗
Motor FDI∗
Radial Midcervical PSP∗
Median (APB)∗
Ulnar (ADM)∗
Table 47–11 Lower Trunk Assessment

Sensory First dorsal interosseus

Ulnar—5th Abductor pollicis brevis
Medial antebrachial Flexor pollicis longus
Median—Index∗ Extensor indicis proprius
Radial∗ Flexor carpi ulnaris
Flexor digitorum profundus
Motor
Median (APB) Deltoid∗
Ulnar (ADM) Biceps∗
Pronator teres∗
Triceps∗
Midcervical PSP∗
(median nerve); and extensor indicis proprius antebrachial cutaneous, both of which may
(EIP) and extensor pollicis brevis (EPB) (radial be abnormal in lateral cord lesions. Needle
nerve) (Table 47–11). examination abnormalities occur in the biceps
and brachialis (musculocutaneous nerve), and
PT and FCR (median nerve). Sparing of
Cord Lesions the axillary nerve-innervated muscles (deltoid,
teres minor), suprascapular-innervated mus-
Posterior cord. NCS used to evaluate the pos- cles (infraspinatus, supraspinatus), and upper
terior cord include the axillary (deltoid), radial trunk or radial nerve-innervated muscles (bra-
(EDC), and radial sensory. Needle examina- chioradialis, supinator) helps to distinguish
tion abnormalities occur in deltoid, teres minor lesions of the lateral cord from the upper trunk
(axillary nerve), and radial innervated muscles or posterior cord (Table 47–13).
(BR, triceps, ECR, supinator, EDC, EIP, etc).
In looking at the list of muscles involved, it
Medial cord. NCS are helpful in assess-
is evident that muscles supplied by all roots
ing median cord lesions, including median
(C5–T1) and all trunks may show abnormalities
motor (APB), ulnar motor (ADM or FDI),
in posterior cord lesions (Table 47–12).
ulnar antidromic (5th), and medial ante-
Lateral cord. NCS that may be useful brachial cutaneous. Needle examination abnor-
include musculocutaneous (biceps) and lateral malities occur in all ulnar-innervated muscles
Table 47–12 Posterior Cord Assessment

Sensory Deltoid
Radial Teres minor
Median—Index∗ Brachioradialis
Ulnar—5th∗ Supinator
Motor Triceps
Radial Extensor carpi radialis
Axillary Extensor digitorum communis
Median (APB)∗ Extensor indicis proprius
Ulnar (ADM)∗ Biceps∗
Pronator teres∗
FDI∗
Mid cervical PSP∗
Table 47–13 Lateral Cord Assessment

Sensory Biceps
Median—Thumb Pronator teres
Median—Index Flexor carpi radialis
Median—Middle
Lateral antebrachial Brachioradialis∗
Radial (dorsum)∗ Deltoid∗
Ulnar—5th∗ Rhomboid major∗
Triceps∗
Motor FDI∗
Musculocutaneous Upper cervical PSP∗
Median (APB)∗
Ulnar (ADM)∗
and several median-innervated muscles (APB, plexopathies is the complex anatomy and
FPL, opponens pollicis). The sparing of mus- sometimes the patchy nature of involvement of
cles supplied by the C8–T1 root, but the poste- different elements of the plexus in many cir-
rior cord or radial nerve (EIP, EPB) localized cumstances. Completely isolated involvement
the lesion to the lower trunk. Another impor- of a single trunk or cord is less common than
tant point is that the NCS in lower trunk and predominant involvement of one site with less
medial cord lesions will be identical with the severe involvement of other sites. Therefore,
exception of the radial (EIP) (Table 47–14). electrodiagnostic testing should study compo-
nents that may not be involved, in addition
Tables 47–15 and 47–16 may assist in the inter- to those that are clinically involved. Since a
pretation of the findings on NCS and needle brachial plexopathy may involve any of the
examination, and may help to find the “com- many nerves of the upper extremity, the eval-
mon link” in localizing to the appropriate site uation of a brachial plexopathy is best modified
of the plexus. on the basis of the clinical deficit and suspected
One of the factors contributing to the site of involvement. The following are the sug-
difficulty in the electrodiagnostic evaluation gested approaches that may be used to evaluate
and interpretation of findings in brachial lesions at different sites of the plexus.
Table 47–14 Medial Cord Assessment
Sensory First dorsal interosseus

Ulnar—5th Abductor pollicis brevis
Medial antebrachial Flexor pollicis longus
Median—Index∗ Extensor indicis proprius
Radial∗ Flexor carpi ulnaris
Flexor digitorum profundus
Motor
Median (APB) Deltoid∗
Ulnar (ADM) Biceps∗
Pronator teres∗
Triceps∗
Midcervical PSP∗
Table 47–15 Segments of the Brachial Plexus Assessed by Specific NCS

Upper trunk (C5–6) Middle trunk (C6–7–8) Lower trunk (C8–T1)
Lateral cord Lateral antebrachial Median sensory (index)

cutaneous
Median sensory (thumb) Median sensory (middle)
Median sensory (index)
Posterior cord Radial sensory Radial sensory
Axillary motor Radial motor (EDC)
Medial cord Ulnar sensory (5th digit)
Dorsal ulnar cutaneous
sensory
Medial antebrachial sensory
Median motor (ABP)
Ulnar motor (ADM, FDI)
Table 47–16 Muscle Innervation According to Trunk, Cord, and Nerve

Upper trunk (C5–6) Middle trunk (C6–7–8) Lower trunk (C8–T1)
Lateral cord Biceps (MC) Pronator teres (M)

Brachialis (MC) FCR (M)
Posterior cord Brachioradialis (R) Triceps (R) Extensor indicis proprius (R)
Supinator (R) Anconeus (R) Extensor carpi ulnaris (R)
Deltoid (Ax) Extensor carpi radialis (R)
Teres minor (Ax) Extensor digiti communis (R)
Triceps (R) Extensor carpi ulnaris (R)
Medial cord Abductor pollicis brevis (M)
First dorsal interosseus (U)
Abductor digiti minimi (U)
Flexor carpi ulnaris (U)
Flexor digitorum profundus (U)
Flexor pollicis longus (M)
SSc, suprascapular nerve; MC, musculocutaneous nerve; Ax, axillary nerve; R, radial nerve; M, median nerve; U, ulnar
nerve.
821
Differentiating root from plexus lesions. It that may account for the symptoms. In most
is often difficult to completely differentiate cases, generalized weakness is not caused by
between processes involving the roots from the a neuromuscular disease. However, disorders
trunks of the plexus. The sparing of SNAPs in that should be considered include myopathies,
clinically affected areas of sensory loss and the NMJ disorders, polyradiculopathies, or motor
identification of abnormalities on needle EMG neuron diseases. The presence of true, objec-
in cervical paraspinals are indicative of a pro- tive weakness raises the possibility of a disorder
cess involving the roots. Several other findings at one of these levels, and the distribution
may be useful: of objective weakness helps to determine the
type and extent of testing that is performed.
• C5–6 vs. upper trunk—The pattern of NCS are important not only to obtain objec-
findings on needle examination may be tive evidence of a disorder of the motor nerves
similar, although the presence of abnor- or muscles, but to assess for diseases of the
malities in the rhomboids or serratus NMJ. Needle examination of distal and prox-
anterior implicates a process proximal imal muscles can help to identify the type of
to the trunks. Sensory nerve conduc- underlying disorder and more precisely deter-
tion abnormalities in the lateral ante- mine the localization.
brachial cutaneous or median nerves
occur in upper trunk plexopathies but not Motor NCS. In patients with weakness, motor
radiculopathies. NCS may demonstrate a number of abnor-
• C7 vs. middle trunk—Distinguishing bet- malities that assist in localizing the process
ween lesions at these sites is diffi- along the peripheral neuroaxis. In motor neu-
cult. Fortunately, isolated middle trunk ron diseases or polyradiculopathies, low CMAP
plexopathies are extremely rare. Low- amplitudes may be present. Focal demyeli-
amplitude radial or median sensory respon- nation, as evidenced by increased tempo-
ses on NCS may occur with middle trunk ral dispersion or conduction block, in motor
plexopathies. Needle examination find- nerves may be seen in acquired demyelinat-
ings in the serratus anterior and cervical ing polyradiculopathies such as CIDP or in
paraspinals indicate a root lesion. multifocal motor neuropathy with conduction
• C8–T1 vs. lower trunk—Lesions of the C8 block. In disorders of neuromuscular transmis-
and T1 roots affect the ulnar and median sion, routine motor NCS may be normal or low
motor NCS but do not affect the ulnar amplitude (especially in Lambert–Eaton myas-
or medial antebrachial cutaneous sensory thenic syndrome). Since myopathies typically
NCS. Needle examination demonstrates involve proximal muscles and standard motor
abnormalities in the intrinsic hand mus- NCS are recorded from distal muscles, rou-
cles and medial forearm muscles in both tine CMAP amplitudes will often be normal in
localizations. primary myopathies.
F waves. F waves may demonstrate pro-
Electrophysiological testing is also used in the longation in patients with weakness due to
preoperative assessment of patients who have polyradiculopathies, particularly early in the
had traumatic injury of the brachial plexus. The course when other abnormalities on conduc-
Mayo Clinic brachial plexus outpatient proto- tion studies may not be evident.
col that is performed in all patients who are
plexus or spinal accessory surgery candidates Repetitive nerve stimulation studies. The per-
includes a thorough performance of NCS. The formance of repetitive stimulation studies
protocol is described in Table 47–17. should be considered in all patients who com-
plain of generalized weakness, as they will
occasionally identify an unsuspected NMJ dis-
GENERALIZED WEAKNESS order. The extent of testing depends on the
level of suspicion of an NMJ disorder; if
Generalized weakness is a common complaint. high, several distal and proximal nerve–muscle
Patients with weakness are often referred combinations should be performed. In patients
for electrophysiologic testing to assess for a in whom motor CMAP amplitudes are low
number of possible neuromuscular disorders and Lambert–Eaton myasthenic syndrome is
Table 47–17 Mayo Clinic Brachial Plexus Outpatient EMG Protocol

Common approach Other considerations
Motor NCS Median (APB) Standard NCS are needed in each

Ulnar (ADM) nerve with weakness or sensory loss
Radial (EDC) in its distribution.
Musculocutaneous
Spinal accessory
Sensory NCS Median (antidromic) Consider doing sensory and/or motor
Ulnar (antidromic) studies on opposite side for comparison if
Radial findings are borderline or where
Lateral antebrachial cutaneous otherwise appropriate.
Needle EMG Cervical paraspinal (two levels)
Biceps (both heads)
Triceps (lateral and medial heads)
First dorsal interosseous
If there is weakness in upper trunk
distribution, test additional C5 & C6
muscles
Rhomboid , infraspinatus, deltoid,
brachioradialis, lateral brachialis, medial
brachialis
If there is weakness in middle trunk
distribution, test additional C7 muscles
Pronator teres, extensor carpi radialis
If there is weakness in lower trunk
distribution, test additional C8 muscles
Extensor carpi ulnaris, flexor carpi
ulnaris, flexor pollicis longus
considered, repetitive stimulation should be possibly long duration if due to a neurogenic

performed before and after 10 seconds of exer- disorder.
cise to assess for facilitation.
Sensory NCS. These are useful to perform in

patients complaining of weakness, even if they MYOPATHY
have no sensory complaints, in order to assess
for subclinical sensory involvement. If abnor- There are a variety of techniques available to
mal, a disorder of the peripheral nerves would study the muscle, including motor and sen-
become more likely than diseases primarily sory NCS, repetitive stimulation, and needle
affecting muscle, NMJ, or anterior horn cells. EMG. Electrodiagnostic testing is an impor-
However, some disorders, such as amyloido- tant step in the evaluation of patients with sus-
sis or sarcoidosis, may affect nerve and muscle pected myopathies and can help to (1) confirm
(neuromyopathies). a clinically suspected myopathy; (2) exclude
other disorders that may mimic myopathies,
Needle EMG. Needle examination can deter- such as polyradiculopathies, motor neuron dis-
mine the underlying pathology, when weak ease, or NMJ disorders; (3) provide clues to
muscles are examined. In a weak muscle, vol- the etiology of the myopathy; (4) assist in the
untary MUPs will be of short duration, low selection of a muscle for biopsy; and (5) assess
amplitude, and polyphasic if due to a myopa- response to treatment. Early in the course of
thy; varying in amplitude if due to an NMJ dis- a myopathy, the findings on NCS and nee-
ease; or demonstrate reduced recruitment and dle EMG may be patchy or subtle, requiring
persistence and thoroughness in widespread nerve and muscle function, and are much
sampling of muscles. EMG is also limited in more sensitive to change in neuropathic dis-
its ability to distinguish between different eti- orders. However, in some myopathies, non-
ologies. Although the electrodiagnostic find- specific abnormalities may occasionally be
ings reflect the underlying muscle fiber pathol- seen.
ogy and physiology, in many disorders similar
pathologic changes may be occurring within Motor NCS. Loss of a sufficient number of
the muscle and the findings are not specific axons in neurogenic disorders may rapidly lead
for individual diseases. Furthermore, in some to reduction in the recorded CMAP amplitude;
types of myopathies, electrodiagnostic studies however, a substantial direct loss of muscle
demonstrate no abnormalities at all, which con- fibers will less significantly reduce the CMAP
founds the overall assessment of the patient’s amplitude. It is estimated that over 50% of
disorder. muscle fiber loss in a muscle is necessary
The findings in certain myopathies, such to produce reduction in the CMAP ampli-
as inflammatory myopathies, also evolve over tude. Conduction velocities, distal latencies,
time, beginning with small MUPs and quickly and F-wave latencies are typically normal and
developing fibrillation potentials and polypha- unaffected in muscle diseases, unless concomi-
sic MUPs. The regenerating muscle fibers and tant nerve dysfunction is present. The muscle
fibers that have lost their innervation because fiber CV may be slowed in some myopathies,
of nerve terminal damage, segmental necrosis, although this requires specialized technique of
or fiber splitting produce a number of fibrilla- direct muscle stimulation to identify.
tion potentials that roughly parallel the degree In most myopathies, standard motor NCS
of disease severity. As the disease subsides, fib- (e.g., median, ulnar, peroneal, and tibial) are
rillation potentials become less prominent and usually normal, since these study the dis-
MUPs have a more normal size. The number tal nerve or muscle sites and the majority
of muscle fibers in some motor units increases, of myopathies affect predominantly proximal
resulting in larger than normal MUPs late in muscles. Myopathies with distal muscle weak-
the disorder. ness are more likely to demonstrate low CMAP
NCS are an integral part of the electrodi- amplitudes (Table 47–18). Proximal NCS may
agnostic evaluation of patients with suspected demonstrate low CMAP amplitudes in more
myopathies, even though in most patients the severe myopathies.
results are normal. NCS assess either nerve The major utility of motor NCS is to exclude
function in isolation or the combination of alternative etiologies that may be confused
Table 47–18 Myopathies with Low CMAP

Amplitudes on Distal NCS
Muscular dystrophies
Myotonic dystrophy
Facioscapulohumeral dystrophy
Emery–Dreifuss muscular dystrophy
Distal muscular dystrophies
Welander, Miyoshi, Markesberry–Griggs
Inflammatory myopathy
Inclusion body myositis
Severe polymyositis
Metabolic myopathy
Debrancher enzyme deficiency
Congenital myopathy
Nemaline myopathy
Central core myopathy
Centronuclear myopathy
Myofibrillary myopathy (desmin)
clinically with myopathies, such as motor neu- Needle EMG. Needle EMG is a way of
ron disease, multifocal motor neuropathy with efficiently sampling multiple muscles in a
conduction block, or NMJ disorders. widespread distribution and extrapolating the
underlying pathologic changes that may be
Repetitive nerve stimulation. RNS is useful occurring within the muscle fibers. The find-
to exclude NMJ disorders and should be ings on needle examination often correlate
normal in myopathies. In patients with sus- with the pathologic changes on muscle biopsy,
pected weakness due to myopathy, baseline thereby assisting in the selection of an appro-
2-Hz repetitive stimulation in one or two priate muscle for biopsy. The combination of
nerve/muscle groups should be performed. In spontaneous activity and MUP changes reflects
some myotonic disorders with a disturbance the underlying pathologic reactions occurring
of sarcolemma function, such as myotonic dys- in the muscle fibers.
trophy, myotonia congenita, and paramyotonia
Fibrillation potentials. In myopathic dis-
congenita, RNS may demonstrate decrement
orders characterized by fiber necrosis,
at low or high rates of stimulation at baseline,
fibers splitting, or intracellular vacuole
which repairs immediately following exercise
formation, fragments of muscle fibers are
and worsens after several minutes, similar to
separated from the innervation terminal
MG. In these disorders, however, the CMAP
nerve twig. This “functional denervation”
amplitude is typically reduced immediately
leads to the development of fibrillations
after the exercise.
in those separated segments. Fibrillation
Sensory NCS. Sensory NCS are routinely potentials in myopathies often demon-
normal in myopathies. In cases where the strate features that are different from
SNAP amplitudes are reduced, disorders those in neurogenic disorders. They may
producing both myopathy and neuropathy be of low amplitude, have a slower firing
(“neuromyopathy”) should be considered rate, are more often positive waveform,
(Table 47–19). However, two separate pro- and may occur in a patchy distribution
cesses could also account for the combina- within a muscle. Myopathies that occur
tion of findings, especially in patients with with fibrillation potentials are listed in the
underlying medical diseases that are known to table (Table 47–20).
produce peripheral neuropathy, such as dia- Myotonic discharges. Myotonic discharges
betes. are seen in disorders where instability of
Table 47–19 Disorders Causing Myopathy and Neuropathy

Connective tissue disorders Infiltrative disorders
SLE Amyloidosis
Rheumatoid arthritis Sarcoidosis
Mixed connective tissue disease
Sjogren’s
Polyarteritis nodosa
Endocrine disorders Infectious diseases
Acromegaly HIV
Addison’s Toxoplasmosis
Hyperthyroid
Hypothyroid
Metabolic myopathies Medications
Acid maltase deficiency Vincristine
Debrancher enzyme deficiency Chloroquine
Colchicine
Cyclosporin
Congenital myopathies Alcohol
Myofibrillary (desmin)
Mitochondrial disorders
Table 47–20 Myopathies Associated with Fibrillation Potentials

Inflammatory Congenital myopathies
Polymyositis Myofibrillary (desmin)
Dermatomyositis Nemaline
Inclusion body myositis Centronuclear
Dystrophies Metabolic myopathies
Dystrophinopathies (Duchenne’s, Becker’s) Acid maltase deficiency
Sarcoglycanopathies Debrancher enzyme deficiency
Limb-girdle muscular dystrophy
Fascioscapulohumeral dystrophy
Distal dystrophies
Myotonic dystrophy
Toxic myopathy Critical illness myopathy
Cholesterol lowering agent myopathy (CLAM)
Acute alcohol myopathy
Colchicine Disorders producing rhabdomyolysis
Azidothymidine (AZT) Trauma
Cyclosporin Drugs (heroin, phencyclidine)
Emetine Prolonged coma
Penicillamine
Infiltrative Infectious
Amyloid Trichinosis
Sarcoidosis Viral myositis
HIV myopathy
muscle fiber membrane, due to defec- myotonic disorders. Infrequent myotonic

tive sodium or chloride channels, pro- discharges may be seen in a variety
duces repetitive spontaneous firing of of myopathies (Table 47–21) and even
muscle fiber action potentials. The iden- briefly in long-standing neurogenic dis-
tification of myotonic discharges is use- orders. However, profuse and promi-
ful in narrowing the differential diag- nent myotonic discharges are only seen
nosis; however, there are no specific in a few types of disorders, particularly
features of myotonic discharges that the myotonic dystrophies and channelo-
allow distinction between the different pathies.
Table 47–21 Myopathies Associated with Myotonic Discharges

Prominent
Myotonic dystrophy (DM1)
Proximal myotonic myopathy (PROMM, DM2)
Myotonia congenita
Paramyotonia congenita
Hyperkalemic periodic paralysis
Schwartz–Jampel syndrome
Infrequent
Polymyositis
Dermatomyositis
Inclusion body myositis
Toxic myopathies
Cholesterol lowering agents
Chloroquine
Cyclosporin
Amyloid myopathy
Motor unit potentials. In myopathic dis- myopathies where the pathologic pro-
eases, random loss, atrophy, or variation cess affects the muscle contractile appa-
of size of muscle fibers in a motor unit ratus or muscle membrane, muscle fiber
leads to short-duration, low-amplitude, necrosis, regeneration, and motor unit
and polyphasic MUPs. The total number remodeling often will not occur, and
of motor units within a muscle is usu- therefore the findings on needle exam-
ally unchanged in myopathies; however, ination are normal. Alternatively, some
loss of individual fibers within motor units disorders will only affect type II mus-
leads to less force production from each cle fibers, which are not assessed ade-
motor unit. Therefore, more motor units quately by routine EMG. Myopathies that
than normal must be recruited to gener- may demonstrate a normal EMG include
ate a force. The initial firing frequency steroid-induced myopathies, some con-
of the motor units is normal, but more genital myopathies, metabolic myopathies,
motor units fire with low effort (rapid endocrine myopathies, and some
recruitment). Rapid recruitment makes sarcoglycanopathies.
assessment of individual MUPs difficult,
since multiple potentials fire at low effort.
In severe or end-stage myopathies, loss of MYALGIAS, MUSCLE STIFFNESS,
all muscle fibers constituting entire motor
units may produce reduced recruitment,
AND EPISODIC MUSCLE
which can sometimes be mistaken for a WEAKNESS
neurogenic process.
A number of myopathies may be character- Many different central and peripheral disor-
ized pathologically by muscle fiber atro- ders may present with muscle stiffness. Central
phy or dysfunction of structural com- processes such as rigidity and spasticity are best
ponents, without fiber destruction. In assessed clinically, because EMG test results
these cases, while clinical weakness may are normal. The evaluation of peripheral disor-
be present and changes in MUP mor- ders is based on the character of the symptoms.
phology may occur, fibrillation poten- If the major complaint is episodic weakness,
tials and other spontaneous discharges with or without muscle stiffness, the patient
are notably absent. Furthermore, short- should be tested for periodic paralysis. If the
duration MUPs can be seen in severe or major complaint is episodic myalgia without
long-standing NMJ disorders, which can true muscle stiffness, the patient should be
mimic myopathies. tested for a myopathy. A muscle biopsy or
Long-duration MUPs may also be seen with ischemic forearm exercise or lactate test (or
any chronic or long-standing myopathy. both) is helpful if myalgia or contractures
This finding often leads to a misdiagnosis develop with exercise. If the major complaint
of a neurogenic disorder, such as amy- is muscle stiffness, consider doing the needle
otrophic lateral sclerosis (ALS), as iden- examination first to confirm the presence and
tification of the shorter duration, lower nature of spontaneous activity. If myotonic dis-
amplitude MUP may be masked by the charges are found, NCS may help define their
superimposed larger units. In this sit- source.
uation, quantitative MUP analysis may
result in normal mean values of ampli-
tude and duration of the recorded MUP. NMJ DISORDERS
Several explanations have been proposed
for the origin of long-duration polypha- Electrophysiologic testing is an important step
sic MUPs, including (1) reinnervation of in the evaluation of patients with suspected
regenerating muscle fibers with slowed NMJ disorders. While in some cases the clin-
conduction in immature nerve terminals, ical presentation of patients with diseases such
(2) increased scatter of the endplate as MG or Lambert–Eaton myasthenic syn-
zone, (3) slowed conduction in regen- drome is classic, leaving little doubt about the
erating muscle fibers, and (4) hypertro- diagnosis, in many instances patients present
phy of regenerated muscle fibers. In with vague or mild weakness, or symptoms that
are difficult to easily localize to the NMJ. For for NMJ disorders. The diagnosis of a defect
example, a patient who presents with a feeling of neuromuscular transmission is supported
of generalized weakness may have a disorder by identification of a reproducible decrement
involving the muscle (myopathy), peripheral of 10% or more in the CMAP amplitude
nerves (polyradiculopathy), anterior horn cells and area following RNS at 2–5 Hz, ideally
(ALS), or even no neuromuscular disease. In in two or more nerves.10, 11 The sensitivity of
these cases, electrophysiologic testing can help RNS depends on the distribution of clini-
to localize the disease to the NMJ or iden- cally affected muscles and disease severity.12, 13
tify another mimicking disease. Furthermore, In patients with generalized MG, bulbar and
when a disease of the NMJ is identified, EMG proximal muscles are usually more affected
can help to define whether the disorder affects clinically than distal muscles and the diagnos-
the presynaptic (Lambert–Eaton myasthenic tic yield of RNS is typically higher in proximal
syndrome) or postsynaptic junction (MG), can nerves compared to distal nerves, such as mus-
help to define the severity of the disease, and culocutaneous or biceps, axillary or deltoid,
in some instances can be used to follow the and spinal accessory or trapezius.10–12 Despite
patient to objectively assess response to treat- the higher sensitivity, RNS of proximal nerve–
ment. The approach to the patient with a muscle combinations is technically more dif-
suspected NMJ disease depends on the index ficult to obtain reliable and consistent results
of clinical suspicion for an NMJ disorder. In due to instability of baseline as a result of more
patients in whom an NMJ disease is strongly limb and stimulator movement. Proximal stim-
suspected, testing may be more comprehen- ulation may also be more uncomfortable for
sive, particularly when the clinical features are the patient. As a result, a standard protocol in
relatively mild, than in those in whom there is many laboratories begins initially with RNS on
a low suspicion. a distal nerve–muscle combination, such as the
Several types of electrophysiologic studies ulnar or peroneal nerves, followed by proximal
are important in assessing the NMJ, each of or cranial nerves.
which provides important and often comple- The choice of the nerve–muscle combi-
mentary information that is used to determine nations for RNS ultimately depends on the
the underlying disease: distribution of clinical weakness. A common
approach consists of ulnar (ADM) or peroneal
Routine motor NCS (assessing the CMAP (AT) if symptoms are worse in the legs, spinal
amplitudes): Routine motor NCS are important accessory (trapezius), and facial. If these nerves
to perform prior to repetitive stimulation stud- do not demonstrate significant decrement and
ies. In most cases of postsynaptic NMJ dis- there is a high clinical suspicion, one should
orders, such as myasthenia gravis, the CMAP strongly consider performing RNS on other
amplitudes are normal on routine motor NCS. nerves, such as radial (EIP or anconeus), mus-
However, in severe cases of myasthenia gravis, culocutaneous (biceps), axillary (deltoid), or
the CMAP amplitudes may be low as a result trigeminal (masseter).
of severe blockade of neuromuscular trans- Performing repetitive stimulation at differ-
mission. In contrast, in presynaptic NMJ dis- ent rates can provide important information
orders, such as Lambert–Eaton myasthenic about the type of NMJ disorder and can assist
syndrome or botulism, the CMAP ampli- in determining whether the disorder involves
tudes on motor NCS are typically low. Fol- the presynaptic or postsynaptic junction. How-
lowing exercise or electrical stimulation at ever, it is important to understand that both
rapid rates (e.g., >20 Hz), the CMAP ampli- a severe postsynaptic NMJ disorder and a
tudes increase (facilitate). In LEMS, facilita- presynaptic disorder may demonstrate similar
tion occurs rapidly following brief (10 sec- findings with slow and fast rates of stimula-
onds) exercise or electrical stimulation; in tion. Stimulation at slow rates (2–5 Hz) will
botulism, facilitation may occur only after maximize the degree of decrement by max-
longer (e.g. 1–2 minutes) of exercise or elec- imizing the release of immediately available
trical stimulation. stores of acetylcholine. Decrement at these
slow rates of stimulation may be seen with
Repetitive nerve stimulation. RNS is a reli- both presynaptic and postsynaptic junction
able and important technique used to assess disorders.
Cholinesterase inhibitors can repair abnor- findings as are seen in length-dependent

malities on RNS and lessen the severity peripheral neuropathies, sometimes leading
of findings on single fiber electromyography to misinterpretation of the electrodiagnos-
(SFEMG). Pyridostigmine should be discon- tic study. By definition, the pathology in
tinued 6 hours (long acting form 12 hours) polyradiculopathies is located at the root level.
prior to performing an electrodiagnostic study. In some cases, the dorsal root ganglion or a
Immunosuppressive therapy may also render more distal component of the nerve may also
a patient seronegative and improve abnormali- be involved (“polyradiculoneuropathy”). Clini-
ties on electrodiagnostic testing, although mild cal symptoms include sensory loss (often distal
changes on SFEMG usually persist. predominant), weakness, and areflexia. Weak-
ness is often diffuse, patchy, and symmetric
Needle EMG. While decrement on RNS is or asymmetric, and involves proximal mus-
strongly supportive of an NMJ disorder, it is cles to the same degree or more than distal
not specific and decrements (even greater than muscles.
10%) may be seen in other disorders, such The electrophysiologic features of polyradi-
as progressing neurogenic diseases or mus- culopathies may be similar to those of periph-
cle channelopathies. As a result, the findings eral neuropathies. The key distinguishing fea-
on NCS alone are not completely sufficient tures are the identification of abnormalities in
to confirm the diagnosis. Needle EMG pro- proximal nerve segments (ideally roots) and the
vides important complementary information identification of neurogenic abnormalities in
about the underlying disorder. In NMJ dis- proximal muscles in addition to distal muscles.
eases, the most prominent finding on nee-
dle EMG is MUP variation. MUP durations
and amplitudes are typically normal in mild Motor NCS. Motor NCS may demonstrate
cases, but may be short or low in more low CMAP amplitudes, particularly in axonal
severe cases. Long-duration or highly polypha- polyradiculopathies. Careful assessment for
sic MUPs should not be seen in NMJ disorders; abnormal temporal dispersion, particularly in
if they are identified, another disease, such proximal segments of the nerve, is important
as motor neuron disease or a superimposed when assessing for demyelinating polyradicu-
neurogenic process, should be considered. lopathies such as CIDP. This may require
Single fiber EMG. SFEMG is very sensitive stimulation in proximal nerve segments such as
(100% abnormal if a clinically involved muscle Erb’s point.
is examined), but is nonspecific. Abnormalities Sensory NCS. Sensory NCS may be entirely
of increased jitter and blocking are frequently normal if the process involves the pregan-
observed in myopathies and neurogenic disor- glionic root. However, many disorders may
ders associated with denervation. SFEMG is involve the dorsal root ganglia or even periph-
usually reserved for patients with normal find- eral portions of the nerves, thereby leading to
ings on standard NCS (including RNS) and low amplitude or absent SNAPs.
concentric needle EMG. It is most useful in
making the diagnosis of MG in very mild cases F waves. F waves are particularly important
or excluding it in cases with atypical features in assessing polyradiculopathies. F-wave laten-
or chronic fatigue. SFEMG is time-consuming, cies may be prolonged, particularly if there is
requires a cooperative patient, and is abnormal demyelination in the proximal portions of the
in any disease associated with a mild disorder roots or nerves. In more severe cases, F waves
of NMT. may be absent.
Needle EMG. The NCS findings (apart from
F waves) in polyradiculopathies may be similar
POLYRADICULOPATHY to that in length-dependent peripheral neuro-
pathies. Needle EMG can efficiently assess
Disorders involving multiple nerve roots, with proximal components of the nerves and roots,
or without distal peripheral nerve involvement, and are important to define the anatomic dis-
may clinically present similar to length- tribution of involvement. In contrast to length-
dependent peripheral neuropathy. The elect- dependent peripheral neuropathies, fibrillation
rodiagnostic findings may demonstrate similar potentials and long-duration MUPs are found
in proximal upper (and possibly lower) extrem- Sensory NCS. Sensory NCS are typically
ity muscles, especially in the paraspinals. normal in pure motor neuron diseases. How-
Mild or early polyradiculopathies may only ever, minor abnormalities, such as slightly low
involve the roots, with changes seen only in amplitudes, may be occasionally seen in ALS.
the paraspinals. Furthermore, in disorders such as spinobulbar
muscular atrophy (Kennedy’s disease), SNAP
amplitudes are commonly low.
MOTOR NEURON DISEASE Needle EMG. Gradual loss of anterior horn
cells in motor neuron disease produces
Electrodiagnostic studies are important in sus- changes in the EMG findings during the course
pected motor neuron diseases to localize the of the disease (Table 47–22). These changes
problem to the anterior horn cells or motor allow electromyographers to assess the evo-
nerves, exclude other mimicking disorders lution of the disease as well as its severity.
such as myopathies or polyradiculopathies, and In the initial stages of the disease, before
to assess for temporal profile and progression clinical weakness is evident, collateral sprout-
of the disease. ing of viable motor neuron axons maintains
innervation of all muscle fibers; thus, few if
Motor NCS. Motor NCS may be normal or any fibrillation potentials are evident. How-
of low amplitude in motor neuron diseases. ever, the loss of MUPs can be recognized.
Assessment for focal motor conduction block Later, MUP size increases with innervation of
or abnormal temporal dispersion is particu- greater numbers of muscle fibers. If a sig-
larly important to identify and, when present, nificant amount of collateral sprouting has
may indicate a disorder such as multifocal occurred, some MUPs vary in configuration. As
motor neuropathy with conduction block or a these changes progress to the stage where rein-
demyelinating polyradiculopathy. nervation cannot keep pace with denervation,
Table 47–22 Evolution of Electromyographic (EMG) Changes in Motor Neuron

Disease
Subacute active Chronic active Inactive or
residual
Motor NCS
CMAP amplitude Normal, unless Low (if severe) Low (if severe)
severe
Conduction velocity Normal Up to 30% slowed Up to 30% slowed
(if severe) (if severe)
Distal latency Normal Up to 30% Up to 30%
prolonged (if prolonged (if
severe) severe)
Repetitive stimulation Mild decrement in Mild decrement in Normal
some some
Needle EMG
Fibrillation potentials Many Many Few, small
Fasciculation potentials Frequent in mildly Frequent in mildly Rare of absent
affected, absent in affected, absent in
severly affected severly affected
Complex repetitive discharges None Rare Occasional
Motor unit potentials Reduced Reduced Reduced
recruitment, long recruitment, long recruitment, long
duration, high duration, high duration, high
amplitude amplitude amplitude
May be unstable May be unstable
fibrillation potentials become prominent. Dur- diagnoses when in fact the study may be
ing this time, larger numbers of regenerating normal. The most commonly encountered
fibers are present and intermittent blocking anomalous variations are the median-to-ulnar
of the components of an MUP, motor unit anastomoses (“Martin-Gruber anastomosis”,
potential variation, becomes more evident. “median-to-ulnar crossover”) and accessory
The potentials become increasingly polypha- peroneal nerve.
sic, with satellite potentials. This combination
of polyphasic and varying MUPs is evidence of
a severe, progressing disorder. At times, it is Martin-Gruber Anastomosis
accompanied by a decrement on slow repeti-
tive stimulation. The Martin-Gruber anastomosis (MGA) is a
common variation that occurs in 15%–31% of
The diagnosis of definite amyotrophic lat- individuals and is bilateral in up to 68% of
eral sclerosis requires upper motor neuron individuals. In this variation, a communication
and lower motor neuron signs at three lev- exists between the median and the ulnar nerve
els of the nervous system. Other lower motor fibers in the forearm, whereby fibers that are
neuron syndromes may have similar EMG destined to supply ulnar-innervated muscles
findings, including spinal muscular atrophy, course through the median nerve in the upper
residuals of poliomyelitis, hexosaminidase A arm and upper forearm, and “cross over” to
deficiency, multifocal motor neuropathy, pure the ulnar nerve in the forearm before inner-
motor inflammatory neuropathy, demyelinat- vating the destined muscles. The fibers may
ing neuropathy, lead neuropathy, porphyria, branch off the median nerve proper or the
Fazio–Londe disease (cranial), focal motor anterior interosseus branch. Sensory fibers are
neuron disease (Sobue’s), arteriovenous mal- not involved. In rare instances, proximal MGA
formation of the cord, syrinx, and paraneoplas- may occur with origin of the crossing over
tic syndromes such as lymphoma or radiation. fibers located at an above elbow site, thereby
resembling an ulnar neuropathy at the elbow.14
The muscles supplied by the crossing over
FACIAL WEAKNESS fibers vary among individuals and include one
or more of the following: (1) first dorsal
NCS and needle EMG can be used to evaluate interosseus, (2) abductor digiti minimi, and
patients with unilateral or bilateral facial weak- (3) adductor pollicis or flexor pollicis bre-
ness. Electrophysiologic studies are helpful to vis. In approximately half of individuals with
localize the process to the facial nerve and, in MGA, only one muscle is innervated by the
some instances, determine the site of injury crossing over fibers (FDI > adductor pollicis
along the nerve. Additionally, testing can help >ADM).15–18 Three types of MGA occur, all of
to identify subclinical involvement of other which produce a different pattern on routine
cranial nerves or assess for a more gener- median and ulnar motor NCS.
alized neuromuscular disorder causing facial
weakness. Finally, testing is useful for prog- TYPE I MGA. CROSSOVER FIBERS
nostication. In particular, comparison of the SUPPLY HYPOTHENAR MUSCLES
facial motor amplitude on the involved side (ADM)—FALSELY “ABNORMAL”
with that on the unaffected side in a unilateral ULNAR NCS
facial neuropathy can be used to determine In type I MGA, the crossing over fibers sup-
prognosis. ply the ADM through the crossover. Therefore,
during a routine ulnar motor NCS recording
from the ADM, stimulation of the ulnar nerve
ANOMALOUS INNERVATION at the elbow does not include the crossing over
fibers to the ADM, whereas stimulation at the
Variations in peripheral nerve anatomy are wrist includes these fibers. The result is often a
common and are important to recognize by drop in CMAP amplitude and area (sometimes
the electrodiagnostician. Failure to recognize more than 20%) between the wrist and the
anomalous anatomy may lead to erroneous elbow, simulating a focal conduction block such
Table 47–23 Findings of a Type I Martin-Gruber anastomosis

a) Ulnar nerve stimulation (recorded from ADM)
a. CMAP with above elbow (AE) stimulation > 20% lower than wrist (Wr) stimulation
b. CMAP with below elbow (BE) stimulation > 20% lower than wrist stimulation (AE = BE)
b) Median nerve stimulation (recorded from APB)
a. Normal results. CMAP with above elbow stimulation slightly lower than wrist stimulation
c) Median nerve stimulation (recorded from ADM)
a. Initial negative CMAP obtained with elbow stimulation (normal individuals will not have any
response)
b. No response with wrist stimulation
Figure 47–3. Type I Martin-Gruber anastomosis. Ulnar nerve stimulation recording from the ADM (left traces) demon-
strating CMAP amplitude drop between wrist and below elbow stimulation sites. Stimulation of the median nerve
demonstrates a response from the ADM with elbow stimulation from the crossing fibers, but not with wrist stimulation
(right traces).
as in an ulnar neuropathy. Below elbow stimu- at the elbow, in addition to stimulating all
lation will yield a similar result as above elbow true median-innervated muscles, will also stim-
stimulation since the fibers have not crossed ulate muscles in the thenar region supplied
over yet (Table 47–23). by the crossing over fibers. This produces
Type I MGA can be confirmed by stimulat- a higher CMAP amplitude than would nor-
ing the median nerve at the wrist and elbow mally be produced due to the additive mus-
while still recording over the ADM. With a cle action potentials of the adductor pollicis,
crossover, a CMAP response with an initial flexor pollicis brevis, and/or FDI. When the
negative deflection will occur with elbow stim- median nerve is stimulated at the wrist dis-
ulation but no response will occur with wrist tal to the site of the crossover, the fibers to
stimulation (Fig. 47–3). these muscles are not depolarized (Fig. 47–4)
(Table 47–24).
In contrast to type I MGA, type II MGA
TYPE II MGA. CROSSOVER FIBERS
is difficult to confirm, since volume conduc-
SUPPLY THENAR-REGION
tion from ulnar muscles in the thenar region
MUSCLES (ADDUCTOR POLLICIS,
will always produce a CMAP response with
FLEXOR POLLICIS BREVIS, OR
ulnar nerve stimulation at the elbow and wrist
FIRST DORSAL
sites, even in individuals without MGA. In
INTEROSSEUS)–FALSELY
some cases, a much higher response with ulnar
“ABNORMAL” MEDIAN NCS
wrist stimulation than elbow stimulation sup-
In type II MGA, crossing over fibers supply ports a type II MGA. In addition, a similar
ulnar muscles adjacent to the thenar eminence. pattern of findings can be seen with overstim-
The result is that median nerve stimulation ulation of the median nerve at the elbow or
Figure 47–4. Type II Martin-Gruber anastomosis. With median nerve stimulation recording from the APB, the elbow
CMAP amplitude (upper trace) is higher than the wrist CMAP (second trace). Stimulation of the ulnar nerve while record-
ing from the ABP (traces 3 and 4) demonstrates the volume-conducted response from ulnar muscles near the thenar
eminence.
Table 47–24 Findings of a Type II Martin-Gruber anastomosis

a) Median nerve stimulation (recorded from the APB)
a. Higher CMAP amplitude with above elbow stimulation than wrist stimulation
b. Change in CMAP morphology between the above elbow and the wrist stimulation sites
b) Ulnar nerve stimulation (recorded from the ADM)
a. Normal results. CMAP with above elbow stimulation is slightly lower than wrist stimulation
c) Ulnar nerve stimulation (recorded from the APB)
a. CMAP recorded at wrist and above elbow stimulations, with larger amplitude at wrist
stimulation
b. Possible initially negative CMAP with above elbow stimulation (this is not always seen)
c. CMAP recorded at wrist stimulation, with initially positive takeoff
understimulation at the wrist in individuals an initial positivity at both sites of stimulation

without MGA. indicates that the problem is not due to inap-
propriate placement of the active (G1) elec-
trode over the endplate region. The reason
TYPE II MGA IN CARPAL TUNNEL
for this finding is that, with elbow stimula-
SYNDROME
tion, there is focal slowing or partial conduction
A distinctive pattern is seen in up to 20% block at the wrist of the true median fibers
of individuals with a type II MGA and innervating the median muscles in the thenar
a median neuropathy at the wrist (“carpal eminence (e.g., APB, opponens pollicis). How-
tunnel syndrome”).19, 20 In this situation, an ever, the crossing over fibers that supply other
initial positive deflection is seen in the CMAP adjacent muscles in the thenar region (such
waveform with median nerve stimulation at the as the adductor pollicis or flexor pollicis bre-
elbow, whereas with wrist stimulation no initial vis) do not course under the transverse carpal
positivity occurs (Fig. 47–5). The absence of ligament and therefore have slightly faster CV.
Figure 47–5. Type II Martin-Gruber anastomosis with superimposed CTS. The positive deflection with elbow stimulation
reflects CMAP originating from the adductor pollicis and deep head of flexor pollicis brevis that are supplied through the
crossing over fibers. The CMAP recorded from the APB is slowed at the site of compression at the wrist, thereby unmasking
this volume-conducted response.
Since these muscles are at a distance away from TYPE III MGA. (COMBINATION OF
the G1 electrode and depolarize slightly sooner TYPES I AND II)
than the median-innervated thenar muscles,
they are recorded as a volume-conducted Type III MGA is simply a combination of types
response (positive deflection). I and II. Table 47–25 describes the findings
seen on routine median and ulnar motor
In fact, this volume-conducted response NCS.
from the ulnar-innervated crossed over fibers
occurs in all individuals with type II MGA;
however, because the median fibers (in the Riche–Cannieu Anastomosis—“All
absence of CTS) conduct slightly faster than Ulnar Hand”
the crossing over fibers, the positive deflection
is “hidden” in the major negative CMAP and The Riche–Cannieu anastomosis is an uncom-
is not seen. The finding of a positive deflec- mon anatomic variation, but one that very often
tion with elbow stimulation has been rarely leads to misinterpretation of findings. While
described in patients with CTS with otherwise this anastomosis has been identified in approx-
normal conduction studies.20 imately 19% of hands in cadaveric studies, it
Table 47–25 Findings of Type III Martin-Gruber anastomosis

a) Ulnar NCS (recorded from the ADM)
a. CMAP with above elbow (AE) stimulation > 20% lower than wrist (Wr) stimulation
b. CMAP with below elbow (BE) stimulation > 20% lower than wrist stimulation (AE = BE)
b) Median NCS (recorded from the APB)
a. Higher CMAP amplitude with AE stimulation than wrist stimulation
b. Change in CMAP morphology between the AE and the wrist stimulation sites
c) Median NCS (recorded from the ADM)
a. Initial negative response obtained with elbow stimulation
b. No response with wrist stimulation
is seen in a very small percentage of patients

undergoing NCS.21 In this anastomosis, the
thenar muscles and therefore all of the hand
muscles are supplied by the ulnar nerve. In
most cases, the deep branch of the ulnar nerve,
which normally lies deep in the palm of the
hand, courses over to innervate the thenar
muscles. In the purest form, all hand muscles
are innervated by the ulnar nerve.22 Therefore,
on median motor NCS no response is obtained
in the APB with median nerve stimulation at
the elbow or wrist.
Clues to the presence of this anastomosis are
normal muscle bulk and strength and normal
findings on needle examination in the thenar
Figure 47–6. Accessory peroneal nerve. Higher per-
muscles, despite an absent CMAP on median oneal CMAP amplitude with knee stimulation compared
motor conduction studies. In a “partial” Riche– to ankle stimulation. Stimulation behind the lateral mal-
Cannieu anastomosis, there is some median leaolus produces a CMAP.
nerve supply to the thenar muscles, although
the ulnar nerve supplies the majority of mus-
cles or fibers. In this case, a present but low nerve and course around the lateral malleolus
CMAP amplitude will occur with median nerve before innervating the EDB.25, 26 This anatomic
stimulation. anomaly is termed an accessory deep peroneal
In most cases, the Riche–Cannieu anasto- nerve or accessory peroneal nerve. The acces-
mosis is not clinically significant. However, sev- sory peroneal nerve may innervate fibers of the
eral cases have been described where needle EDB to a variable degree. In most cases, only a
examination abnormalities were seen in the percentage of fibers (usually the lateral fibers)
APB with an ulnar nerve lesion or abnormali- are supplied by the accessory branch, with the
ties in the ADM with CTS.23, 24 Without knowl- medial fibers supplied directly by the deep
edge of this anomaly, these findings could be peroneal nerve. However, in some cases, the
misinterpreted. accessory peroneal nerve supplies the entire
The Riche–Cannieu anastomosis is con- muscle (Fig. 47–6) (Table 47–26).
firmed by the presence of a normal CMAP In most situations, this finding is of no
response, with an initially negative deflection, clinical significance. However, in lesions of
with stimulation of the ulnar nerve and record- the deep peroneal nerve, clinical and electro-
ing from the APB. Caution must be made physiologic sparing of the toe extensors could
in interpreting this finding, as a response will occur in the presence of an accessory per-
always be obtained due to volume conduc- oneal nerve. Also, in laboratories where the
tion from neighboring ulnar-innervated mus- peroneal nerve is stimulated at the ankle before
cles. Additionally, if there is a true median the knee, caution should be made when a
nerve lesion, with significant atrophy of the low CMAP amplitude is obtained, which could
thenar muscles, a much higher response may be misinterpreted as the result of underlying
occur with ulnar nerve stimulation, indicating pathology, and stimulation at the knee should
the need for a careful clinical examination prior always be performed.
to or during the study.
Accessory Peroneal Nerve UNEXPECTED FINDINGS ON

NERVE CONDUCTION STUDIES:
In most individuals, the EDB is supplied by CAUSE AND ACTION
a branch of the deep peroneal nerve. In up
to 28% of the population, the axons supplying During the electrodiagnostic evaluation of
the EDB travel within the superficial peroneal most patients, the examining physician usually
Table 47–26 Findings of an Accessory Peroneal Nerve

a. Higher CMAP amplitude with stimulation at the knee and fibular head than with stimulation at the ankle.
(This finding can also be seen with understimulation at the ankle or overstimulation at the knee.)
b. Stimulation behind the lateral malleolus produces an initially negative CMAP, recorded from the EDB.
c. CMAP amplitude recorded from lateral malleolus stimulation typically equals the difference between the
knee and the ankle CMAP amplitudes.
has a general hypothesis of what the underly- of cranial conduction studies and brain stem
ing disease may be. Therefore, certain findings auditory evoked potentials. Spinal cord dis-
are often expected to be found or not found ease produces alterations in EMG, NCS, and
during NCS and needle examination. How- somatosensory evoked potentials. Peripheral
ever, the expected findings are not always seen, diseases show changes on NCS, EMG, and
which may be due to a different localization or autonomic function testing.
disease than what was initially expected. How- The multiplicity of different neurophysio-
ever, technical and other physiologic factors logic measures that can be applied in periph-
may also result in unexpected findings on test- eral disorders is sometimes assisted by applying
ing. It is important for the electromyographer guideline protocols based on the patient’s clini-
to be able to recognize and rectify any technical cal findings and what is found during testing.
problems during a study. Although a clinical neurophysiologic assess-
ment rarely provides evidence for a specific
diagnosis, it can provide valuable information
SUMMARY about the severity, progression, and prognosis
of the disease.
The major value and primary application of
clinical neurophysiology is in the assessment
and characterization of neurologic disease.
Selection of appropriate studies for the prob- REFERENCES
lem of an individual patient requires a careful
clinical evaluation to determine possible causes 1. AAEM Quality Assurance Committee. 1999. Litera-
of the patient’s symptoms. The nature of the ture review of the usefulness of nerve conduction stud-
symptoms and the conclusions of the clinical ies and electromyography in the evaluation of patients
with ulnar neuropathy at the elbow. Muscle & Nerve
evaluation are the best guides to appropriate 22:S408–11.
use of clinical neurophysiologic testing. 2. England, J. D., G. S. Gronseth, G. Franklin, et al.
The approach to testing can be assisted 2005. Distal symmetrical polyneuropathy: Definition
by deciding which structures are likely to for clinical research. Muscle & Nerve 31:113–23.
be involved. For example, motor and sensory 3. American Academy of Neurology. 1991. Research cri-
teria for diagnosis of chronic inflammatory demyelinat-
symptoms are best assessed using the dif- ing polyneuropathy (CIDP). Report from an Ad Hoc
ferent methods of motor and sensory NCS. Subcommittee of the American Academy of Neurol-
Electroencephalography, autonomic function ogy CIDP Task Force. Neurology 41:617–18.
testing, and polysomnography provide distinct 4. Bromberg, M. B. 1991. Comparison of electrodiag-
nostic criteria for primary demyelination in chronic
assessment of disturbances of consciousness, polyneuropathy. Muscle & Nerve 14:968–76.
cognition, visceral function, and sleep. The 5. Albers, J. W., and J. J. Kelly. 1989. Acquired
level of the nervous system that is likely to inflammatory demyelinating polyneuropathies: Clini-
be involved by the disease process can also cal and electrodiagnostic features. Muscle & Nerve 12:
guide selection of the neurophysiologic meth- 435–51.
6. Thaisetthawatkul, P., E. L. Logigian, and D. N.
ods that will be most helpful in sorting out Herrmann. 2002. Dispersion of the distal compound
the clinical problem. Disorders of the cere- muscle action potential as a diagnostic criterion for
bral hemisphere are best characterized elec- chronic inflammatory demyelinating polyneuropathy.
trically by electroencephalography, somatosen- Neurology 59:1526–32.
7. Van den Bergh, P. Y. K., and F. Pieret. 2004. Elec-
sory evoked potentials, polysomnography, and trodiagnostic criteria for acute and chronic inflamma-
movement recordings. Lesions in the pos- tory demyelinating polyradiculoneuropathy. Muscle &
terior fossa may benefit from the addition Nerve 29:565–74.
8. Lewis, R. A., A. J. Sumner, and M. E. Shy. 2000. 17. Kimura, I., and D. R. Ayyar. 1984. The hand
Electrophysiological features of inherited demyelinat- neural communication between the ulnar and
ing neuropathies: A reappraisal in the era of molecular median nerves: Electrophysiological detection. Elec-
diagnosis. Muscle & Nerve 23:1472–87. tromyography and Clinical Neurophysiology 24:
9. Olney, R. K. 1999. American Association of Electro- 409–14.
diagnostic Medicine. Consensus criteria for the diag- 18. Kimura, J., M. J. Murphy, and D. J. Varda. 1976.
nosis of partial conduction block. Muscle & Nerve Electrophysiological study of anomalous innervation
22:S225–9. of intrinsic hand muscles. Archives of Neurology
10. Martinez, A. C., M. T. Ferrer, E. D. Tejedor, M. C. 33:842–4.
Perez Conde, B. Anciones, and A. Frank. 1982. 19. Gutmann, L. 1977. Median-ulnar nerve communica-
Diagnostic yield of single fiber electromyography and tions and carpal tunnel syndrome. Journal of Neurol-
other electrophysiological techniques in myasthenia ogy, Neurosurgery, and Psychiatry 40:982–6.
gravis. Electromyography and Clinical Neurophysiol- 20. Gutmann, L., A. Gutierrez, and J. E. Riggs. 1986.
ogy 22:377–93. The contribution of median to ulnar communication
11. Ozdemir, C., and R. R. Young. 1976. The results to in diagnosis of mild carpal tunnel syndrome. Muscle &
be expected from electrical testing in the diagnosis of Nerve 9:319–21.
myasthenia gravis. Annals of the New York Academy of 21. Rowntree, T. 1949. Anomalous innervation of the
Sciences 274:203–22. hand muscles. The Journal of Bone and Joint Surgery
12. Oh, S. J., N. Eslami, T. Nishirhira, et al. 1982. Elec- 31-B:505–10.
trophysiological and clinical correlation in myasthenia 22. Sachs, G. M., E. M. Raynor, and J. M. Shefner. 1995.
gravis. Annals of Neurology 12:348–54. The all ulnar motor hand without forearm anastomo-
13. Oh, S. J., D. E. Kim, R. Kuruoglu, R. J. Bradley, sis. Muscle & Nerve 18:309–13.
and D. Dwyer. 1992. Diagnostic sensitivity of the lab- 23. Dimitru, D., N. E. Walsh, and C. F. Weber.
oratory tests in myasthenia gravis. Muscle & Nerve 1988. Electrophysiologic study of the Riche-Cannieu
15:720–4. anomaly. Electromyography and Clinical Neurophysi-
14. Whitaker, C. H., and K. J. Felice. 2004. Apparent ology 28:27–31.
conduction block in patients with ulnar neuropathy at 24. Seradge, H., and E. Seradge. 1990. Median innervated
the elbow and proximal Martin-Gruber anastomosis. hypothenar muscle: Anomalous branch of median
Muscle & Nerve 30:808–11. nerve in carpal tunnel syndrome. The Journal of Hand
15. Wilbourn, A. J., and E. H. Lambert. 1976. The fore- Surgery 15A:356–9.
arm median-to-ulnar nerve communication: Electro- 25. Gutmann, L. 1993. AAEM minimonograph #2: Impor-
diagnostic aspects. Neurology 26:368. tant anomalous innervations of the extremities. Muscle
16. Sun, S. F., and E. W. Streib. 1983. Martin- & Nerve 16:339–47.
Gruber anastomosis: Electromyographic studies: Part 26. Lambert, E. H. 1969. The accessory deep peroneal
II. Electromyography and Clinical Neurophysiology nerve: A common variation in innervation of extensor
23:271–85. digitorum brevis. Neurology 19:1169–76.
Glossary of Electrophysiologic
Terms∗
A Wave: A compound muscle action potential stimulus is at or above threshold, the action
that follows the M wave, evoked consistently potential generated has a constant size and
from a muscle by submaximal electric stim- configuration. See also compound action
uli and frequently abolished by supramaxi- potential and motor unit action potential.
mal stimuli. Its amplitude is similar to that of Activation: (1) In physiology, a general term
an F wave, but the latency is more constant. for the initiation of a process. (2) The pro-
Usually occurs before the F wave, but may cess of motor unit action potential firing. The
occur afterwards. Thought to be due to extra force of muscle contraction is determined by
discharges in the nerve, ephapses, or axonal the number of motor units and their firing
branching. This term is preferred over axon rate.
reflex, axon wave, or axon response. Com- Activation Procedure: A technique used to
pare with the F wave. detect defects of neuromuscular transmis-
Absolute Refractory Period: See refractory sion during repetitive nerve stimulation test-
period. ing. Most commonly a sustained voluntary
Accommodation: In neuronal physiology, a contraction is performed to elicit facilitation
rise in the threshold transmembrane depo- or postactivation depression. See also tetanic
larization required to initiate a spike, when contraction.
depolarization is slow or a subthreshold Active Electrode: Synonymous with explor-
depolarization is maintained. In the older lit- ing electrode. See recording electrode.
erature, the observation that the final inten- Acute Inflammatory Neuropathy: An acute,
sity of current applied in a slowly rising monophasic polyneuropathy. Characterized
fashion to stimulate a nerve was greater than by a time course of progression to maximum
the intensity of a pulse of current required deficit within 4 weeks of onset of symp-
to stimulate the same nerve. The latter may toms. Most common clinical presentation
largely be an artifact of the nerve sheath and is an ascending sensory-motor neuropathy.
bears little relation to true accommodation Electrodiagnostic studies most commonly
as measured intracellularly. reveal evidence for demyelination, but
Accommodation Curve: See strength–duration axonal degeneration also occurs. Distinguish
curve. from chronic inflammatory demyelinating
Acoustic Myography: The recording and polyradiculoneuropathy (CIDP). See also
analysis of sounds produced by the contract- Guillain–Barré syndrome.
ing muscle. The muscle contraction may be Adaptation: A decline in the frequency of the
produced by stimulation of the nerve supply spike discharge as typically recorded from
to the muscle or by volitional activation of sensory axons in response to a maintained
the muscle. stimulus.
Action Potential (AP): The brief regenera- ADEMG: Abbreviation for automatic decom-
tive electric potential that propagates along position electromyography.
a single axon or muscle fiber membrane. AEP: Abbreviation for auditory-evoked pot-
An all-or-none phenomenon; whenever the ential.
∗ Reproduced by permission of the American Association of Electrodiagnostic Medicine.
839
840 Glossary of Electrophysiologic Terms
Afterdischarge: (1) The continuation of motor nerve fibers away from the muscle and
action potentials in a neuron, axon, or mus- conduction along sensory fibers away from
cle fiber following the termination of an the spinal cord. Contrast with orthodromic.
applied stimulus. (2) The continuation of AP: Abbreviation for action potential.
firing of muscle action potentials after ces- Artifact (also Artefact): A voltage change
sation of voluntary activation, for example in generated by a biologic or nonbiologic
myotonia. source other than the ones of interest. The
Afterpotential: The membrane potential bet- stimulus artifact (or shock artifact) repre-
ween the end of the spike and the time sents cutaneous spread of stimulating cur-
when the membrane potential is restored to rent to the recording electrode and the delay
its resting value. The membrane during this in return to baseline which is dependent on
period may be depolarized or hyperpolar- the ability of filters to respond to high volt-
ized at different times. age. Stimulus artifacts may precede or over-
Akinesia: Lack or marked delay of intended lap the activity of interest. Movement artifact
movement, often observed in patients with refers to a change in the recorded activ-
Parkinson’s disease. Often used synony- ity caused by movement of the recording
mously with bradykinesia. electrodes.
Amplitude: With reference to an action Asterixis: A quick involuntary movement
potential, the maximum voltage difference caused by a brief lapse in tonic muscle acti-
between two points, usually baseline-to-peak vation. It can be appreciated only during
or peak-to-peak. By convention, the ampli- voluntary movement. Is usually irregular, but
tude of potentials which have an initial neg- can be rhythmic and confused with action
ative deflection from the baseline, such as tremor.
the compound muscle action potential and Ataxia: Clumsiness of movement. Specific fea-
the antidromic sensory nerve action poten- tures include dysmetria (incorrect distance
tial are measured from baseline to the most moved) and dysdiadochokinesis (irregularity
negative peak. In contrast, the amplitude of of attempted rhythmic movements). Most
a compound sensory nerve action potential, commonly due to a disorder of the cere-
motor unit potential, fibrillation potential, bellum or proprioceptive sensory system.
positive sharp wave, fasciculation potential, Referred to, respectively, as cerebellar ataxia
and most other action potentials is measured or sensory ataxia.
from the most positive peak to the most Auditory Evoked Potential (AEP): Elec-
negative peak. tric waveforms of biologic origin elicited in
Amplitude Decay: The percent change in the response to sound stimuli. Classified by their
amplitude of the M wave or the compound latency as short-latency brain stem auditory
sensory nerve action potential between evoked potential (BAEP) with a latency of up
two different stimulation points along the to 10 ms, middle latency with a latency of
nerve. Decay = 100 · (amplitudedistal – 10–50 ms, and long latency with a latency of
amplitudeproximal )/amplitudedistal . Useful in the over 50 ms. See brain stem auditory evoked
evaluation of conduction block. Abnormal potential.
decay without increased temporal dispersion Automatic Decomposition EMG (ADEMG):
may indicate a conduction block. Computerized method for extracting indi-
Anodal Block: A local block of nerve conduc- vidual motor unit action potentials from an
tion caused by membrane hyperpolarization interference pattern.
under a stimulating anode. Does not occur in Averager: See signal averager.
routine clinical studies, since it is possible for Averaging: A method for extracting time-
the anode to routinely result in nerve depo- locked potentials from random background
larization if sufficient current intensities are noise by sequentially adding traces and
used. dividing by the total number of traces.
Anode: The positive terminal of an electric Axon Reflex: Use of term discouraged as it is
current source. See stimulating electrode. incorrect. No reflex is thought to be involved.
Antidromic: Propagation of a nerve impulse See preferred term, A wave.
in the direction opposite to physiologic Axon Response: See preferred term, A wave.
conduction; for example, conduction along Axon Wave: See A wave.
Axonal Degeneration: Degeneration of the are flush with the level of the cannula which
segment of a nerve distal to the cell body may serve as a ground.
with preferential distal pathology. Bipolar Stimulating Electrode: See stimu-
Axonotmesis: Nerve injury characterized by lating electrode.
axon and myelin sheath disruption with Bizarre High-Frequency Discharge: See
supporting connective tissue preservation, preferred term, complex repetitive dis-
resulting in axonal degeneration distal to charge.
the injury site. Compare neurapraxia and Bizarre Repetitive Discharge: See pre-
neurotmesis. ferred term, complex repetitive discharge.
Backaveraging: Averaging a signal which Bizarre Repetitive Potential: See preferred
occurs in a time epoch preceding a trig- term, complex repetitive discharge.
gering event. Often used to extract a time- Blink Reflex: See blink responses.
locked EEG signal preceding voluntary or Blink Responses: Compound muscle action
involuntary movement, usually triggered by potentials evoked from orbicularis oculi
the onset of the EMG activity of the muscles as a result of brief electric or
movement. An example is the Bereitschaft- mechanical stimuli applied to the cuta-
spotential. neous area innervated by the supraorbital
Backfiring: Discharge of an antidromically (or less commonly, the infraorbital) branch
activated motor neuron. of the trigeminal nerve. Typically, there is
BAEP: Abbreviation for brain stem auditory an early compound muscle action potential
evoked potential. (R1 wave) ipsilateral to the stimulation site
BAER: Abbreviation for brain stem auditory with a latency of about 10 ms and a bilat-
evoked response. See preferred term, brain eral late compound muscle action poten-
stem auditory evoked potential. tial (R2 wave) with a latency of approxi-
Baseline: (1) The potential recorded from a mately 30 ms. Generally, only the R2 wave
biologic system while the system is at rest. is associated with a visible contraction of the
(2) A flat trace on the recording instrument; muscle. The configuration, amplitude, dura-
an equivalent term, isoelectric line, may be tion, and latency of the two components,
used. along with the sites of recording and stim-
Benign Fasciculation Potential: A firing ulation, should be specified. The R1 and
pattern of fasciculation potentials occurring R2 waves are oligosynaptic and polysnap-
in association with a clinical syndrome of tic brain stem reflexes, respectively. Together
fasciculations in an individual with a non- they are called the blink reflex. The afferent
progressive neuromuscular disorder. Use of arc is provided by the sensory branches of
term discouraged. the trigeminal nerve and the efferent arc is
BER: Abbreviation for brain stem auditory provided by facial nerve motor fibers.
evoked responses. See preferred term, brain Blocking: Term used in single fiber elec-
stem auditory evoked potentials. tromyography to describe dropout of one
Bereitschaftspotential (BP): A component or more components of the potential dur-
of the movement-related cortical potential. ing sequential firings. If more than one
The slowly rising negativity in the EEG pre- component drops out simultaneously it is
ceding voluntary movement. The German described as concomitant blocking. Usually
term means “readiness potential.” Has two seen when jitter values exceed 80–100 μs.
phases called BP1 and BP2 or BP and NS9 A sign of abnormal neuromuscular trans-
(negative slope). See backaveraging. mission, which may be due to primary neu-
Biphasic Action Potential: An action poten- romuscular transmission disorders, such as
tial with one baseline crossing, producing myasthenia gravis and other myasthenic
two phases. syndromes. Also seen as a result of degen-
Biphasic End Plate Activity: See end plate eration and reinnervation in neuropathies or
activity (biphasic). myopathies. Concomitant blocking may be
Bipolar Needle Electrode: Recording elec- generated by a split muscle fiber or failure of
trode that measures voltage between two conduction at an axon branch serving several
insulated wires cemented side-by-side in a muscle fibers.
steel cannula. The bare tips of the electrodes BP: Abbreviation for Bereitschaftspotential.
Brachial Plexus: An anatomical structure Cathode: The negative terminal of an electric

which is formed by the spinal roots from current source. See stimulating electrode.
C5 to T1, traverses the shoulder region, and Center Frequency: The mean or median fre-
culminates in the named peripheral nerves quency of a waveform decomposed by fre-
in the arm. It is composed of roots, trunks, quency analysis. Employed in the study of
divisions, cords, and terminal nerves. muscle fatigue.
Bradykinesia: Slowness of movement, often Central Electromyography: Use of elec-
observed in patients with Parkinson’s dis- trodiagnostic recording techniques to study
ease. Often used synonymously with reflexes and the control of movement by the
akinesia. spinal cord and brain. See electrodiagnosis.
Brain Stem Auditory Evoked Potential Central Motor Conduction: The time taken
(BAEP): Electric waveforms of biologic ori- for conduction of action potentials in the
gin elicited in response to sound stimuli. central nervous system from motor cortex
Normally consists of a sequence of up to to alpha motoneurons in the spinal cord or
seven waves, designated I to VII, which brain stem. Calculated from the latencies
occur during the first 10 ms after the onset of the motor evoked potentials produced by
of the stimulus and have positive polarity at transcranial magnetic stimulation or tran-
the vertex of the head. scranial electrical stimulation, subtracting
Brain Stem Auditory Evoked Response the time for peripheral conduction.
(BAER, BER): See preferred term, brain Chorea: Clinical term used to describe irregu-
stem auditory evoked potentials. lar, random, brief, abrupt, involuntary move-
BSAP: Abbreviation for brief, small, abun- ments of the head or limbs due to a dis-
dant potentials (see BSAPP). Use of term is order of the basal ganglia. Most commonly
discouraged. observed in patients with Huntington’s dis-
BSAPP: Abbreviation for brief, small, abun- ease and Sydenham’s chorea.
dant, polyphasic potentials. Used to describe Chronaxie (also Choronaxy): See strength–
a recruitment pattern of brief duration, duration curve.
small amplitude, overly abundant, polypha- Chronic Inflammatory Demyelinating
sic motor unit action potentials, with respect Poly-radiculoneuropathy (CIDP): A poly-
to the amount of force generated; usually a neuropathy or polyradiculoneuropathy char-
minimal contraction. Use of term discour- acterized by generalized demyelination of
aged. Quantitative measurements of motor the peripheral nervous system. In most cases
unit action potential duration, amplitude, there is also a component of axonal degener-
numbers of phases, and recruitment fre- ation. Some cases are associated with a mon-
quency are preferred. See motor unit action oclonal gammopathy of undetermined sig-
potential. nificance (MGUS). Distinguish from acute
C Reflex: An abnormal reflex response rep- inflammatory neuropathy.
resenting the electrophysiologic correlate of Clinical Electromyography: Term used
sensory evoked myoclonus. The term “C” commonly to describe the scientific methods
was chosen to indicate that the reflex might of recording and analysis of biologic electri-
be mediated in the cerebral cortex. This is cal potentials from human peripheral nerve
sometimes, but not always, true. and muscle. See preferred term, electrodiag-
c/s (also cps): Abbreviation for cycles per nostic medicine.
second. See preferred term, hertz (Hz). CMAP: Abbreviation for compound muscle
Carpal Tunnel Syndrome: A mononeuropa- action potential.
thy affecting the median nerve at the wrist. Coaxial Needle Electrode: See synonym,
As the nerve passes through the carpal tun- concentric needle electrode.
nel, a space bounded dorsally by the bones Collision: When used with reference to nerve
of the wrist, laterally by the forearm flexor conduction studies, the interaction of two
tendons, and volarly by the transverse carpal action potentials propagated toward each
ligament, it is subject to compression by other from opposite directions on the same
any of these structures. Repetitive hand and nerve fiber so that the refractory periods of
wrist movement is thought to contribute to the two potentials prevent propagation past
the compression. each other.
Complex Motor Unit Action Potential: A amplitude, duration, and latency of the neg-
motor unit action potential that is polyphasic ative phase should be noted, along with
or serrated. See preferred terms, polyphasic details of the method of stimulation and
action potential or serrated action potential. recording. Use of specific named potentials
Complex Repetitive Discharge: A type of is recommended, for example, M wave, F
spontaneous activity. Consists of a regu- wave, H wave, T wave, A wave, and R1 or
larly repeating series of complex polyphasic R2 wave (blink responses).
or serrated potentials that begin abruptly Compound Nerve Action Potential (Com-
after needle electrode movement or spon- pound NAP): The summation of nearly
taneously. The potentials have a uniform synchronous nerve fiber action potentials
shape, amplitude, and discharge frequency recorded from a nerve trunk, commonly pro-
ranging from 5 to 100 Hz. The discharge duced by stimulation of the nerve directly
typically terminates abruptly. May be seen or indirectly. Details of the method of stim-
in both myopathic and neurogenic disor- ulation and recording should be specified,
ders, usually chronic. Thought to be due together with the fiber type (sensory, motor,
to ephaptic excitation of adjacent mus- or mixed nerve).
cle fibers in a cyclic fashion. This term Compound Sensory Nerve Action Poten-
is preferred to bizarre high-frequency dis- tial (Compound SNAP): A compound
charge, bizarre repetitive discharge, bizarre nerve action potential recorded from the
repetitive potential, pseudomyotonic dis- afferent fibers of a sensory nerve, a sensory
charge, and synchronized fibrillation. See branch of a mixed nerve or in response to
also ephapse and ephaptic transmission. stimulation of a sensory nerve or a dorsal
Compound Action Potential: A potential nerve root. May also be elicited when an
or waveform resulting from the summa- adequate stimulus is applied synchronously
tion of multiple individual axon or mus- to sensory receptors. The amplitude, latency,
cle fiber action potentials. See compound duration, and configuration should be noted.
mixed nerve action potential, compound Generally, the amplitude is measured as
motor nerve action potential, compound the maximum peak-to-peak voltage when
nerve action potential, compound sensory there is an initial positive deflection or from
nerve action potential, and compound mus- baseline-to-peak when there is an initial neg-
cle action potential. ative deflection. The latency is measured as
Compound Mixed Nerve Action Poten- either the time to the initial deflection or
tial: A compound nerve action potential the negative peak, and the duration as the
recorded from a mixed nerve when an elec- interval from the first deflection of the wave-
tric stimulus is applied to a segment of the form from the baseline to its final return
nerve that contains both afferent and effer- to the baseline. Also referred to by the less
ent fibers. The amplitude, latency, duration, preferred terms sensory response, sensory
and phases should be noted. potential, or SNAP.
Compound Motor Nerve Action Potential Concentric Needle Electrode: Recording
(Compound Motor NAP): A compound electrode that measures an electric poten-
nerve action potential recorded from effer- tial difference between a centrally insulated
ent fibers of a motor nerve or a motor wire and the cannula of the needle through
branch of a mixed nerve. Elicited by stim- which it runs.
ulation of a motor nerve, a motor branch Conditioning Stimulus: See paired stimuli.
of a mixed nerve, or a ventral nerve root. Conduction Block: Failure of an action
The amplitude, latency, duration, and num- potential to propagate past a particular point
ber of phases should be noted. Distinguish in the nervous system whereas conduction is
from compound muscle action potential. possible below the point of the block. Doc-
Compound Muscle Action Potential umented by demonstration of a reduction in
(CMAP): The summation of nearly synchro- the area of a compound muscle action poten-
nous muscle fiber action potentials recorded tial greater than that normally seen with
from a muscle, commonly produced by stim- stimulation at two different points on a nerve
ulation of the nerve supplying the muscle trunk; anatomic variations of nerve path-
either directly or indirectly. Baseline-to-peak ways and technical factors related to nerve
stimulation must be excluded as the cause of temporary muscle shortening seen in some
the reduction in area. myopathies (e.g., muscle phosphorylase
Conduction Distance: The length of nerve deficiency).
or muscle over which conduction is deter- Coupled Discharge: See preferred term,
mined, customarily measured in centimeters satellite potential.
or millimeters. cps (also c/s): Abbreviation for cycles per
Conduction Time: See conduction velocity. second. See preferred term, hertz (Hz).
Conduction Velocity (CV): Speed of propa- Cramp Discharge: Involuntary repetitive fir-
gation of an action potential along a nerve ing of motor unit action potentials at a high
or muscle fiber. The nerve fibers studied frequency (up to 150 Hz) in a large area of a
(motor, sensory, autonomic, or mixed nerve) muscle usually associated with painful mus-
should be specified. For a nerve trunk, the cle contraction. Both discharge frequency
maximum conduction velocity is calculated and number of motor unit action potentials
from the latency of the evoked potential activated increase gradually during develop-
(muscle or nerve) at maximal or supramax- ment and subside gradually with cessation.
imal intensity of stimulation at two differ- See muscle cramp.
ent points. The distance between the two Cross Talk: (1) A general term for abnormal
points (conduction distance) is divided by communication between excitable mem-
the difference between the corresponding branes. See ephapse and ephaptic transmis-
latencies (conduction time). The calculated sion. (2) Term used in kinesiologic EMG for
result is the conduction velocity of the fastest signals picked up from adjacent muscles.
fibers and is usually expressed as meters per Crossed Leg Palsy: Synonym for peroneal
second (m/s). As commonly used, refers to neuropathy at the knee.
the maximum conduction velocity. By spe- Cubital Tunnel Syndrome: A mononeuropa-
cialized techniques, the conduction velocity thy involving the ulnar nerve in the region
of other fibers can also be determined and of the elbow. An entrapment neuropathy
should be specified, for example, minimum caused by compression of the nerve as it
conduction velocity. passes through the aponeurosis (the cubital
Congenital Myasthenia: A heterogeneous tunnel) of the two heads of the flexor carpi
group of genetic disorders of the neuromus- ulnaris approximately 1.5–3.5 cm distal to
cular junction manifest by muscle weakness the medial epicondyle of the elbow. The
and fatigue. mechanism of entrapment is presumably
Contraction: A voluntary or involuntary narrowing of the cubital tunnel during elbow
reversible muscle shortening that may or flexion. See also tardy ulnar palsy and ulnar
may not be accompanied by action potentials neuropathy at the elbow.
from muscle. Contrast the term contracture. Cutaneous Reflex: A reflex produced by cuta-
Contraction Fasciculation: Clinical term for neous stimulation. There are several phases
visible twitching of a muscle with weak vol- to cutaneous reflexes, and, if the muscle has
untary or postural contraction which has the a background contraction, the phases can be
appearance of a fasciculation. More likely to seen to be inhibitory as well as excitatory.
occur in neuromuscular disorders in which CV: Abbreviation for conduction velocity.
the motor unit territory is enlarged and the Cycles Per Second (c/s, cps): Unit of fre-
tissue covering the muscle is thin, but may quency. See preferred term hertz (Hz).
also be observed in normal individuals. Decomposition EMG: Synonym for auto-
Contracture: (1) Fixed resistance to stretch matic decomposition EMG.
of a shortened muscle due to fibrous con- Decremental Response: See preferred term,
nective tissue changes and loss of sarcom- decrementing response.
eres in the muscle. Limited movement of Decrementing Response: A reproducible
a joint may be due to muscle contracture decline in the amplitude and/or area of the
or to fibrous connective tissue changes in M wave of successive responses to repeti-
the joint. Contrast with contraction, which tive nerve stimulation. The rate of stimula-
is a rapidly reversible painless shortening tion and the total number of stimuli should
of the muscle. (2) The prolonged, painful, be specified. Decrementing responses with
electrically silent, and involuntary state of disorders of neuromuscular transmission
are most reliably seen with slow rates potentials only. Synonymous with action
(2–5 Hz) of nerve stimulation. A decrement- potential.
ing response with repetitive nerve stimula- Discharge Frequency: The rate at which
tion commonly occurs in disorders of neu- a potential discharges repetitively. When
romuscular transmission, but can also be potentials occur in groups, the rate of recur-
seen in some neuropathies, myopathies, and rence of the group and rate of repetition
motor neuron disease. An artifact resem- of the individual components in the groups
bling a decrementing response can result should be specified. See also firing rate.
from movement of the stimulating or record- Discrete Activity: See interference pattern.
ing electrodes during repetitive nerve stimu- Distal Latency: The interval between the
lation (see pseudodecrement). Contrast with delivery of a stimulus to the most distal point
incrementing response. of stimulation on a nerve and the onset of a
Delay: (1) The time between the beginning response. A measure of the conduction prop-
of the horizontal sweep of the oscilloscope erties of the distal-most portion of motor
and the onset of an applied stimulus. (2) A or sensory nerves. See motor latency and
synonym for an information storage device sensory latency.
(delay line) used to display events occurring Double Discharge: Two sequential firings of
before a trigger signal. a motor unit action potential of the same
Delay Line: An information storage device form and nearly the same amplitude, occur-
used to display events which occur before ring consistently in the same relationship to
a trigger signal. A method for displaying a one another at intervals of 2–20 ms. See also
waveform at the same point on a sweep from multiple discharge and triple discharge.
a free-running electromyogram. Doublet: Synonym for the preferred term,
Demyelination: Disease process affecting the double discharge.
myelin sheath of central or peripheral nerve DSEP: Abbreviation for dermatomal somato-
fibers, manifested by conduction velocity sensory evoked potential.
slowing, conduction block, or both. Duration: The time during which something
Denervation Potential: Sometimes used as a exists or acts. (1) The interval from the
synonym for fibrillation potential. Use of this beginning of the first deflection from the
term is discouraged, since fibrillation poten- baseline to its final return to the baseline of
tials can occur in the absence of denervation. an action potential or waveform, unless oth-
See preferred term, fibrillation potential. erwise specified. If only part of the waveform
Depolarization: A change in the existing is measured, the points of the measurement
membrane potential to a less negative value. should be specified. For example, the dura-
Depolarizing an excitable cell from its rest- tion of the M wave may be measured as the
ing level to threshold typically generated an negative phase duration and refers to the
action potential. interval from the deflection of the first neg-
Depolarization Block: Failure of an excitable ative phase from the baseline to its return to
cell to respond to a stimulus due to preexist- the baseline. (2) The interval of the applied
ing depolarization of the cell membrane. current or voltage of a single electric stimu-
Depth Electrodes: Electrodes which are lus. (3) The interval from the beginning to
inserted into the substance of the brain for the end of a series of recurring stimuli or
electrophysiological recording. Most often action potentials.
inserted using stereotactic techniques. Dynamic EMG: See kinesiologic EMG.
Dermatomal Somatosensory Evoked Pote- Dyskinesia: An abnormal involuntary move-
ntial (DESP): Scalp-recorded waveforms ment of a choreic or dystonic type. The term
generated from repeated stimulation of a is nonspecific and is often used in association
specific dermatome. Different from typical with a modifier that describes its etiology,
somatosensory evoked potentials which are for example, tardive dyskinesia or l-DOPA
recorded in response to stimulation of a dyskinesia.
named peripheral nerve. Dystonia: A disorder characterized by invol-
Discharge: The firing of one or more excitable untary movements caused by sustained mus-
elements (neurons, axons, or muscle fibers); cle contraction, producing prolonged move-
as conventionally used, refers to all-or-none ments or abnormal postures.
E-1: Synonymous with input terminal 1. See electrical potentials from the central, periph-
recording electrode. eral, and autonomic nervous systems and
E-2: Synonymous with input terminal 2. See muscles. See also clinical electromyog-
recording electrode. raphy, electromyography, electroneurogra-
E:I Ratio: In autonomic testing, the ratio of phy, electroneuromyography, evoked
the longest electrocardiographic R–R inter- potentials, electrodiagnostic medicine, elec-
val during expiration to the shortest during trodiagnostic medicine consultation, and
inspiration. Primarily a measure of parasym- electrodiagnostic medicine consultant.
pathetic control of heart rate. Electrodiagnostic Medicine: A specific area
Early Recruitment: A recruitment pattern of medical practice in which a physician inte-
which occurs in association with a reduction grates information obtained from the clinical
in the number of muscle fibers per motor history, observations from physical examina-
unit or when the force generated by the tion, and scientific data acquired by record-
fibers is reduced. At low levels of muscle ing electrical potentials from the nervous
contraction more motor unit action poten- system and muscle to diagnose, or diagnose
tials are recorded than expected, and a full and treat, diseases of the central, periph-
interference pattern may be recorded at rela- eral, and autonomic nervous systems, neu-
tively low levels of muscle contraction. Most romuscular junctions, and muscle. See also
often encountered in myopathy. electrodiagnosis, electrodiagnostic medicine
Earth Electrode: Synonymous with ground consultation, and electrodiagnostic medicine
electrode. consultant.
EDX: Abbreviation for electrodiagnosis. Can Electrodiagnostic Medicine Consultant:
also be used for electrodiagnostic and elec- A physician specially trained to obtain a
trodiagnostic medicine. medical history, perform a physical exam-
Electric Inactivity: See preferred term, elec- ination, and to record and analyze data
tric silence. acquired by recording electrical potentials
Electric Silence: The absence of measurable from the nervous system and muscle to
electric activity due to biologic or nonbio- diagnose and/or treat diseases of the cen-
logic sources. The sensitivity and signal-to- tral, peripheral, and autonomic nervous sys-
noise level of the recording system should be tems, neuromuscular junction, and muscle.
specified. See also electrodiagnosis, electrodiagnos-
Electrocorticography: Electrophysiologic tic medicine, and electrodiagnostic medicine
recording directly from the surface of the consultation.
brain. In the intraoperative setting, record- Electrodiagnostic Medicine Consultation:
ings are made of ongoing spontaneous The medical evaluation in which a spe-
electroencephalogram activity, or potentials cially trained physician (electrodiagnostic
evoked by stimulation of peripheral sensory medicine consultant) obtains a medical his-
pathways. tory, performs a physical examination, and
Electrode: A conducting device used to integrates scientific data acquired by record-
record an electric potential (recording elec- ing electrical potentials from the nervous
trode) or to deliver an electric current system and muscle to diagnose and/or treat
(stimulating electrode). In addition to the diseases of the central, peripheral, and
ground electrode used in clinical record- autonomic nervous systems, neuromuscular
ings, two electrodes are always required junction, and muscle. See also electrodiag-
either to record an electric potential or to nosis, electrodiagnostic medicine, and elec-
deliver a stimulus. See ground electrode, trodiagnostic medicine consultant.
recording electrode, and stimulating elec- Electromyogram: The record obtained by
trode. Also see specific needle electrode con- electromyography.
figurations: monopolar, unipolar, concentric, Electromyograph: Equipment used to acti-
bifilar recording, bipolar stimulating, mul- vate, record, process, and display electri-
tilead, single fiber, and macro-EMG needle cal potentials for the purpose of evaluating
electrodes. the function of the central, peripheral, and
Electrodiagnosis (EDX): The scientific autonomic nervous systems, neuromuscular
methods of recording and analyzing biologic junction, and muscles.
Electromyographer: See preferred term, the postsynaptic membrane of a muscle fiber

electrodiagnostic medicine consultant. by release of acetylcholine from the presy-
Electromyography (EMG): Strictly defined, naptic axon terminal in response to an action
the recording and study of insertion, sponta- potential.
neous, and voluntary activity of muscle with End Plate Spike: See end plate activity
a recording electrode (either a needle elec- (biphasic).
trode for invasive EMG or a surface electrode End Plate Zone: The region in a muscle
for kinesiologic studies). The term is also where neuromuscular junctions are concen-
commonly used to refer to an electrodiag- trated.
nostic medicine consultation, but its use in ENG: Abbreviation for electroneurography.
this context is discouraged. ENMG: Abbreviation for electroneuromyo-
Electroneurography (ENG): The recording graphy.
and study of the action potentials of periph- Entrapment Neuropathy: A mononeuropa-
eral nerve. Synonymous with nerve conduc- thy caused by compression of nerve as it
tion studies. passes through an area of anatomical nar-
Electroneuromyography (ENMG): The rowing.
combined studies of electromyography and Ephapse: A point of abnormal communica-
electroneurography. Synonymous with clin- tion where an action potential in one muscle
ical electromyography. See preferred term fiber or axon can cause depolarization of an
electrodiagnostic medicine consultation. adjacent muscle fiber or axon to generate an
EMG: Abbreviation for electromyography. action potential.
End Plate Activity: Spontaneous electric Ephaptic Transmission: The generation of a
activity recorded with a needle electrode nerve fiber action potential from one muscle
close to muscle end plates. These potentials fiber or axon to another through an ephapse.
may have several different morphologies. Postulated to be the basis for complex repet-
1. Monophasic: Low-amplitude (10–20 μV), itive discharges, myokymic discharges, and
short-duration (0.5–1.0 ms), negative poten- hemifacial spasm.
tials occurring in a dense, steady pattern, the EPSP: Abbreviation for excitatory postsynap-
exact frequency of which cannot be defined. tic potential.
These nonpropagated potentials are proba- Erb’s Point: The site at the anterolateral base
bly miniature end plate potentials recorded of the neck where percutaneous nerve stim-
extracellularly. Referred to as end plate ulation activates the axons comprising the
noise or sea-shell sound (sea shell roar or upper trunk of the brachial plexus.
noise). Erb’s Point Stimulation: Percutaneous supr-
2. Biphasic: Moderate-amplitude (100–300 μV), aclavicular nerve stimulation during which
short-duration (2–4 ms), initially negative the upper trunk of the brachial plexus is acti-
spike potentials occurring irregularly in vated. See the more general and preferred
short bursts with a high frequency (50– term, supraclavicular nerve stimulation.
100 Hz). These propagated potentials are Evoked Potential: Electric waveform elicited
generated by muscle fibers excited by activ- by and temporally related to a stimulus, most
ity in nerve terminals. These potentials have commonly an electric stimulus delivered
been referred to as biphasic spike poten- to a sensory receptor or nerve, or applied
tials, end plate spikes, and, incorrectly, nerve directly to a discrete area of the brain,
potentials. May also have a biphasic initially spinal cord, or muscle. See auditory evoked
positive morphology. potential, brain stem auditory evoked poten-
3. Triphasic: Similar to biphasic potentials, tial, spinal evoked potential, somatosensory
but the waveforms have three phases with evoked potential, visual evoked potential,
an initial positive deflection. Fire in an compound muscle action potential, and com-
irregular fashion; contrast with fibrillation pound sensory nerve action potential.
potential. Evoked Potential Studies: Recording and
End Plate Noise: See end plate activity analysis of electric waveforms of biologic
(monophasic). origin elicited in response to electrical,
End Plate Potential (EPP): The graded non- magnetic, or physiological stimuli. Stimuli
propagated membrane potential induced in are applied to specific motor or sensory
receptors, and the resulting waveforms are reproducible increase in the amplitude and
recorded along their anatomic pathways in area of successive M waves during repetitive
the peripheral and central nervous system. nerve stimulation. Postactivation or postte-
A single motor or sensory modality is typ- tanic facilitation—Nerve stimulation studies
ically tested in a study, and the modality performed within a few seconds after a brief
studied is used to define the type of study period (2–60 seconds) of nerve stimulation
performed. See auditory evoked potentials, producing tetanus or after a strong voluntary
brain stem auditory evoked potentials, visual contraction may show changes in the config-
evoked potentials, and somatosensory evoked uration of the M wave(s) compared to the
potentials. results of identical studies of the rested mus-
Evoked Response: Tautology. Use of term cle as follows: (a) repair of the decrement—
discouraged. See preferred term, evoked a diminution of the decrementing response
potential. with slow rates (2–5 Hz) of repetitive nerve
Excitability: Capacity to be activated by or stimulation; (b) increment after exercise—
react to a stimulus. an increase in the amplitude and area of
Excitatory Postsynaptic Potential (EPSP): the M wave elicited by a single supramaxi-
A local, graded depolarization of a neuron mal stimulus. Distinguish from pseudofacil-
in response to activation by a nerve termi- itation, which occurs in normal individuals
nal. Contrast with inhibitory postsynaptic in response to repetitive nerve stimulation
potential. at high rates (20–50 Hz) or after strong
Exploring Electrode: Synonymous with act- volitional contraction. It probably reflects a
ive electrode. See recording electrode. reduction in the temporal dispersion of the
F Reflex: An incorrect term for F wave. summation of a constant number of mus-
F Response: Synonymous with F wave. See cle fiber action potentials and is character-
preferred term, F wave. ized by an increase in the amplitude of the
F Wave: An action potential evoked intermit- successive M waves with a corresponding
tently from a muscle by a supramaximal elec- decrease in their duration. There is no net
tric stimulus to the nerve due to antidromic change in the area of the negative phase of
activation of motor neurons. When com- successive M waves. (2) An increase in the
pared with the maximal amplitude of the amplitude of the motor evoked potential as a
M wave, it is smaller (1%–5% of the M wave) result of background muscle activation.
and has a variable configuration. Its latency Far-Field: A region of electrical potential
is longer than the M wave and is vari- where the isopotential voltage lines associ-
able. It can be evoked in many muscles of ated with a current source change slowly
the upper and lower extremities, and the over a short distance. Some use the term
latency is longer with more distal sites of far-field potential to designate a potential
stimulation. Named “F” wave by Magladery that does not change in latency, amplitude,
and McDougal in 1950, because it was first or polarity over infinite distances; alterna-
recorded from foot muscles. Compare with tive designations include “boundary poten-
the H wave and the A wave. One of the late tial” and “junctional potential.” The terms
responses. near-field and far-field are arbitrary desig-
Facial Neuropathy: Clinical diagnosis of nations as there are no agreed-upon criteria
facial weakness or paralysis due to pathology defining where the near-field ends and the
affecting the seventh cranial nerve (facial far-field begins. Compare with near-field.
nerve). Bell’s palsy refers to a facial neu- Fasciculation: The random, spontaneous
ropathy due to inflammation of the facial twitching of a group of muscle fibers belong-
nerve. ing to a single motor unit. The twitch may
Facilitation: An increase in an electrically produce movement of the overlying skin
measured response following identical stim- (if in limb or trunk muscles) or mucous
uli. Occurs in a variety of circumstances: membrane (if in the tongue). If the motor
(1) Improvement of neuromuscular trans- unit is sufficiently large, an associated joint
mission resulting in activation of previously movement may be observed. The electric
inactive muscle fibers. May be identified activity associated with the twitch is termed
in several ways: Incrementing response—a a fasciculation potential. See also myokymia.
Historically, the term fibrillation was used an associated high-pitched regular sound
incorrectly to describe fine twitching of described as “rain on a tin roof.” In addi-
muscle fibers visible through the skin or tion to this classic form, positive sharp waves
mucous membranes. This usage is no longer may also be recorded from fibrillating mus-
accepted. cle fibers when the potential arises from
Fasciculation Potential: The electric activ- an area immediately adjacent to the needle
ity associated with a fasciculation which electrode.
has the configuration of a motor unit acti- Firing Pattern: Qualitative and quantitative
vation potential but which occurs sponta- descriptions of the sequence of discharge of
neously. Most commonly occur sporadically electric waveforms recorded from muscle or
and are termed “single fasciculation poten- nerve.
tials.” Occasionally the potentials occur as Firing Rate: Frequency of repetition of a
a grouped discharge and are termed a potential. The relationship of the frequency
“brief repetitive discharge.” The repetitive to the occurrence of other potentials and
firing of adjacent fasciculation potentials, the force of muscle contraction may be
when numerous, may produce an undulating described. See also discharge frequency.
movement of muscle (see myokymia). Use Flexor Reflex: A reflex produced by a noxious
of the terms benign fasciculation and malig- cutaneous stimulus, or a train of electrical
nant fasciculation is discouraged. Instead, stimuli, that activates the flexor muscles of
the configuration of the potentials, peak-to- a limb and thus acts to withdraw it from the
peak amplitude, duration, number of phases, stimulus. In humans, it is well characterized
stability of configuration, and frequency of only in the lower extremity.
occurrence, should be specified. Frequency: Number of complete cycles of a
Fatigue: A state of depressed responsiveness repetitive waveform in 1 second. Measured
resulting from activity. Muscle fatigue is in hertz (Hz) or cycles per second (cps or c/s).
a reduction in contraction force following Frequency Analysis: Determination of the
repeated voluntary contraction or electric range of frequencies composing a wave-
stimulation. form, with a measurement of the absolute
Fiber Density: (1) Anatomically, a measure of or relative amplitude of each component
the number of muscle or nerve fibers per frequency.
unit area. (2) In single fiber electromyogra- Full Interference Pattern: See interference
phy, the mean number of muscle fiber action pattern.
potentials fulfilling amplitude and rise time Full Wave Rectified EMG: The absolute
criteria belonging to one motor unit within value of a raw EMG signal. Involves invert-
the recording area of a single fiber needle ing all the waveforms below the isopoten-
electrode encountered during a systematic tial line and displaying them with opposite
search in a weakly, voluntarily contracting polarity above the line. A technique used to
muscle. See also single fiber electromyogra- analyze kinesiologic EMG signals.
phy and single fiber needle electrode. Functional Refractory Period: See refrac-
Fibrillation: The spontaneous contractions of tory period.
individual muscle fibers which are not vis- G1, G2: Abbreviation for grid 1 and grid 2.
ible through the skin. This term has been Generator: In volume conduction theory, the
used loosely in electromyography for the source of electrical activity, such as an action
preferred term, fibrillation potential. potential. See far-field and near-field.
Fibrillation Potential: The action potential “Giant” Motor Unit Action Potential: Use
of a single muscle fiber occurring spon- of term discouraged. Refers to a motor unit
taneously or after movement of a needle action potential with a peak-to-peak ampli-
electrode. Usually fires at a constant rate. tude and duration much greater than the
Consists of biphasic or triphasic spikes of range found in corresponding muscles in
short duration (usually less than 5 ms) with normal subjects of similar age. Quantitative
an initial positive phase and a peak-to-peak measurements of amplitude and duration
amplitude of less than 1 mV. May also have are preferable.
a biphasic, initially negative phase when Giant Somatosensory Evoked Potential:
recorded at the site of initiation. It has Enlarged somatosensory evoked potentials
seen as a characteristic of cortical reflex Habituation: Decrease in size of a reflex

myoclonus and reflecting cortical hyperex- motor response to an afferent stimulus when
citability. the latter is repeated, especially at regular
Grid 1: Synonymous with G1, input termi- and recurring short intervals.
nal 1 (E-1), or active or exploring electrode. Hemifacial Spasm: Clinical condition char-
Use of the term Grid 1 is discouraged. See acterized by frequent, repetitive, unilateral,
recording electrode. involuntary contractions of the facial mus-
Grid 2: Synonymous with G2, input terminal cles. Electrodiagnostic studies demonstrate
2 (E-2), or reference electrode. Use of the brief discharges of groups of motor unit
term Grid 2 is discouraged. See recording action potentials occurring simultaneously
electrode. in several facial muscles. Occasionally high-
Ground Electrode: A connection from the frequency discharges occur.
patient to earth. Used as a common return Hertz (Hz): Unit of frequency. Synonymous
for an electric circuit and as an arbitrary zero with cycles per second.
potential reference point. Hoffmann Reflex: See H wave.
Grouped Discharge: Term used historically Hyperekplexia: Clinical condition character-
to describe three phenomena: (1) irregu- ized by exaggerated startle reflexes. Startle
lar, voluntary grouping of motor unit action reflexes can be exaggerated by being more
potentials as seen in a tremulous mus- extreme than expected (larger amplitude
cular contraction, (2) involuntary group- or more widespread) or by lack of normal
ing of motor unit action potentials as habituation to repeated similar stimuli. Can
seen in myokymia, (3) general term to be either genetic or acquired.
describe repeated firing of motor unit action Hyperpolarization: A change in the exist-
potentials. See preferred term, repetitive ing membrane potential to a more negative
discharge. value.
Guillain–Barré Syndrome: Eponym for Hypertonia: See tone.
acute inflammatory neuropathy. Also referred Hypotonia: See tone.
to as Landry–Guillain–Barré syndrome or Hz: Abbreviation for hertz.
Landry–Guillain–Barré–Strohl syndrome. Impulse Blocking: See blocking.
H Reflex: Abbreviation for Hoffmann reflex. Inching: A nerve conduction study technique
See H wave. consisting of applying stimuli at multiple
H Response: See preferred term, H wave. short distance increments along the course
H Wave: A compound muscle action poten- of a nerve. This technique is used to localize
tial with a consistent latency recorded from an area of focal slowing or conduction block.
muscles after stimulation of the nerve. Reg- Incomplete Activation: Motor unit action
ularly found in adults only in a limited potentials firing, on requested maximal
group of physiologic extensors, particularly effort, in decreased numbers at their nor-
the calf muscles. Compared to the M wave mal physiological rates, within the basal
of the same muscle, has a longer latency firing range of 5–10 Hz. Causes include
and thus is one of the late responses (see upper motor neuron syndrome, pain on mus-
A and F wave). Most reliably elicited with cle contraction, hysteria/conversion reac-
a stimulus of long duration (500–1000 μs). tion, and malingering. Contrast with reduced
A stimulus intensity sufficient to elicit a max- recruitment.
imal amplitude M wave reduces or abol- Increased Insertion Activity: See insertion
ishes the H wave. Thought to be due to activity.
a spinal reflex, with electric stimulation of Increment After Exercise: See facilitation.
afferent fibers in the mixed nerve and activa- Incremental Response: See preferred term,
tion of motor neurons to the muscle mainly incrementing response.
through a monosynaptic connection in the Incrementing Response: A reproducible
spinal cord. The latency is longer with more increase in amplitude and/or area of suc-
distal sites of stimulation. The reflex and cessive M waves to repetitive nerve stim-
wave are named in honor of Hoffmann’s ulation. The rate of stimulation and the
description (1918). Compare the F wave and number of stimuli should be specified. Com-
A wave. monly seen in two situations. First, in normal
subjects the configuration of the M wave electrode, E-2, or less preferred term, grid 2.
may change in response to repetitive nerve See recording electrode.
stimulation so that the amplitude progres- Insertion Activity: Electric activity caused by
sively increases as the duration decreases, insertion or movement of a needle electrode
leaving the area of the M wave unchanged. within a muscle. The amount of the activ-
This phenomenon is termed pseudofacilita- ity may be described as normal, reduced,
tion. Second, in neuromuscular transmission or increased (prolonged), with a description
disorders, the configuration of the M wave of the waveform and repetition rate. See
may change with repetitive nerve stimula- also fibrillation potential and positive sharp
tion so that the amplitude and the area wave.
of the M wave progressively increase. This Integrated EMG: Mathematical integration
phenomenon is termed facilitation. Contrast of the full wave rectified EMG signal.
with decrementing response. Reflects the cumulative EMG activity of a
Indifferent Electrode: Synonymous with ref- muscle over time. See also linear envelope
erence electrode. Use of term discouraged. EMG.
See recording electrode. Interdischarge Interval: Time between con-
Infraclavicular Plexus: Segments of the secutive discharges of the same potential.
brachial plexus inferior to the divisions; Measurements should be made between
includes the three cords and the terminal the corresponding points on each
peripheral nerves. This clinically descriptive waveform.
term is based on the fact that the clavicle Interference: Unwanted electric activity reco-
overlies the divisions of the brachial plexus rded from the surrounding environment.
when the arm is in the anatomic position Interference Pattern: Electric activity recor-
next to the body. ded from a muscle with a needle electrode
Inhibitory Postsynaptic Potential (IPSP): during maximal voluntary effort. A full inter-
A local graded hyperpolarization of a neu- ference pattern implies that no individual
ron in response to activation at a synapse motor unit action potentials can be clearly
by a nerve terminal. Contrast with excitatory identified. A reduced interference pattern
postsynaptic potential. (intermediate pattern) is one in which some
Injury Potential: (1) The potential difference of the individual motor unit action potentials
between a normal region of the surface of may be identified while others are not due
a nerve or muscle and a membrane region to superimposition of waveforms. The term
that has been injured; also called a “demar- discrete activity is used to describe the elec-
cation,” or “killed end” potential. Approx- tric activity recorded when each of several
imates the potential across the membrane different motor unit action potentials can be
because the injured surface has nearly the identified in an ongoing recording due to
same potential as the interior of the cell. limited superimposition of waveforms. The
(2) In electrodiagnostic medicine, the term term single unit pattern is used to describe
is also used to refer to the electrical activ- a single motor unit action potential, firing
ity associated with needle electrode insertion at a rapid rate (should be specified) dur-
into muscle. See preferred terms fibrilla- ing maximum voluntary effort. The force
tion potential, insertion activity, and positive of contraction associated with the interfer-
sharp wave. ence pattern should be specified. See also
Input Terminal 1: The input terminal of early recruitment, recruitment pattern, and
a differential amplifier at which negativity, reduced recruitment pattern.
relative to the other input terminal, pro- Interference Pattern Analysis: Quantitative
duces an upward deflection. Synonymous analysis of the interference pattern. This can
with active or exploring electrode, E-1, or be done either in the frequency domain
less preferred term, grid 1. See recording using fast Fourier transformation (FFT) or
electrode. in the time domain. Can be done using a
Input Terminal 2: The input of a differen- fixed load (e.g., 2 kg) at a given proportional
tial amplifier at which negativity, relative to strength (e.g., 30% of maximum) or at ran-
the other input terminal, produces a down- dom strengths. The following are measured
ward deflection. Synonymous with reference in the time domain: (1) the number of turns
per second and (2) the amplitude, defined as two muscle fiber action potentials belonging
the mean amplitude between peaks. to the same motor unit. Usually expressed
Intermediate Interference Pattern: See quantitatively as the mean value of the dif-
interference pattern. ference between the interpotential intervals
International 10–20 System: A system of of successive discharges (the mean consecu-
electrode placement on the scalp in which tive difference, MCD). Under certain condi-
electrodes are placed either 10% or 20% tions, it is expressed as the mean value of the
of the total distance on a line on the skull difference between interpotential intervals
between the nasion and inion in the sagit- arranged in the order of decreasing interdis-
tal plane and between the right and left charge intervals (the mean sorted difference,
preauricular points in the coronal plane. MSD). See single fiber electromyography.
Interpeak Interval: Difference between the Jolly Test: A technique named after Friedrich
peak latencies of two components of a wave- Jolly, who applied an electric current to
form. excite a motor nerve repetitively while
Interpotential Interval: Time between two recording the force of muscle contraction.
different potentials. Measurement should be Use of the term is discouraged. Inappro-
made between the corresponding parts of priately used to describe the technique of
each waveform. repetitive nerve stimulation.
Intraoperative Monitoring: The use of elec- Kinematics: Technique for description of
trophysiological stimulating and recording body movement without regard to the
techniques in an operating room setting. The underlying forces. See kinesiologic EMG.
term is usually applied to techniques which Kinesiologic EMG: The muscle electrical
are used to detect injury to nervous tis- activity recorded during movement. Gives
sue during surgery or to guide the surgical information about the timing of muscle
procedure. activity and its relative intensity. Either
Involuntary Activity: Motor unit action surface electrodes or intramuscular fine
potentials that are not under volitional con- wire electrodes are used. Synonymous with
trol. The condition under which they occur dynamic EMG.
should be described, for example, sponta- Kinesiology: The study of movement. See
neous or reflex potentials. If elicited by a kinesiologic EMG.
stimulus, its nature should be described. Kinetics: The internal and external forces
Contrast with spontaneous activity. affecting the moving body. See kinesiologic
IPSP: Abbreviation for inhibitory postsynaptic EMG.
potential. Late Component (of a Motor Unit Action
Irregular Potential: See preferred term, ser- Potential): See preferred term, satellite
rated action potential. potential.
Isoelectric Line: In electrophysiologic record- Late Response: A general term used to
ing, the display of zero potential difference describe an evoked potential in motor nerve
between the two input terminals of the conduction studies having a longer latency
recording apparatus. See baseline. than the M wave. Examples include A wave,
Iterative Discharge: See preferred term, F wave, and H wave.
repetitive discharge. Latency: Interval between a stimulus and a
Jiggle: Shape variability of motor unit action response. The onset latency is the inter-
potentials recorded with a conventional val between the onset of a stimulus and
EMG needle electrode. A small amount the onset of the evoked potential. The peak
occurs normally. In conditions of disturbed latency is the interval between the onset of a
neuromuscular transmission, including early stimulus and a specified peak of the evoked
reinnervation and myasthenic disorders, the potential.
variability can be sufficiently large to be eas- Latency of Activation: The time required for
ily detectable by eye. Quantitative methods an electric stimulus to depolarize a nerve
for estimating this variability are not yet fiber (or bundle of fibers as in a nerve trunk)
widely available. beyond threshold and to initiate an action
Jitter: The variability of consecutive dis- potential in the fiber(s). This time is usually
charges of the interpotential interval between of the order of 0.1 ms or less. An equivalent
term, now rarely used, is the “utilization Macro MUAP: Abbreviation for macro motor
time.” unit action potential.
Latent Period: See preferred term, latency. Macroelectromyography (Macro-EMG):
Linear Envelope EMG: Moving average General term referring to the technique and
of the full wave rectified EMG. Obtained conditions that approximate recording of all
by low-pass filtering the full wave rectified muscle fiber action potentials arising from the
EMG. See also integrated EMG. same motor unit. See macro motor unit action
Linked Potential: See preferred term, satel- potential.
lite potential. Macro-EMG: Abbreviation for macroelec-
Lipoatrophy: Pathologic loss of subcutaneous tromyography.
fat and connective tissues overlying mus- Macro-EMG Needle Electrode: A modi-
cle which mimics the clinical appearance of fied single fiber electromyography electrode
atrophy of the underlying muscle. insulated to within 15 mm from the tip and
Long-Latency Reflex: A reflex with many with a small recording surface (25 μm in
synapses (polysynaptic) or a long pathway diameter) 7.5 mm from the tip.
(long-loop) so that the time to its occur- Malignant Fasciculation: Used to describe
rence is greater than the time of occurrence large, polyphasic fasciculation potentials fir-
of short-latency reflexes. See also long-loop ing at a slow rate. This pattern has been seen
reflex. in progressive motor neuron disease, but the
Long-Loop Reflex: A reflex thought to have a relationship is not exclusive. Use of this term
circuit that extends above the spinal segment is discouraged. See fasciculation potential.
of the sensory input and motor output. May Maximal Stimulus: See stimulus.
involve the cerebral cortex. Should be differ- Maximum Conduction Velocity: See con-
entiated from reflexes arising from stimula- duction velocity.
tion and recording within a single segment MCD: Abbreviation for mean consecutive dif-
or adjacent spinal segments (i.e, a segmental ference. See jitter.
reflex). See also long-latency reflex. Mean Consecutive Difference (MCD): See
M Response: See preferred term, M wave. jitter.
M Wave: A compound muscle action poten- Mean Sorted Difference (MSD): See jitter.
tial evoked from a muscle by an electric Membrane Instability: Tendency of a cell
stimulus to its motor nerve. By convention, membrane to depolarize spontaneously in
the M wave elicited by a supramaximal response to mechanical irritation or follow-
stimulus is used for motor nerve conduc- ing voluntary activation. May be used to
tion studies. Ideally, the recording electrodes describe the occurrence of spontaneous sin-
should be placed so that the initial deflec- gle muscle fiber action potentials such as
tion of the evoked potential from the base- fibrillation potentials during needle electrode
line is negative. Common measurements examination.
include latency, amplitude, and duration. MEP: Abbreviation for motor evoked poten-
Also referred to as the motor response. tial.
Normally, the configuration is biphasic and MEPP: Abbreviation for miniature end plate
stable with repeated stimuli at slow rates potential.
(1–5 Hz). See repetitive nerve stimulation. Microneurography: The technique of record-
Macro Motor Unit Action Potential: The ing peripheral nerve action potentials in
average electric activity of that part of an humans by means of intraneural electrodes.
anatomic motor unit that is within the Miniature End Plate Potential (MEPP):
recording range of a macro-EMG elec- The postsynaptic muscle fiber potentials
trode. Characterized by consistent appear- produced through the spontaneous release
ance when the small recording surface of of individual acetylcholine quanta from the
the macro-EMG electrode is positioned to presynaptic axon terminal. As recorded with
record action potentials from one muscle monopolar or concentric needle electrodes
fiber. The following characteristics can be inserted in the end plate region, MEPPs are
specified quantitatively: (1) maximal peak- monophasic, negative, short duration (less
to-peak amplitude, (2) area contained under than 5 ms), and generally less than 20 μV in
the waveform, and (3) number of phases. amplitude.
Minimum Conduction Velocity: The nerve Motor Neuron Disease: A clinical con-
conduction velocity measured from slowly dition characterized by degeneration of
conducting nerve fibers. Special techniques motor nerve cells in the brain, brain stem,
are needed to produce this measurement in and spinal cord. The location of degen-
motor or sensory nerves. eration determines the clinical presenta-
Mixed Nerve: A nerve composed of both tion. Primary lateral sclerosis occurs when
motor and sensory axons. degeneration affects mainly corticospinal
MNCV: Abbreviation for motor nerve conduc- tract motor fibers. Spinal muscular atro-
tion velocity. See conduction velocity. phy occurs when degeneration affects lower
Mononeuritis Multiplex: A disorder char- motor neurons. Amyotrophic lateral scle-
acterized by axonal injury and/or demyeli- rosis occurs when degeneration affects
nation affecting nerve fibers in multiple both corticospinal tracts and lower motor
nerves (multiple mononeuropathies). Usu- neurons.
ally occurs in an asymmetric anatomic dis- Motor Point: The site over a muscle
tribution and in a temporal sequence which where its contraction may be elicited by
is not patterned or symmetric. a minimal intensity short-duration electric
Mononeuropathy Multiplex: A disorder stimulus.
characterized by axonal injury and/or Motor Response: (1) The compound mus-
demyelination affecting nerve fibers exclu- cle action potential (M wave) recorded over
sively along the course of one named a muscle in response to stimulation of
nerve. the nerve to the muscle. (2) The muscle
Monophasic Action Potential: An action twitch or contraction elicited by stimulation
potential with the waveform entirely on one of the nerve to a muscle. (3) The mus-
side of the baseline. cle twitch elicited by the muscle stretch
Monophasic End Plate Activity: See end reflex.
plate activity (monophasic). Motor Unit: The anatomic element consisting
Monopolar Needle Electrode: A solid wire of an anterior horn cell, its axon, the neu-
electrode coated with TeflonTM , except at the romuscular junctions, and all of the muscle
tip. Despite the term monopolar, a separate fibers innervated by the axon.
surface or subcutaneous reference electrode Motor Unit Action Potential (MUAP): The
is required for recording electric signals. compound action potential of a single motor
May also be used as a cathode in nerve unit whose muscle fibers lie within the
conduction studies with another electrode recording range of an electrode. With vol-
serving as an anode. untary muscle contraction, it is character-
Motor Evoked Potential (MEP): A com- ized by its consistent appearance and rela-
pound muscle action potential produced by tionship to the force of the contraction.
either transcranial magnetic stimulation or The following measures may be specified,
transcranial electrical stimulation. quantitatively if possible, after the record-
Motor Latency: Interval between the onset ing electrode is placed randomly within the
of a stimulus and the onset of the resultant muscle:
compound muscle action potential (M wave). 1. Configuration
The term may be qualified, as proxi- (a) Amplitude, peak-to-peak (μV or mV).
mal motor latency or distal motor latency, (b) Duration, total (ms).
depending on the relative position of the (c) Number of phases (monophasic, bipha-
stimulus. sic, triphasic, tetraphasic, and poly-
Motor Nerve: A nerve containing axons which phasic).
innervate extrafusal and intrafusal muscle (d) Polarity of each phase (negative,
fibers. These nerves also contain sensory positive).
afferent fibers from muscle and other deep (e) Number of turns.
structures. (f) Variation of shape (jiggle), if any, with
Motor Nerve Conduction Velocity (MNCV): consecutive discharges.
The speed of propagation of action poten- (g) Presence of satellite (linked) poten-
tials along a motor nerve. See conduction tials, if any.
velocity. (h) Spike duration, including satellites.
2. Recruitment characteristics Multifocal Motor Neuropathy: A disease

(a) Threshold of activation (first recruited, characterized by selective focal block of
low threshold, high threshold). motor nerve conduction in multiple nerves.
(b) Onset frequency. Motor nerve conduction studies may permit
(c) Recruitment frequency (Hz) or recruit- identification and localization of the seg-
ment interval (ms) of individual poten- ments of nerve affected by the underlying
tials. pathology.
Descriptive terms implying diagnostic signif- Multilead Electrode: Three or more insu-
icance are not recommended, for example, lated wires inserted through apertures in a
myopathic, neuropathic, regeneration, nascent, common metal cannula with their bared tips
giant, BSAP, and BSAPP. See polyphasic action flush with the cannula’s outer circumference.
potential and serrated action potential. The arrangement of the bare tips relative
Motor Unit Fraction: See scanning EMG. to the axis of the cannula and the distance
Motor Unit Number Counting: See the between each tip should be specified. See
preferred term motor unit number estimate electrode.
(MUNE). Multiple Discharge: Four or more motor
Motor Unit Number Estimate (MUNE): unit action potentials of the same form and
A quantitative technique for determining nearly the same amplitude occurring con-
the number of functioning motor units in sistently in the same relationship to one
a muscle. A variety of methods, includ- another and generated by the same axon.
ing spike-triggered averaging, incremental See double and triple discharge.
motor nerve stimulation, F-wave measure- Multiplet: See multiple discharge.
ment, or a Poisson statistical technique can MUNE: Abbreviation for motor unit number
be used. Synonyms can include motor unit estimate, motor unit number estimation, and
number estimation and motor unit number motor unit number estimating.
estimating. MUP: Abbreviation for motor unit potential.
Motor Unit Number Estimating (MUNE): See preferred term, motor unit action poten-
See motor unit number estimate (MUNE). tial.
Motor Unit Number Estimation (MUNE): Muscle Action Potential: Term commonly
See motor unit number estimate used to refer to a compound muscle action
(MUNE). potential.
Motor Unit Potential (MUP): See synonym, Muscle Atrophy: Decrease in size of a mus-
motor unit action potential. cle that may be due to disease of nerve or
Motor Unit Territory: The area of a muscle muscle, or todisuse.
cross section within which the muscle fibers Muscle Cramp: An involuntary, painful mus-
belonging to an individual motor unit are cle contraction associated with electrical
distributed. activity. Cramp discharges are most com-
Movement Artifact: See artifact. mon, but other types of repetitive discharges
Movement-Related Cortical Potential: can also be seen.
Electroencephalogram activity associated Muscle Fiber Action Potential: Action
with (before and after) a voluntary move- potential recorded from a single muscle
ment. There are several components includ- fiber.
ing the Bereitschaftspotential before the Muscle Fiber Conduction Velocity: The
movement and the motor potential at speed of propagation of a single muscle fiber
about the time of the movement. See also action potential, usually expressed as meters
Bereitschaftspotential. per second. Usually less than most nerve
MSD: Abbreviation for mean sorted difference. conduction velocities, varies with the rate of
See jitter. discharge of the muscle fiber, and requires
MUAP: Abbreviation for motor unit action special techniques for measurement.
potential. Muscle Hypertrophy: Increase in the size
Multi MUP Analysis: A template match- of a muscle due to an increase in the size
ing, decomposition EMG method used for of the muscle fibers or replacement or dis-
MUAP analysis. placement of muscle fibers by other tissues.
Multielectrode: See multilead electrode. The latter is also referred to by the term
pseudohypertrophy, because the muscle is Myopathic Motor Unit Potential: Low-

enlarged but weak. Muscle fibers increase amplitude, short-duration, polyphasic motor
in size as a physiologic response to repeti- unit action potentials. Use of term discour-
tive and forceful voluntary contraction or as aged. It incorrectly implies specific diag-
a pathologic response to involuntary electric nostic significance of a motor unit action
activity in a muscle, for example, myotonic potential configuration. See motor unit
discharges or complex repetitive discharges. action potential.
Muscle Stretch Reflex: Activation of a mus- Myopathic Recruitment: Used to describe
cle which follows stretch of the muscle, for an increase in the number and firing rate of
example, by percussion of a muscle tendon. motor unit action potentials compared with
See stretch reflex and T wave. normal for the strength of muscle contrac-
Muscle Tone: See tone. tion. Use of term discouraged.
Myasthenia Gravis: A disease characterized Myopathy: Disorder affecting the structure
by muscle weakness which increases with and/or function of muscle fibers. Etiologies
repetitive muscle activation. Most com- include hereditary, congenital, mitochon-
monly, an autoimmune disease caused by the drial, inflammatory, metabolic, infectious,
presence of antibodies to the acetylcholine neoplastic, vascular, and traumatic diseases.
receptors at the neuromuscular junction. Most, but not all of these disorders, show
Myoclonus: A quick jerk of a body part pro- abnormalities on needle electromyography.
duced by a brief muscle contraction typically Myotonia: Delayed relaxation of a muscle
originating from activity in the central ner- after voluntary contraction or percussion.
vous system. Based on the anatomic location Associated with propagated electric activ-
of the pathology, may be classified as spinal, ity, such as myotonic discharges, complex
segmental, brain stem, or cortical. repetitive discharges, or neuromyotonic dis-
Myoedema: Focal muscle contraction pro- charges.
duced by muscle percussion. Not associ- Myotonic Discharge: Repetitive discharge
ated with propagated electric activity. May which occurs at rates of 20–80 Hz. There are
be seen in hypothyroidism (myxedema) and two types: (1) biphasic (positive–negative)
chronic malnutrition. spike potentials less than 5 ms in duration
Myokymia: Continuous quivering or undu- resembling fibrillation potentials. (2) posi-
lating movement of surface and overly- tive waves of 5–20 ms duration resembling
ing skin and mucous membrane associated positive sharp waves. Both potential forms
with spontaneous, repetitive discharge of are recorded after needle electrode insertion,
motor unit action potentials. See myokymic after voluntary muscle contraction, or after
discharge, fasciculation, and fasciculation muscle percussion and are due to indepen-
potential. dent, repetitive discharges of single muscle
Myokymic Discharge: A form of involun- fibers. The amplitude and frequency of the
tary activity in which motor unit action potentials must both wax and wane. This
potentials fire repetitively and may be asso- change produces a characteristic musical
ciated with clinical myokymia. Two firing sound in the audio output of the electromyo-
patterns have been described: (1) Com- graph due to the corresponding change in
monly, the discharge is a brief, repetitive pitch, which has been likened to the sound
firing of single motor unit action poten- of a “dive bomber.” Contrast with waning
tials for a short period (up to a few sec- discharge.
onds) at a uniform rate (2–60 Hz) followed Myotonic Potential: See preferred term,
by a short period (up to a few seconds) of myotonic discharge.
silence, with repetition of the same sequence NAP: Abbreviation for nerve action potential.
for a particular potential at regular inter- See compound nerve action potential.
vals. (2) Rarely, the potential recurs con- Nascent Motor Unit Potential: From the
tinuously at a fairly uniform firing rate Latin nascens, “to be born.” Refers to
(1–5 Hz). Myokymic discharges are a sub- very low-amplitude, short-duration, highly
class of grouped discharges and repetitive polyphasic motor unit action potentials
discharges. See also ephapse and ephaptic observed during early states of reinnervation.
transmission. Use of term is discouraged, as it incorrectly
implies diagnostic significance of a motor assumed to refer to the maximum speed of

unit action potential configuration. See propagation unless otherwise specified. See
motor unit action potential. conduction velocity.
NCS: Abbreviation for nerve conduction study. Nerve Fiber Action Potential: Action poten-
NCV: Abbreviation for nerve conduction veloc- tial recorded from a single axon.
ity. See conduction velocity. Nerve Potential: Equivalent to nerve action
Near-Field: A region of electrical activity potential. Also commonly, but inaccurately,
where the isopotential voltage lines associ- used to refer to the biphasic form of end
ated with a current source change rapidly plate activity observed during needle elec-
over a short distance. The terms near-field trode examination of muscle. The latter use
and far-field are arbitrary designations, as is incorrect, because muscle fibers, not nerve
there are no agreed-upon criteria defining fibers, are the source of these potentials.
where the near-field ends and the far-field Nerve Trunk Action Potential: See pre-
begins. Compare with far-field. ferred term, compound nerve action poten-
Needle Electrode: An electrical device tial.
used for recording or stimulating that is Neurapraxia: Clinical term used to describe
positioned near the tissue of interest by pen- the reversible motor and sensory deficits
etration of the skin. See specific electrodes: produced by focal compressive or traction
bifilar (bipolar) needle recording electrode, lesions of large myelinated nerve fibers.
concentric needle electrode, macro-EMG It is due to conduction block, most often
needle electrode, monopolar needle elec- caused by focal demyelination, but, when
trode, multilead electrode, single fiber needle very short-lived, presumably caused by focal
electrode, and stimulating electrode. ischemia. The axon is not injured at the
Nerve Action Potential (NAP): Strictly lesion site. Compare with axonotmesis and
defined, refers to an action potential neurotmesis.
recorded from a single nerve fiber. The term Neuromuscular Transmission Disorder:
is commonly used to refer to the compound Clinical disorder associated with pathol-
nerve action potential. See compound nerve ogy affecting the structure and function of
action potential. the neuromuscular junction and interfer-
Nerve Conduction Study (NCS): Record- ing with synaptic transmission at that site.
ing and analysis of electric waveforms of Specific diseases include myasthenia gravis,
biologic origin elicited in response to elec- Lambert–Eaton myasthenic syndrome, and
tric or physiologic stimuli. The waveforms botulism.
are compound sensory nerve action poten- Neuromyopathy: Clinical disorder associated
tials, compound muscle action potentials, with pathology affecting both nerve and
or mixed nerve action potentials. The com- muscle fibers.
pound muscle action potentials are generally Neuromyotonia: Clinical syndrome of con-
referred to by letters which have historical tinuous muscle fiber activity manifested
origin: M wave, F wave, H wave, T wave, as continuous muscle rippling and stiff-
A wave, and R1, R2 waves. It is possible ness. It may be associated with delayed
under standardized conditions to establish relaxation following voluntary muscle con-
normal ranges for amplitude, duration, and traction. The accompanying electric activ-
latency of the waveforms and to calculate ity may be intermittent or continuous.
the maximum conduction velocity of sensory Terms used to describe related clinical syn-
and motor nerves. The term generally refers dromes are continuous muscle fiber activ-
to studies of waveforms generated in the ity syndrome, Isaac syndrome, Isaac–Merton
peripheral nervous system, whereas evoked syndrome, quantal squander syndrome, gen-
potential studies refers to studies of wave- eralized myokymia, pseudomyotonia, nor-
forms generated in both the peripheral and mocalcemic tetany, and neurotonia. Distin-
central nervous systems. Synonymous with guish from myotonia.
electroneurography. Neuromyotonic Discharge: Bursts of motor
Nerve Conduction Velocity (NCV): The unit action potentials that fire at high rates
speed of action potential propagation along (150–300 Hz) for a few seconds, often start-
a nerve fiber or nerve trunk. Generally ing or stopping abruptly. The amplitude of
the waveforms typically wanes. Discharges Order of Activation: The sequence of

may occur spontaneously or be initiated by appearance of different motor unit action
needle electrode movement, voluntary effort, potentials with increasing strength of volun-
ischemia, or percussion of a nerve. The tary contraction. See recruitment.
activity originates in motor axons. Distin- Orthodromic: Propagation of a nerve impulse
guish from myotonic discharges and complex in the same direction as physiologic conduc-
repetitive discharges. One type of electri- tion; for example, conduction along motor
cal activity recorded in patients who have nerve fibers toward the muscle and conduc-
clinical neuromyotonia. tion along sensory nerve fibers toward the
Neuropathic Motor Unit Potential: Abnor- spinal cord. Contrast with antidromic.
mally high-amplitude, long-duration, poly- Paired Stimuli: Two consecutive stimuli deliv-
phasic motor unit action potential. Use of ered in a time-locked fashion. The time
term discouraged. Incorrectly implies a spe- interval between the two stimuli and the
cific diagnostic significance of a motor unit intensity of each stimulus can be varied but
action potential configuration. See motor should be specified. The first stimulus is
unit action potential. called the conditioning stimulus and the sec-
Neuropathic Recruitment: A recruitment ond stimulus is the test stimulus. The condi-
pattern characterized by a decreased num- tioning stimulus may modify tissue excitabil-
ber of motor unit action potentials firing at a ity, which is then evaluated by the response
rapid rate. Use of term discouraged. See pre- to the test stimulus.
ferred terms, reduced interference pattern, Parasite Potential: See preferred term, satel-
discrete activity, and single unit pattern. lite potential.
Neuropathy: Disorder of the peripheral Peak Latency: Interval between the onset of
nerves. May be classified by the anatomi- a stimulus and a specified peak of an evoked
cal structure of the nerve most affected by waveform.
the disease: the cell body (neuronopathy), Peroneal Neuropathy at the Knee: A mono-
the axon (axonopathy), or the myelin sheath neuropathy involving the common peroneal
(demyelinating neuropathy). May selectively nerve as it passes around the head of the
affect motor or sensory nerves or both simul- fibula. The presumed mechanism is com-
taneously. The etiology may be hereditary, pression of the nerve against the fibula. See
metabolic, inflammatory, toxic, or unknown. also crossed leg palsy.
Neurotmesis: Partial or complete nerve sever- Phase: That portion of a waveform between
ance including the axons, associated myelin the departure from and the return to the
sheaths, and supporting connective tissues, baseline.
resulting in axonal degeneration distal to the Plexopathy: Axonal and/or demyelinating dis-
injury site. Compare with axonotmesis and order affecting the nerve fibers exclusive to
neurapraxia. the cervical, brachial, lumbar, or sacral rear-
Neurotonic Discharges: Repetitive motor rangement of spinal nerve roots into periph-
unit action potentials recorded from intra- eral nerves.
muscular electrodes during intraoperative Polarization: The presence of an elec-
monitoring. Thought to arise from irritation tric potential difference usually across an
or injury of nerves supplying the muscle excitable cell membrane.
from which the recording is made. Polyneuropathy: Axonal and/or demyelinat-
Noise: Electric activity not related to the sig- ing disorder affecting nerve fibers, usually in
nal of interest. In electrodiagnostic medicine, a symmetrical fashion. The distal segments
waveforms generated by electrodes, cables, of the longer nerves in the lower extremities
amplifier, or storage media and unrelated to are usually the most severely affected. May
potentials of biologic origin. The term has be classified as sensory, motor, or sensorimo-
also been used loosely to refer to one form tor depending on the function of nerve fibers
of end plate activity. affected.
Onset Frequency: The lowest stable firing Polyphasic Action Potential: An action pot-
rate for a single motor unit action poten- ential with four or more baseline crossings,
tial that can be voluntarily maintained by a producing five or more phases. See phase.
subject. Contrast with serrated action potential.
Polyradiculoneuropathy: See radiculopathy. Postactivation Facilitation: See facilitation.

Positive Sharp Wave: A biphasic, positive Postactivation Potentiation: An increase
then negative action potential of a sin- in the force of contraction (mechanical
gle muscle fiber. It is initiated by needle response) after a strong voluntary contrac-
electrode movement (insertional or unsus- tion. Contrast postactivation facilitation.
tained positive sharp wave) or occurs spon- Posttetanic: The period following tetanus.
taneously. Typically discharge in a uniform, Contrast with postactivation.
regular pattern at a rate of 1–50 Hz; the Posttetanic Depression: See postactivation
discharge frequency may decrease slightly depression.
just before cessation of discharge. The ini- Posttetanic Facilitation: See facilitation,
tial positive deflection is rapid (<1 ms), its posttetanic.
duration is usually less than 5 ms, and the Posttetanic Potentiation: (1) The increment-
amplitude is up to 1 mV. The negative phase ing mechanical response of muscle during
is of low amplitude, and its duration is 10– and after repetitive nerve stimulation. (2) In
100 ms. A sequence of positive sharp waves central nervous system physiology, enhance-
is commonly referred to as a train of posi- ment of excitability or reflex outflow of neu-
tive sharp waves. Assumed to be recorded ronal systems following a long period of
from a damaged area of a muscle fiber. This high-frequency stimulation. See facilitation
configuration may result from the position and potentiation.
of the needle electrode which is believed Potential: (1) A difference in charges, measur-
to be adjacent to the depolarized segment able in volts, that exists between two points.
of a muscle fiber injured by the electrode. Most biologically produced potentials arise
Note that the positive sharp waveform is not from the difference in charge between two
specific for muscle fiber damage. May occur sides of a cell membrane. (2) A term for a
in association with fibrillation potentials and physiologically recorded waveform.
are thought by some to be equivalent dis- Potentiation: Physiologically, the enhance-
charges. Motor unit action potentials and ment of a response. The convention used
potentials in myotonic discharges may have in this glossary is to use the term potentia-
the configuration of positive sharp waves. tion to describe the incrementing mechani-
Positive Wave: Loosely defined, the term cal response of muscle elicited by repetitive
refers to a positive sharp wave. See pre- nerve stimulation, for example, posttetanic
ferred term positive sharp wave. potentiation, whereas the term facilitation
Postactivation: The period following volun- is used to describe the incrementing elec-
tary activation of a nerve or muscle. Contrast trical response elicited by repetitive nerve
with posttetanic. stimulation, for example, postactivation
Postactivation Depression: A reduction in facilitation.
the amplitude and area of the M wave(s) Prolonged Insertion Activity: See insertion
in response to a single stimulus or train of activity.
stimuli which occurs within a few minutes Propagation Velocity of a Muscle Fiber:
following a 10–60 second strong voluntary The speed of transmission of a muscle fiber
contraction. Postactivation exhaustion refers action potential.
to the cellular mechanisms responsible for Pseudodecrement: An artifact produced by
the observed phenomenon of postactivation movement of the stimulating or recording
depression. Also used to describe reduction electrodes during repetitive nerve stimula-
of the M wave following a tetanus, which tion. The amplitude and area of the M wave
should more logically be termed posttetanic can vary in a way that resembles a decre-
depression. menting response; however, the responses
Postactivation Exhaustion: A reduction in are generally irregular and not reproducible.
the safety factor (margin) of neuromuscu- Pseudofacilitation: See facilitation.
lar transmission after sustained activation at Pseudohypertrophy: See muscle hypertro-
the neuromuscular junction. The changes phy.
in the configuration of the M wave due to Pseudomyotonic Discharge: Formerly used
postactivation exhaustion are referred to as to describe complex repetitive discharges.
postactivation depression. Use of term discouraged.
Pseudopolyphasic Action Potential: Use of Reciprocal Inhibition: Inhibition of a motor

term discouraged. See preferred term, ser- neuron pool secondary to the activation of
rated action potential. the motor neuron pool of its antagonist. It
QEMG: Abbreviation for quantitative elec- is one of several important spinal mecha-
tromyography. nisms of motor control that help to make
QSART: Abbreviaton for quantitative sudomo- movements smoother and utilize less energy.
tor axon reflex test. There are multiple mechanisms for recip-
QST: Abbreviaton for quantitative sensory rocal inhibition, including one mediated by
testing. the Ia inhibitory interneuron that actives Ia
Quantitative Electromyography (QEMG): afferents and disynaptically inhibits the mus-
A systematic method for measuring the cle that is the antagonist to the source of the
recordings made by an intramuscular nee- Ia afferents.
dle electrode. Measurements include motor Recording Electrode: Device used to record
unit action potential characteristics such as electric potential difference. All electric
amplitude, duration, and phases, or interfer- recordings require two electrodes. The elec-
ence pattern characteristics. See turns and trode close to the source of the activity to
amplitude analysis. be recorded is called the active or explor-
Quantitative Sensory Testing (QST): An ing electrode, and the other recording elec-
instrumented method for measuring cuta- trode is called the reference electrode. Active
neous sensation. electrode is synonymous with input termi-
Quantitative Sudomotor Axon Reflex Text nal 1, or E-1 (or older terms whose use
(QSART): Test of postganglionic sympa- is discouraged, i.e., grid 1 and G1). Ref-
thetic sudomotor axons function by mea- erence electrode is synonymous with input
suring sweat output following activation of terminal 2, or E-2 (or older terms whose
axon terminals by local application of acetyl- use is discouraged, i.e., grid 2 and G2).
choline. Antidromic transmission of the In some recordings it is not certain which
impulse from the nerve terminals reaches electrode is closer to the source of the bio-
a branch point, then travels orthodromi- logic activity, for example, recording with a
cally to release acetylcholine from the nerve bifilar needle recording electrode, or when
terminals, inducing a sweating response. In attempting to define far-field potentials. In
small fiber polyneuropathy, the response this situation, it is convenient to refer to one
may be reduced or absent. In painful neu- electrode as input electrode 1, or E-1, and
ropathies, and in reflex sympathetic dys- the other as input electrode 2, or E-2. By
trophy, the response may be excessive and present convention, a potential difference
persistent or reduced. that is negative at the active electrode (input
R1, R2 Waves: See blink responses. terminal 1, E-1) relative to the reference
Radiculopathy: Axonal and/or demyelinating electrode (input terminal 2, E-2) causes an
disorder affecting the nerve fibers exclusive upward deflection on the display screen. The
to one spinal nerve root or spinal nerve. term “monopolar recording” is not recom-
May affect the anterior (motor) or poste- mended, because all recordings require two
rior (sensory) spinal nerve roots, or both, at electrodes; however, it is commonly used to
one spinal cord segment level. The resulting describe the use of one type of intramuscu-
clinical syndrome may include pain, sensory lar needle electrode. A similar combination
loss, paresthesia, weakness, fasciculations, of needle electrodes has been used to record
and muscle atrophy. If more than one spinal nerve activity and also has been referred to
root is involved, the term polyradiculopathy as “monopolar recording.”
may be used as a descriptor. Recruitment: The successive activation of
Raster: A method for display of a free- the same and additional motor units with
running sweep in electromyography. Sweeps increasing strength of voluntary muscle con-
are offset vertically so that each succes- traction. See motor unit action potential.
sive sweep is displayed below the one Recruitment Frequency: Firing rate of a
preceding it. motor unit action potential (MUAP) when a
Raw EMG: Unprocessed EMG signal recorded different MUAP first appears during gradu-
with surface or intramuscular electrodes. ally increasing voluntary muscle contraction.
This parameter is essential to assessment of produce another action potential. The abso-
recruitment pattern. lute refractory period is the time following
Recruitment Interval: The interdischarge an action potential during which no stimulus,
interval between two consecutive discharges however strong, evokes a further response.
of a motor unit action potential (MUAP) The relative refractory period is the time fol-
when a different MUAP first appears during lowing an action potential during which a
gradually increasing voluntary muscle con- stimulus must be abnormally large to evoke
traction. The reciprocal of the recruitment a second response. The functional refractory
interval is the recruitment frequency. See period is the time following an action poten-
also interdischarge interval. tial during which a second action potential
Recruitment Pattern: A qualitative and/or cannot yet excite the given region.
quantitative description of the sequence of Refractory Period of Transmission: Interval
appearance of motor unit action potentials following an action potential during which a
during increasing voluntary muscle contrac- nerve cannot conduct a second one. Distin-
tion. The recruitment frequency and recruit- guish from refractory period, as commonly
ment interval are two quantitative measures used, which deals with the ability of a stimu-
commonly used. See interference pattern, lus to produce an action potential.
early recruitment, and reduced recruitment Regeneration Motor Unit Potential: Use
for qualitative terms commonly used. of term discouraged. See motor unit action
Recurrent Inhibition: Decreased probabil- potential.
ity of firing of a motor neuron pool medi- Relative Refractory Period: See refractory
ated by Renshaw cells. Renshaw cells are period.
activated by recurrent collaterals from the Repair of the Decrement: See facilitation.
axons of alpha-motoneurons. Such inhibi- Repetitive Discharge: General term for the
tion influences the same cells that orig- recurrence of an action potential with the
inate the excitatory impulses and their same or nearly identical form. May refer
neighbors. to recurring potentials recorded in muscle
Reduced Insertion Activity: See insertion at rest, during voluntary contraction, or in
activity. response to a single nerve stimulus. See dou-
Reduced Interference Pattern: See inter- ble discharge, triple discharge, multiple dis-
ference pattern. charge, myokymic discharge, complex repet-
Reduced Recruitment Pattern: A descrip- itive discharge, neuromyotonic discharge,
tive term for the interference pattern when and cramp discharge.
the number of motor units available to gen- Repetitive Nerve Stimulation: The tech-
erate a muscle contraction are reduced. One nique of repeated supramaximal stimula-
cause for a reduced interference pattern. tion of a nerve while recording succes-
See interference pattern and recruitment sive M waves from a muscle innervated
pattern. by the nerve. Commonly used to assess
Reference Electrode: See recording elec- the integrity of neuromuscular transmission.
trode. The number of stimuli and the frequency of
Reflex: A stereotyped motor response elicited stimulation should be specified. Activation
by a sensory stimulus and a response. Its procedures performed as a part of the test
anatomic pathway consists of an afferent, should be specified, for example, sustained
sensory input to the central nervous sys- voluntary contraction or contraction induced
tem, at least one synaptic connection, and by nerve stimulation. If the test includes
an efferent output to an effector organ. an activation procedure, the time elapsed
The response is most commonly motor, but after its completion should also be speci-
reflexes involving autonomic effector organs fied. For a description of specific patterns of
also occur. Examples include the H reflex responses, see incrementing response, decre-
and the sudomotor reflex. See H wave and menting response, facilitation, and postacti-
quantitative sudomotor axon reflex test. vation depression.
Refractory Period: General term for the Repolarization: A return in membrane poten-
time following an action potential when an tial from a depolarized state toward the
excitable membrane cannot be stimulated to normal resting level.
Residual Latency: The calculated time differ- common branch of an axon. These groups of
ence between the measured distal latency of fibers form a motor unit fraction.
a motor nerve and the expected latency, cal- Sea Shell Sound (Sea Shell Roar or Noise):
culated by dividing the distance between the Use of term discouraged. See end plate
stimulating cathode and the active recording activity and monophasic.
electrode by the maximum conduction veloc- Sensory Latency: Interval between the onset
ity measured in a more proximal segment of a stimulus and the onset of the negative
of the nerve. It is due in part to neuro- deflection of the compound sensory nerve
muscular transmission time and to slowing action potential. This term has been used
of conduction velocity in terminal axons due loosely to refer to the sensory peak latency.
to decreasing diameter and the presence of May be qualified as proximal sensory latency
unmyelinated segments. or distal sensory latency, depending on the
Response: An activity elicited by a stimulus. relative position of the stimulus.
Resting Membrane Potential: Voltage across Sensory Nerve: A nerve containing only sen-
the membrane of an excitable cell in the sory fibers, composed mainly of axons inner-
absence of a stimulus. See polarization. vating cutaneous receptors.
Rheobase: See strength–duration curve. Sensory Nerve Action Potential (SNAP):
Rigidity: A velocity-independent increase in See compound sensory nerve action poten-
muscle tone and stiffness with full range tial.
of joint motion as interpreted by the clin- Sensory Nerve Conduction Velocity: The
ical examiner from the physical examina- speed of propagation of action potentials
tion. Often associated with simultaneous along a sensory nerve.
low-grade contraction of agonist and antag- Sensory Peak Latency: Interval between the
onist muscles. Like muscle spasticity, the onset of a stimulus and the peak of the
involuntary motor unit action potential activ- negative phase of the compound sensory
ity increases with activity or passive stretch. nerve action potential. Contrast with sensory
Does not seem to change with the veloc- latency.
ity of stretch, and, on passive stretch, the Sensory Potential: Synonym for the more
increased tone has a “lead pipe” or constant precise term, compound sensory nerve
quality. It is a cardinal feature of central action potential.
nervous system disorders affecting the basal Sensory Response: Synonym for the more
ganglia. Contrast with spasticity. precise term, compound sensory nerve
Rise Time: The interval from the onset of a action potential.
polarity change of a potential to its peak. SEP: Abbreviation for somatosensory evoked
The method of measurement should be potential.
specified. Serrated Action Potential: A waveform with
Satellite Potential: A small action poten- several changes in direction (turns) which
tial separated from the main motor unit do not cross the baseline. Most often used
action potential by an isoelectric interval to describe a motor unit action potential.
which fires in a time-locked relationship to The term is preferred to complex motor
the main action potential. It usually follows, unit action potential and pseudopolyphasic
but may precede, the main action potential. action potential. See also turn and polypha-
Less preferred terms include late compo- sic action potential.
nent, parasite potential, linked potential, and SFEMG: Abbreviation for single fiber elec-
coupled discharge. tromyography.
Scanning EMG: A technique by which a nee- Shock Artifact: See artifact.
dle electrode is advanced in defined steps Short-Latency Reflex: A reflex with one
through muscle while a separate SFEMG (monosynaptic) or few (oligosynaptic) syna-
electrode is used to trigger both the dis- pses. Used in contrast to long-latency reflex.
play sweep and the advancement device. Short-Latency Somatosensory Evoked
Provides temporal and spatial information Potential (SSEP): That portion of the
about the motor unit. Distinct maxima in waveforms of a somatosensory evoked poten-
the recorded activity are considered to be tial normally occurring within 25 ms after
generated by muscle fibers innervated by a stimulation of the median nerve in the upper
extremity at the wrist, 40 ms after stimula- firings produces a distinctive sound, hence
tion of the common peroneal nerve in the the name. Seen most often in those with
lower extremity at the knee, and 50 ms after mesomorphic builds, especially young adult
stimulation of the posterior tibial nerve at males. Found most often in lower extrem-
the ankle. ity muscles, especially the medial gastrocne-
Signal Averager: A digital device that mius.
improves the signal-to-noise ratio of an elec- Somatosensory Evoked Potential (SEP):
trophysiological recording by adding suc- Electric waveforms of biologic origin elicited
cessive time-locked recordings to preced- by electric stimulation or physiologic acti-
ing traces and computing the average value vation of peripheral sensory nerves and
of each data point. A signal acquired by recorded from peripheral and central ner-
this method is described as an “averaged” vous system structures. Normally is a com-
waveform. plex waveform with several components
Silent Period: A pause in the electric activ- which are specified by polarity and aver-
ity of a muscle that may be produced by age peak latency. The polarity and latency
many different stimuli. Stimuli used com- of individual components depend upon (1)
monly in clinical neurophysiology include subject variables, such as age, gender, and
rapid unloading of a muscle, electrical stim- body habitus, (2) stimulus characteristics,
ulation of a peripheral nerve, or transcranial such as intensity and rate of stimulation,
magnetic stimulation. and (3) recording parameters, such as ampli-
Single Fiber Electromyography (SFEMG): fier time constants, electrode placement, and
The technique and conditions that permit electrode combinations. See short-latency
recording of single muscle fiber action poten- somatosensory evoked potentials.
tials. See single fiber needle electrode, block- Spasticity: A velocity-dependent increase in
ing, and jitter. muscle tone due to a disease process that
Single Fiber EMG: See single fiber elec- interrupts the suprasegmental tracts to the
tromyography. alpha motor neurons, gamma motor neu-
Single Fiber Needle Electrode: A nee- rons, or segmental spinal neurons. May
dle electrode with a small recording surface be elicited and interpreted by the clini-
(usually 25 μm in diameter) which permits cal examiner during the physical examina-
the recording of single muscle fiber action tion by brisk passive movement of a limb
potentials between the recording surface at the joint. Almost uniformly accompa-
and the cannula. See single fiber electromyo- nied by hyperreflexia, a Babinski sign, and
graphy. other signs of upper motor neuron pathol-
Single Unit Pattern: See interference pat- ogy, including clonus and the clasp-knife
tern. phenomenon. The clasp-knife phenomenon
SNAP: Abbreviation for sensory nerve action is a rapid decrease of tone following a period
potential. See compound sensory nerve of increased tone during passive rotation of
action potential. the joint. The pathophysiology is not certain
Snap, Crackle, and Pop: A benign type and may include more than dysfunction of
of increased insertion activity that follows, the corticospinal tracts.
after a very brief period of electrical silence, Spike: (1) A short-lived (1–3 ms), all-or-none
the normal insertion activity generated by waveform that arises when an excitable
needle electrode movement. It consists of membrane reaches threshold. (2) The elec-
trains of potentials that vary in length; how- tric record of a nerve or muscle impulse.
ever, they can persist for a few seconds. Spinal Evoked Potential: Electric wave-
Each train consists of a series of up to 10 forms of biologic origin recorded over the
or more potentials in which the individ- spine in response to electric stimulation or
ual components fire at irregular intervals. physiologic activation of peripheral sensory
The potentials consistently vary in ampli- fibers. See preferred term, somatosensory
tude, duration, and configuration. Individual evoked potential.
potentials may be mono-, bi-, tri-, or multi- Spontaneous Activity: Electric activity recor-
phasic in appearance; they often have a pos- ded from muscle at rest after insertion
itive waveform. The variation on sequential activity has subsided and when there is not
voluntary contraction or an external stimu- single fiber electromyography. The method

lus. Compare with involuntary activity. is used in patients who are unable to produce
SSEP: Abbreviation for short-latency somatosen- a steady voluntary muscle contraction. The
sory evoked potential. stimulation can be delivered to intramuscu-
Staircase Phenomenon: The progressive lar axons, nerve trunks, or muscle fibers.
increase in muscle contraction force observed Stimulus: Any external agent, state, or change
in response to continued low rates of muscle that is capable of influencing the activity of
activation. a cell, tissue, or organism. In clinical nerve
Startle (Reflex): A response produced by an conduction studies, an electric stimulus is
unanticipated stimulus that leads to alerting applied to a nerve. It may be described in
and protective movements such as eye lid absolute terms or with respect to the evoked
closure and flexion of the limbs. Auditory potential of the nerve or muscle. In abso-
stimuli are typically most efficacious. lute terms, it is defined by a duration (ms), a
Stiff-man Syndrome: A disorder charac- waveform (square, exponential, linear, etc.),
terized by continuous muscle contraction and a strength or intensity measured in volt-
giving rise to severe stiffness. Axial mus- age (V) or current (mA). With respect to
cles are typically affected most severely. the evoked potential, the stimulus may be
Patients have difficulty moving. Walking graded as subthreshold, threshold, submaxi-
and voluntary movements are slow. Sensory mal, maximal, or supramaximal. A threshold
stimulation often induces severe spasms. stimulus is one just sufficient to produce a
Electromyography demonstrates continuous detectable response. Stimuli less than the
activity of motor unit action potentials in a threshold stimulus are termed subthreshold.
normal pattern that cannot be silenced by The maximal stimulus is the stimulus inten-
contraction of the antagonist muscle. It is sity after which a further increase in intensity
often associated with circulating antibodies causes no increase in the amplitude of the
to glutamic acid decarboxylase (GAD), and evoked potential. Stimuli of intensity below
the resulting deficiency of GABA may play a this level but above threshold are submax-
role in its pathophysiology. Since women are imal. Stimuli of intensity greater than the
affected in equal or greater numbers than maximal stimulus are termed supramaximal.
men, the term stiff-person syndrome may be Ordinarily, supramaximal stimuli are used
preferable. for nerve conduction studies. By conven-
Stiff-person Syndrome: Synonym for stiff- tion, an electric stimulus of approximately
man syndrome. 20% greater voltage/current than required
Stigmatic Electrode: A term of historic inter- for the maximal stimulus is used for supra-
est. Used by Sherrington for active or explor- maximal stimulation. The frequency, num-
ing electrode. ber, and duration of a series of stimuli should
Stimulated SFEMG: See preferred term be specified.
stimulation SFEMG. Stimulus Artifact: See artifact.
Stimulating Electrode: Device used to Strength–Duration Curve: Graphic presen-
deliver electric current. All electric stimu- tation of the relationship between the inten-
lation requires two electrodes; the negative sity (Y axis) and various durations (X axis) of
terminal is termed the cathode, and the pos- the threshold electric stimulus of a nerve or
itive terminal is the anode. By convention, muscle. The rheobase is the intensity of an
the stimulating electrodes are called bipolar electric current of infinite duration neces-
if they are encased or attached together and sary to produce a minimal action potential.
are called monopolar if they are not. Elec- The chronaxie is the time required for an
tric stimulation for nerve conduction studies electric current twice the rheobase to elicit
generally requires application of the cathode the first visible action potential. Measure-
in the vicinity of the neural tissue to produce ment of the strength–duration curve is not a
depolarization. common practice in modern electrodiagnos-
Stimulation Single Fiber Electromyogra- tic medicine.
phy (Stimulation SFEMG): Use of elec- Stretch Reflex: A reflex produced by pas-
trical stimulation instead of voluntary sive lengthening of a muscle. The principal
activation of motor units for the analysis of sensory stimuli come from group Ia and
group II muscle spindle afferents. It con- tunnel syndrome and ulnar neuropathy at
sists of several phases. The earliest compo- the elbow.
nent is monosynaptic and is also called the Template Matching: An automated method
myotatic reflex, or tendon reflex. There are used in quantitative electromyography for
also long-latency stretch reflexes. See also selecting motor unit action potentials for
muscle stretch reflex and T wave. measurement by extracting only poten-
Submaximal Stimulus: See stimulus. tials which resemble an initially identified
Subnormal Period: A time interval that potential.
immediately follows the supernormal period Temporal Dispersion: Relative desynchro-
of nerve which is characterized by reduced nization of components of a compound mus-
excitability compared to the resting state. cle action potential due to different rates
Its duration is variable and is related to the of conduction of each synchronously evoked
refractory period. component from the stimulation point to
Subthreshold Stimulus: See stimulus. the recording electrode. It may be due to
Supernormal Period: A time interval that normal variability in individual axon conduc-
immediately follows the refractory period tion velocities, especially when assessed over
which corresponds to a very brief period of a long nerve segment, or to disorders that
partial depolarization. It is characterized by affect myelination of nerve fibers.
increased nerve excitability and is followed Terminal Latency: Synonymous with pre-
by the subnormal period. ferred term, distal latency. See motor
Supraclavicular Plexus: That portion of the latency and sensory latency.
brachial plexus which is located superior to TES: Abbreviation for transcranial electrical
the clavicle. stimulation.
Supraclavicular Stimulation: Percutaneous Test Stimulus: See paired stimuli.
nerve stimulation at the base of the neck Tetanic Contraction: The contraction pro-
which activates the upper, middle, and/or duced in a muscle through repetitive maxi-
lower trunks of the brachial plexus. This mal direct or indirect stimulation at a suffi-
term is preferred to Erb’s point stimulation. ciently high frequency to produce a smooth
Supramaximal Stimulus: See stimulus. summation of successive maximum twitches.
Surface Electrode: Conducting device for The term may also be applied to maximum
stimulating or recording placed on the skin voluntary contractions in which the firing
surface. The material (metal, fabric, etc.), frequencies of most or all of the component
configuration (disk, ring, etc.), size, and sep- motor units are sufficiently high that succes-
aration should be specified. See electrode sive twitches of individual motor units fuse
(ground, recording, stimulating). smoothly. Their combined tensions produce
Sympathetic Skin Response: Electrical poten- a steady, smooth, maximum contraction of
tial resulting from electrodermal activity in the whole muscle.
sweat glands in response to both direct and Tetanus: (1) The continuous contraction of
reflex peripheral or sympathetic trunk stim- muscle caused by repetitive stimulation or
ulation of autonomic activity. discharge of nerve or muscle. Contrast with
Synkinesis: Involuntary movement made by tetany. (2) A clinical disorder caused by cir-
muscles distant from those activated volun- culating tetanus toxin. Signs and symptoms
tarily. It is commonly seen during recovery are caused by loss of inhibition in the cen-
after facial neuropathy. It is due to aberrant tral nervous system and are characterized
reinnervation and/or ephaptic transmission. by muscle spasms, hyperreflexia, seizures,
T Wave: A compound muscle action poten- respiratory spasms, and paralysis.
tial evoked from a muscle by rapid stretch Tetany: A clinical syndrome manifested by
of its tendon, as part of the muscle stretch muscle twitching, cramps, and carpal and
reflex. pedal spasm. These clinical signs are mani-
Tardy Ulnar Palsy: A type of mononeu- festations of peripheral and central nervous
ropathy involving the ulnar nerve at the system nerve irritability from several causes.
elbow. The nerve becomes compressed or In these conditions, repetitive discharges
entrapped due to deformity of the elbow (double discharge, triple discharge, multiple
from a previous injury. See also cubital discharge) occur frequently with voluntary
activation of motor unit action potentials is low, this is called hypotonia. Two types of
or may appear as spontaneous activity. This hypertonia are rigidity and spasticity.
activity is enhanced by systemic alkalosis or Train of Positive Sharp Waves: See positive
local ischemia. sharp wave.
Tetraphasic Action Potential: Action poten- Train of Stimuli: A group of stimuli. The
tial with three baseline crossings, producing duration of the group or the number of
four phases. stimuli as well as the stimulation frequency
Thermography: A technique for measuring should be specified.
infrared emission from portions of the body Transcranial Electrical Stimulation (TES):
surface. The degree of emission depends Stimulation of the cortex of the brain
upon the amount of heat produced by the through the intact skull and scalp by means
region that is studied. Its use in the diagno- of a brief, very high voltage, electrical stim-
sis of radiculopathy, peripheral nerve injury, ulus. Activation is more likely under the
and disorders of the autonomic nervous sys- anode rather than the cathode. Because
tem is controversial. it is painful, this technique has largely
Thermoregulatory Sweat Test: A technique been replaced by transcranial magnetic
for assessing the integrity of the central and stimulation.
peripheral efferent sympathetic pathways. It Transcranial Magnetic Stimulation (TMS):
consists of measuring the sweat distribution Stimulation of the cortex of the brain
using an indicator powder while applying a through the intact skull and scalp by means
controlled heat stimulus to raise body tem- of a brief magnetic stimulus. In practice,
perature sufficient to induce sweating. a brief pulse of strong current is passed
Thoracic Outlet Syndrome: An entrapment through a coil of wire in order to produce
neuropathy caused by compression of the a time-varying magnetic field in the order of
neurovascular bundle as it traverses the 1–2 Tesla. Contrast with transcranial electri-
shoulder region. Compression arises from cal stimulation.
acquired or congenital anatomic variations Tremor: Rhythmical, involuntary oscillatory
in the shoulder region. Symptoms can be movement of a body part.
related to compression of vascular struc- Triphasic Action Potential: Action potential
tures, portions of the brachial plexus, or with two baseline crossings, producing three
both. phases.
Threshold: The level at which a clear and Triple Discharge: Three motor unit action
abrupt transition occurs from one state to potentials of the same form and nearly the
another. The term is generally used to refer same amplitude, occurring consistently in
to the voltage level at which an action poten- the same relationship to one another and
tial is initiated in a single axon or muscle generated by the same axon. The interval
fiber or a group of axons or muscle fibers. between the second and third action poten-
Threshold Stimulus: See stimulus. tials often exceeds that between the first
Tic: Clinical term used to describe a sudden, two, and both are usually in the range of
brief, stereotyped, repetitive movement. 2–20 ms. See also double discharge and mul-
When associated with vocalizations, may be tiple discharge.
the primary manifestation of Tourette syn- Triplet: Synonym for the preferred term,
drome. triple discharge.
Tilt Table Test: A test of autonomic function Turn: Point of change in polarity of a wave-
that is performed by measuring blood pres- form and the magnitude of the voltage
sure and heart rate before and a specified change following the turning point. It is
period of time after head-up tilt. The dura- not necessary that the voltage change pass
tion of recording and amount of tilt should through the baseline. The minimal excursion
be specified. required to constitute a change should be
TMS: Abbreviation for transcranial magnetic specified.
stimulation. Turns and Amplitude Analysis: See pre-
Tone: The resistance to passive stretch of a ferred term interference pattern analysis.
joint. When the resistance is high, this is Refers to the interference pattern analysis
called hypertonia, and when the resistance developed by Robin Willison in the 1960s.
Ulnar Neuropathy at the Elbow: A mononeu- (N75). The precise range of normal val-
ropathy involving the ulnar nerve in the ues for the latency and amplitude of P100
region of the elbow. At least two sites of depends on several factors: (1) subject vari-
entrapment neuropathy have been recog- ables, such as age, gender, and visual acu-
nized. The nerve may be entrapped or com- ity, (2) stimulus characteristics, such as type
pressed as it passes through the retrocondy- of stimulator, full-field or half-field stim-
lar groove at the elbow. Alternatively, it may ulation, check size, contrast and lumines-
be entrapped just distal to the elbow as it cence, and (3) recording parameters, such
passes through the cubital tunnel. Anatomic as placement and combination of recording
variations or deformities of the elbow may electrodes.
contribute to nerve injury. See also cubital Visual Evoked Response (VER): Synonym
tunnel syndrome and tardy ulnar palsy. for preferred term, visual evoked potential.
Unipolar Needle Electrode: See synonym, Volitional Activity: Synonymous with volun-
monopolar needle recording electrode. tary activity.
Upper Motor Neuron Syndrome: A clinical Voltage: Potential difference between two
condition resulting from a pathological pro- recording sites usually expressed in volts (V)
cess affecting descending motor pathways or millivolts (mV).
including the corticospinal tract or its cells Volume Conduction: Spread of current from
of origin. Signs and symptoms include weak- a potential source through a conducting
ness, spasticity, and slow and clumsy motor medium, such as body tissues.
performance. On electromyographic exami- Voluntary Activity: In electromyography,
nation of weak muscles, there is slow motor the electric activity recorded from a mus-
unit action potential firing at maximal effort. cle with consciously controlled contrac-
Utilization Time: See preferred term, latency tion. The effort made to contract the
of activation. muscle may be specified relative to that
Valsalva Maneuver: A forcible exhalation of a corresponding normal muscle, for
against the closed glottis which creates an example, minimal, moderate, or maximal.
abrupt, transient elevation of intrathoracic If the recording remains isoelectric dur-
and intra-abdominal pressure. This results ing the attempted contraction and equip-
in a characteristic pattern of heart rate and ment malfunction has been excluded, it can
blood pressure changes that can be used to be concluded that there is no voluntary
quantify autonomic function. See Valsalva activity.
ratio. Wake-up Test: A procedure used most com-
Valsalva Ratio: The ratio of the fastest heart monly in spinal surgery. During critical por-
rate occurring at the end of a forced exhala- tions of an operation in which the spinal
tion against a closed glottis (phase II of the cord is at risk for injury, the level of gen-
Valsalva maneuver), and the slowest heart eral anesthesia is allowed to decrease to the
rate within 30 seconds after the forced exha- point where the patient can respond to com-
lation (phase IV). In patients with disorders mands. The patient is then asked to move
of the autonomic nervous system, the ratio hands and feet, and a movement in response
may be reduced. to commands indicates the spinal cord is
VEP: Abbreviation for visual evoked potential. intact. This procedure is used routinely in
VER: Abbreviation for visual evoked response. some centers. Somatosensory evoked poten-
See visual evoked potential. tial monitoring has supplanted its use in
Visual Evoked Potential (VEP): Electric most centers, except sometimes in the situ-
waveforms of biologic origin recorded over ation where they indicate the possibility of
the cerebrum and elicited in response to spinal cord injury.
visual stimuli. They are classified by stimu- Wallerian Degeneration: Degeneration of
lus rate as transient or steady state, and they the segment of an axon distal to nerve injury
can be further divided by stimulus presen- that destroys its continuity.
tation mode. The normal transient VEP to Waning Discharge: A repetitive discharge
checkerboard pattern reversal or shift has a that gradually decreases in frequency or
major positive occipital peak at about 100 ms amplitude before cessation. Contrast with
(P100), often preceded by a negative peak myotonic discharge.
Wave: A transient change in voltage represen- needle. After the needle is withdrawn, the
ted as a line of differing directions over time. wire remains in place. Wire electrodes are
Waveform: The shape of a wave. The term is superior to surface electrodes for kinesiologic
often used synonymously with wave. EMG, because they are less affected by cross
Wire Electrodes: Thin wires that are insu- talk from adjacent muscles. They also record
lated except for the tips, which are bared. selectively from the muscle into which they
The wire is inserted into muscle with a are inserted.
Index
Note: In this index, figures are indicated by “f” and tables by “t”
A waves. See Axon reflexes Alternation, 503

Abductor digiti minimi (ADM), 809, 832 Alzheimer’s disease, 565
Abductor pollicis brevis (APB), 808 Ambulatory electroencephalography (AEEG), 187
Absence seizures, 141, 176f clinical applications, 189–91
Absolute latencies of BAEP waves, 307 indications, 187–88
Accommodation, 82, 82f purpose and role, 187
Acetylcholine, 371 technology, 188–89
Acoustic neuroma, 289f, 289–90 Amplification, 17–18
Acoustic reflex, 298–99, 299t, 300f Amplifiers, 16–18, 478
CN VIII vs. cochlear findings, 299t, 299–300 differential, 18f, 18–19, 19f
decay, 299 Amplitude, 458
disorders of brain stem, 300–301 Amyotrophic lateral sclerosis (ALS), 255, 275f, 431, 504f,
disorders of CN VII, 299t, 300 506f
Acoustic reflex tests, 295 and nerve conduction variables, 352t
purpose and role, 296 Amyotrophy, neuralgic, 363
Actigraphy, 704f, 704–5 Analog signals, 56
indications, 705 Analog-to-digital conversion, 57–59. See also Digitization
purpose and role, 698 Analog-to-digital converter (ADC), 56–58
Action potentials, 82 scheme of a 4-bit, 57f
excitability, 84–86, 85f Anesthesia, 316, 728f
ionic basis, 83–84, 84f bispectral analysis of EEG for monitoring depth of, 736
vs. local potentials, 80t light steady-state, 731–32
patterns of activity, 87–88 motor evoked potentials and, 760
propagation, 86–87 preoperative abnormalities seen with, 732–33
threshold, 82–83 somatosensory evoked potentials and, 757f, 757–58
Action tremors, 555 symmetrical EEG patterns during, 731–32
Active reference electrodes, 45 Angular head movements, 577–79
Active sleep, 168 Anhidrosis, 646t, 654. See also Thermoregulatory sweat
Active sources (of bioelectric potentials), 33 test
Active transport, 70–71 Anodal block, 245, 332. See also Hyperpolarization
Active zones, 90 Anomalous innervation, 831
Acute inflammatory demyelinating Anoxia, 276
polyradiculoneuropathy (AIDP). See Anterior horn cell disorders, 489–90
Guillain-Barré syndrome Anterior maximum, rhythmic fast (ARF) pattern, 731–32
Acute process, 805 Anterior maximum, triangular, slow (ATS) waves, 731
Adenosine triphosphate (ATP), 70, 93–94, 95f Anterior temporal and frontal spikes, 138
Adrenergic function, 637. See also Skin vasomotor reflexes Anterior visual pathway lesions, 317–21
After-hyperpolarization, 84, 88 Anticoagulation, needle exams and, 409–10
Afterpotential, 84 Antidromic (sensory) technique, 243, 244f, 253
Airflow and sleep, 713 Apnea hypopnea index (AHI), 715
Alexander’s Law, 580, 581 Apneas, 712. See also Obstructive sleep apnea
Aliasing, 59 mixed, 712
Allodynia, 680 Appearance threshold, 678
Alpha activity, 124 Arm-diaphragm synkinesis, 437
Alpha–delta sleep, 707 Arousal, disorders of, 719
Alpha-frequency coma pattern, 163 Arousal index, 710
Alpha intrusion, 709 Arousals, 7097
Alpha rhythm, 124, 124f Arrhythmia. See also Hypsarrhythmia; Positive,
Alpha variants, 130 arrhythmic, slow (PAS) waves
Altered reactivity, 153 respiratory sinus, 624, 627
Alternating current EMF, 12, 12f Arterial baroreflexes. See Baroreflexes
Alternating (polarities), 287 Arteriovenous malformations, 155
869
870 Index
Artifacts, 111, 123–24 Base (transistor), 17

common forms, 112t Basement syndrome, 608
electric, 267 Behavioral events, nonepileptic
nonphysiologic, 111–13, 113f PVEEG and, 196–97
physiologic, 111 Behavior disorder, REM sleep, 720
Artifactual waveforms, 111–13 Bending energy, 207
Asterixis, 566 Benign epileptiform transients of sleep (BETS), 132–33
Asymmetric reactivity, 153 Benign paroxysmal positioning vertigo (BPPV), 586, 587,
Asymmetry, 153, 170 607
Athetosis, 571 Benign rolandic epilepsy of childhood (BREC), 139–40
Atonia, REM sleep without, 717, 717f Benign sporadic sleep spikes (BSSS), 132–33
Atypical spike and wave discharges, 142–43 characteristics, 134t
Audiogram, 296–98, 297f, 303 Benign temporal slow wave transients, 126–27
applications, 299t, 303 Bereitschaftspotential (BP), 229, 230, 568
purpose and role, 296 early vs. late, 230
Auditory acuity, 287 Beta activity, 124–25, 125f
Auditory anatomy and physiology, 282f, 283 Beta-frequency coma pattern, 163
Auditory brain stem response (ABR), 309 Bilateral periodic lateralized epileptiform discharges
Auditory evoked potentials (AEPs), 788 (BiPLEDs), 140
Auditory neuropathy/dys-synchrony, 302 Biphasic potentials, 41
Auditory pathways, 100. See also Brain stem auditory Biphasic waveforms, 48
evoked potentials Bipolar potentials, 39
neuroanatomy, 282f, 283 Bispectral index (BIS), 736
Auditory pattern recognition, 414 Bizarre repetitive potentials. See Complex repetitive
Autocorrelation analysis, 63–64, 207f discharges
Autocorrelation function (ACF), 64 Bleeding disorders, needle exams and, 409–10
Autoimmune myasthenia gravis. See Myasthenia gravis Blepharospasm, 569
Automated analysis of single MUPs, 464 Blink reflex, 530
Automated decomposition EMG (ADEMG), 464–65 applications, 531–35
Automated methods of analysis of spontaneous activity, methods of eliciting, 530–31, 531f
472
neuroanatomy, 530, 530f
Autonomic axons. See Axons
Blink reflex latencies, 817
Autonomic control of heart rate, 624
Blink reflex responses, factors affecting, 535
Autonomic failure, severity of, 649
Blocking, 477, 481–82
Autonomic laboratory evaluation, indications for, 630t
Blood gases, 714–15
Autonomic nervous system, 617. See also Sympathetic
Blood pressure (BP), 637, 643
function
general organization, 619f, 619–21 intraoperative, 758, 764
symptoms and diseases, 617–19 Blood pressure recovery time (PRT), 639–40
Autonomic testing, 618, 630t Body position, 709
purpose and role, 618–19 Bottom tracing, 242f, 467f
Autonomic tests, 681–83 Botulism, 255, 381, 488
Averaging, 246–47, 261–62. See also under Quantitative Bouton, terminal, 370
MUNE Brachial plexopathy, 817–23, 821t, 822
back, 62, 553 Brain abscess, 180
digital, 60–62 Brain death, 164–65, 276, 292
forward, 62 Brain stem, 99, 276. See also Intraoperative monitoring, of
jerk-locked, 230 cranial nerves
signal, 600 disorders, 300–301
Averaging MUNE, spike-triggered, 498–500, 499f Brain stem auditory evoked potentials (BAEPs), 44, 100,
Axonal degeneration, 803 281–82, 284f, 285f, 292–93, 310
Axonal neuropathies, 350–52, 351t, 813–14 anatomical localization and, 283f
Axonal sensorimotor peripheral neuropathies, etiologies applications, 288–92, 308–9
of, 814t electrodes, 306
Axon reflexes, 99f, 342–43. See also Quantitative factors affecting BAEP response, 286–87
sudomotor axon reflex test generators of, 283
Axons, 97 interpretation, 287, 306–8
intraoperative, 739–40, 745, 745f
Back averaging, 62, 553 methods, 284–86
Balance, 573–75, 607–8 recording, 284–86
Ballistic pattern, 554 stimulation, 284
Baroreflexes, arterial, 624–25 purpose and role of
Baroreflex indices, 643 in central disorders, 283
Baroreflex sensitivity, 667–68, 672 in peripheral acoustic disorders, 305
Basal ganglia, 517, 788, stimuli, 306
Index 871
Brain stem auditory evoked response (BAER). See Brain Cervical dystonia, 569
stem auditory evoked potentials Cervical radiculopathy, 806–7
Brain stem auditory pathways. See also Auditory pathways Cervical spine disease, 770–71
neuroanatomy, 282f, 283 Cervical spondylitic myelopathy, 271, 272f
Brain stem lesions, 274 Chiasmatic lesions, 317–18
intrinsic, 291–92 Childhood
Brain stimulation, deep, 31–32 benign rolandic epilepsy of, 139–40
Breach rhythm, 130, 131f, 132 slow lambdas of, 175
Breathing, deep. See under Heart rate response Children. See also under EEG
Breathing arm or hand, 437 drowsiness, 173–74
Bruxism, 717 epileptiform abnormalities, 176–78
Buildup (EEG), 120, 122 somatosensory evoked potentials, 276
Burst patterns of EMG waveforms, 415 tumors, 181
Burst suppression, 148, 149f Chorea, 571
Burst-suppression pattern, 163, 163f, 170 Chronic disorder, 805
Chronic inflammatory demyelinating polyradiculopathy
C reflex, 545 (CIDP), 354–55, 539f, 814, 816f
Calcium, role of extracellular, 78 Chrono-dispersion, 340
Calcium channels, voltage-gated, 74, 74t Circuit analysis, 7–9
Calcium spike, low-threshold, 88 Circuits. See also specific types of circuits
Caloric irrigation, 588, 589–93, 589–93f containing inductors and capacitors, 10–14
Capacitance, 7 ideal, 10
Capacitative properties, 41 Click sensation level, 284
Capacitors, 7 Closed fields, 35
rules for seats of, 8 Closely spaced potentials, 39
Cardiac rhythm, 715 Cognition, assessing impairment of, 792, 797
Cardiac surgery, 736 “Cogwheeling” pursuit, 586
Cardiopulmonary receptors, 625 Coherence function, 207
Cardiovagal function, tests of, 672. See also Heart rate Coils, 7. See also Inductors
response; Valsalva maneuver Cold hyperalgesia, 680, 680f, 681f
Cardiovagal scoring, 670 Collector (transistor), 17
Cardiovascular heart rate testing. See also Heart rate Coma patterns, 162f, 162–64
response BAEPS, 292
purpose and role, 661–62 Coma, prognosis in, 275–76
Cardiovascular reflexes, 624–26. See also Valsalva Combined sensory index (CSI), 357–59
maneuver Common mode, 18
Carotid endarterectomy Common mode rejection ratio (CMRR), 19
EEG changes during, 728f, 729f, 733–35 Complex MUPs, 459, 479
monitoring techniques during, 735–36 Complex regional pain syndrome (CRPS), 681–83
SSEP recording during, 735 Complex repetitive discharges (CRDs), 414, 429–30, 830f
Carotid stump pressure measurement, 735 disorders associated with, 429–30, 430t
Carpal tunnel syndrome (CTS), 253–54, 335, 808–9 Composite autonomic severity score (CASS), 670
defining the severity of, 358t Compound muscle action potentials (CMAPs), 327–28,
Martin-Gruber anastomosis in, 833–34 363, 381, 496
NCS abnormalities in, 358t, 358–59 axon reflexes (A waves), 342–43
Cathode–anode relationship, 332–33 CMAP changes in disease
Cauda equina, 772–73 mechanisms of conduction in myelinated fibers, 345
Cell membrane, 69–76 mechanisms of slow conduction in disease, 345–46
Central apnea, 712 pathophysiology, 344
Central commands, 624 disorders producing low motor CMAP without sensory
Central conduction time (CCT), 392 involvement, 812t
Central disorders, 649. See also Central nervous system F waves, 338, 341–42, 341t
(CNS) symptoms latency, 338–40f, 340–42, 341t
Central nervous system (CNS) excitability, H reflex and, measurements, 340–41
526 recording, 338–39
Central nervous system (CNS) symptoms, 648 repeater, 352
assessing, 791–92 stimulation, 339–40
with EEG, 793–800 general clinical applications, 328
identifying disease types, 793 intraoperative, 741, 745f, 747, 763–64, 766–67,
localization of disease, 792–93 778–79
prognosis, 793 measurements, 334
Central-temporal spikes, 176, 177f amplitude and area, 334, 496
Cerebellar tremor, 557 conduction velocity, 335–36, 336f
Cerebral cortex, 100–101 duration, 334, 335f
Cerebral infarction, EEG and, 154 latency, 334–35
872 Index
Compound muscle action potentials (CMAPs) Cranial reflexes, 529, 541. See also specific reflexes
(Continued) purpose and role, 529
normal values in CMAP recordings, 337–38 Creutzfeldt–Jakob disease (CJD), 160–61, 161f, 565
potential errors in, 336–37 Critical period (surgery), 753
peripheral nerve disorders and, 346–63 Cross-correlation analysis, 206
physiologic variables affecting Cross-spectral analysis, 207
age, 343–34 Cumulative amplitude, 468
temperature, 343 Current density, 6
recording Current sources, 38–41
location of recording electrode, 330f, 330–32, 331f Cutaneous nerve stimulation SEPs, 266
type of recording electrode, 329t, 329–30 Cutaneous silent period (CSP), 548
stimulating electrode
position of, 332–34 Deafness, 296
type of, 332 Decomposition-based quantitative EMG (DQEMG)
Computer-assisted quantitation of MUPs, 462–64, 463f MUNE, 500, 501f
Computerized dynamic posturography (CDP), 603–6 Decomposition quantitative EMG (DQEMG), 408, 464,
purpose and role, 603 470f, 500
Computerized rotary chair tests, 594–95, 594–98 Deep brain stimulation, 31–32
purpose and role, 595 Degenerative disorders, 179. See also specific disorders
Condensation, 287 Delay (SFEMG), 478
Conductance changes during action potentials, 83f, 83–84 Delirium, EEG evaluation of, 795
Conduction block, 346–48, 347f, 803, 815–16 Delta activity, 125
complete, 815 Demyelinating disease, 271–73, 290–91, 345–46, 394. See
partial, 816, 816t also Multiple sclerosis
Conduction slowing, 346–49 Demyelinating neuropathies, 814–17. See also
Conduction velocity (CV), 247, 248f, 335–36, 336f, 808, Inflammatory demyelinating polyradiculopathy
809, 814 inherited, 353t
Conductive hearing losses, 297, 305 predominantly, 815t
Conductivity, 41 segmental, 352t, 352–53. See also specific disorders
Conductors, 6 focal neuropathies, 355–62
Congenital abnormalities, 173 Demyelination, 242
Congenital myasthenia, 381 partial, 803
Congenital myasthenic syndromes, 488 Denervation supersensitivity, 642
Consciousness, assessing impairment of, 792, 795 Depolarization, 86
Contact heat evoked potential stimulator (CHEPS), Depolarization blockade, 94
540–41, 688–89 Dermatomal SEPs, 266
Continuous positive airway pressure (CPAP), 711f, 712, Diabetic peripheral neuropathies, 648
713, 714f Diagnosis. See also under Disease
Continuous waves, 109–10 confirming a clinical diagnosis, 802
measurable variables of, 109t differential, 802
Contraction fasciculations, 431 Dielectric constant, 41
Contralateral preponderance of negativity (CPN), 230 Differential diagnosis, 802
Cortex, cerebral, 100–101 Diffusion pressure, 71
Cortically generated potentials. See also Digital averaging, 60–62
Movement-related cortical potentials operation of an averager, 60f
in volume conductors, 34–35 Digital clinical neurophysiology, 54
Cortical malformations, 172 Digital computers, utility of, 53
Cortical origin myoclonus. See also Myoclonus Digital electroencephalography (EEG), 54–56
without reflex activation, 565 inter-reader agreement in classification of EEG
Cortical projection techniques, 210 records, 56t
Cortical reflex myoclonus, 544, 545f, 563–65 Digital filtering, example of, 64f
Cortical silent period, 547 Digital filters
Cortical stimulation, therapeutic, 31–32 characteristics, 63
Corticobasal degeneration, myoclonus of, 566 types of, 62–63
Cough test, 672 Digital recording technology
Coupling discharges, 459. See also Satellite potentials capabilities, 54
Cramp discharges, 435–36, 436f disadvantages, 54
disorders associated with, 436t Digital signal processing
Cramp fasciculation, 435 time and frequency domain analysis, 63–67
Cranial fossa, intraoperative monitoring of, 743–44, 745f, uses, 59–60
746–47 Digital systems, construction of, 56
Cranial nerve (CN) VII, disorders of, 300 Digitization, 56–59
Cranial nerve (CN) VIII, 299t, 299–300, 305 principles, 56–57
Cranial nerve (CN) VIII tumors, 308f, 308–9, 309f, 310f Diodes, 16–17
Cranial nerves (CN), 620 Dipoles, 38, 39f, 50, 51
Index 873
Dipole source localization, 46–47, 211–12 drowsiness and, 173–74

Direct current (DC) circuit, 12 neonatal EEG patterns, 168–73
Directional preponderance (DP), 591 movement disorders and, 553
Direction-changing nystagmus, 588 normal EEG activity of adults
Direction-fixed positional nystagmus, 588 awake state, 124–26
Disappearance threshold, 678 older adults, 126–27, 129–30
Discharge (event), 107 sleep state, 127–30
Discrete firing, 419 volume conduction and EEG applications, 45–46
Disease volume conductor resistive-capacitive (RC) properties
defining stage and evolution of, 804–5 and, 41
identifying. See also Diagnosis waveforms and, 103
in patients who are difficult to examine, 802 EEG analysis, quantitative methods of, 203, 212. See also
identifying subclinical, 802–3 EEG special studies; Fourier (spectral) analysis
localizing, 792, 802 cortical projection techniques, 210
Disease pathophysiology, characterizing, 803–4 cross-correlation analysis, 206
Disease severity, defining, 804 cross-spectral analysis, 207
Disease types, identifying, 805–6 interpolation, 207–8
Display times of common signals, 110t montage reformatting, 205–6
Distal anhidrosis, 654 multivariate statistical methods and topographic
Distal latency (DL), 808 analysis, 209–10
prolongation of, 815 source dipole localization, 210–11
Distal motor latency (DML), 335 spike, sharp-wave, high-frequency oscillation and
Distant potentials, 39 seizure detection, 204–5
Distant reference, 39, 41, 45, 46f topographic displays (mapping), 208–9, 209f
Distortion product evoked otoacoustic emissions EEG-EMG polygraphy with back-averaging, 553
(DPOAEs), 301, 302f EEG manifestations of diffuse disorders, 159–64
Diving reflex, 672 coma patterns, 162–64
generalized periodic patterns, 160–62
Dix–Hallpike test, 586, 587f
specific patterns, 160
Doping, 16
EEG special studies. See also EEG analysis
Dorsal column entry (DCV), 264
purpose and role, 203
Dorsal rhizotomy, 772
Effusion, subdural, 180–81
Drowsiness. See also under EEG
E:I ratio, 662
in children, 173–74
Electric artifact, 267
Drugs and EEG, 171
Electric charges and force, 5–6, 8
Drug withdrawal, 171
Electric current, 6–7
Duration, 456–57, 477, 482 physiologic effect, 23, 24f
Dying-back neuropathies, 346 requirements for it to flow through body, 22–23
Dynamic compensation, 581 Electric potential. See Electrical potentials
Dysgenetic disorders, 172 Electric power distribution systems, 21–22
Dystonia, 445, 569–71 Electric safety. See also Electric stimulation safety
recording techniques, 569 principles and implementation, 27–30
“Dystonic myoclonus” of SSPE, 565 procedures for technicians, 30
rules, 29–30
EEG (electroencephalography), 18, 117, 119, 136, 151, Electric shock, 22–23
165. See also specific topics in hospitals, 24–25
abnormalities, 175–83. See also EEG manifestations of risk of
diffuse disorders factors increasing, 24–25
focal intracranial processes causing, 154–59 factors reducing, 23
types of, 152–54 Electric stimulation safety, 30–32. See also Electric safety
activation procedures, 120, 122–23, 123f Electrical potentials, 6. See also specific topics
artifacts, 123–24 calculating in infinite homogeneous media, 50
assessing clinical CNS disorders with, 793–800 in nonhomogeneous media, 50–51
benign variants, 175 sources, 34–36
during drowsiness and sleep, 127, 132–34, 135f, 136, spatial distribution, 39, 40f, 41
173–75 Electrical pressure, 71
during wakefulness, 130, 131f, 132 Electrical surface stimulator, handheld, 332
dipole source localization in, 46–47 Electricity, basic principles and definitions in, 5–7
display of EEG activity, 119–20 Electrocerebral inactivity (ECI), 164f, 164–65
evaluation for suspected brain death, 164–65 Electrocorticography (ECoG), 223, 224f, 225
of infants and children, 175–83 Electrodecremental pattern, 143
conditions giving rise to abnormal EEG patterns, Electroencephalogram (EEG), 119, 120–23f. See also
171–73, 178–82 EEG
developmental changes during infancy, childhood, recording the, 119
and adolescence, 173 Electroencephalographic analysis. See EEG analysis
874 Index
Electromagnetic tomography, low resolution, 211 Epileptiform patterns, generalized, 141–45

Electromagnets, 7. See also Inductors Epileptiform transients of sleep, benign, 132–33
Electromotive force (EMF), seat of, 6 Episodic myalgia, 827
rules for, 8 Episodic weakness, 827
Electronystagmography (ENG). See ENG/VNG test Epochs, 706
battery Equilibrium potential, 71–72
Electrooculogram recording (EOG), 582, 583f Equipment ground, 25
Electrophysiologic generators, 97–101 Equipment grounding, 27–28
Electroretinogram, 100 tests for, 28–29
EMF. See Electromotive force; Evoked magnetic fields Equipotential grounding system, 26
EMG (electromyography), 18. See also Needle EMG; Equipotential lines, 38, 39f
specific topics Ergoreflexes, 625
intraoperative, 740, 779. See also under Intraoperative Esophageal pH measurements, 718
spinal cord monitoring Essential myoclonus, 559, 567
semiquantitative, 415, 452 Essential tremor, 555–56, 557f
volume conductor resistive-capacitive (RC) properties Event recording, waveforms and, 106–8
and, 41–42 Event-related potentials, 232–33
EMG potentials, origin of, 415–17 Evoked magnetic fields (EMFs), 47
EMG waveforms, 103 Evoked otoacoustic emissions (EOAEs), 295–96, 301–3
burst patterns, 415 purpose and role, 296
origin, 417t Evoked potentials. See also specific topics
Emitter (transistor), 17 digital averaging devices for, 60–62
Empyemas, subdural, 180–81 intraoperative, 742–43
Encephalitis, 179–80. See also Panencephalitis Ewald’s second law, 578
herpes simplex, 158–59, 180 Exaggerated physiologic tremor, 555
End plate activity, 407, 417–18 Exaggerated startle, 568
End plate noise, 407, 417 Excessive daytime somnolence, 719
End plate potential (EPP), 371. See also Miniature end Excitatory postsynaptic potential (EPSP), 34–35, 92
plate potentials Exponential change, 414–15
End plate spikes, 98 Extrapyramidal disease, 535
End plate zone, 453
End-tidal anesthetic (ETAG), 736 F waves, 506, 521t, 807, 817, 829. See also under
Energy failure, 93–94 Compound muscle action potentials
ENG/VNG test battery, 589, 594t Facial nerve lesions, 532f, 532–33
purpose and role, 584–85 Facial nerve response, 740f
Entrapment neuropathies, 780–81 lateral spread, 535–36, 536f
Ephaptic activation, 485 Facial synkinesis, 437
Epilepsy monitoring unit (EMU), video-EEG monitoring assessment, 533, 533f, 535–36
in, 217–18, 221, 799–800 Facial weakness, 831
Epilepsy, surgical evaluation of, 225. See also under Facilitation, 372
Seizures postactivation, 374
background, 215–16 False double, 485
preoperative video-EEG monitoring, 218–19 Familial tremor, 555–56
presurgical evaluation with continuous or chronic Far-field potential (FFP), 43, 261, 284
intracranial monitoring, 219 Fasciculation, cramp, 435
depth wire electrodes, 219–21 Fasciculation potentials, 430–31
subdural electrode monitoring, 221–22 disorders associated with, 431t
presurgical selection and evaluation Fast alpha variant, 130
clinical evaluation, 216 Fast spike-and-wave, 134
imaging studies, 216–17 Feedback data flow, 63
intracarotid amobarbital (ICA), 217f, 217–18 Feed-forward data flow, 63
purpose and role of EEG in, 216 Fiber density, 422, 453, 475, 479
Epilepsy surgery. See also Epilepsy, surgical evaluation of Fibrillation potentials, 416, 425–28, 426f, 427f, 803, 826
and quality of life (QOL), 215, 216 diseases associated with, 427t
Epileptic myoclonus, 559 Filter circuits, 14f, 14–16
Epileptic vs. nonepileptic events, PVEEG and, 195–96 Filters
Epileptiform abnormalities in children, 176–78 high-pass, 14f, 14–15
Epileptiform activity, 153, 170 low-pass, 15–16
definitions and overview, 137–38 Finite impulse response (FIR), 63
with a potential seizure association, 148, 170, 171f FIRDA (frontal intermittent rhythmic delta activity)
Epileptiform discharges. See also Periodic lateralized pattern, 152, 731
epileptiform discharges First dorsal interosseus (FDI), 809
ictal, 145, 147 First-night effect, 710
interictal, 796–97 Flat disc electrodes, 332
specific focal, 138–40 Flexor carpi radialis technique, 522
Index 875
Flexor reflex afferents, 546 Head tilts, 579. See also Head-up tilt
Flexor reflexes, 546 Head trauma, 181, 275–76
Focal cerebral lesions, 159 and EEG, 157, 157f
Focal delta slowing, 180, 180f Head-up tilt (HUT), 638. See also Head tilts
Focal intracranial lesions, EEG in, 154 Hearing impairment, 296, 305. See also Brain stem
purpose and role of, 154 auditory evoked potentials
Focal intracranial processes causing EEG abnormalities, Hearing level (HL), 296. See also Brain stem auditory
154–59 evoked potentials
Focal motor seizures, 565 Heart period range, 662–63
Focal myoclonus, 559 Heart rate range, 662
Force, 5–6, 8 Heart rate response, 662. See also Cardiovagal function
Forward averaging, 62 to deep breathing
Forward biasing, 16–17 factors affecting, 663–64
Forward problem, 47 methods of analysis, 662–63, 663f
Fossa normative data, 664t
cranial, 743–44, 745f, 744–47 physiologic basis, 662
posterior, 534, 745f, 744–47 problems and controversies, 664–65
4, 2, and 1 stepping algorithm, 678 reproducibility, 663
Fourier (spectral) analysis, 64–66, 203–4 technique, 662
Frequency of signals, 103–6, 105f, 110t to standing, 671
Frequent stimulus, 233 Heat hyperalgesia, 680, 680f, 681f
Friedreich’s ataxia, 274, 274f Hemiconvulsions, hemiplegia, and epilepsy syndrome
Frontal intermittent rhythmic delta activity (FIRDA) (HHE), 181
pattern, 152, 731 Hemifacial spasm, assessment of, 535–36, 536f
Frontal lobe seizures, 198 Hemiplegia, 181
Frontal spikes, 138 Hemorrhage
Full recruitment, 419 cerebral, 276
Fundamental frequency, 64 intracerebral, 155
F-wave measurements, 503f, 506, 507f intraventricular, 171
Hereditary hyperekplexia, 568
Ganglionopathies, 813t
Gastrocnemius technique, 522
(HNPP), 349
Gaze nystagmus, 581–82
Herpes simplex encephalitis, 158–59, 180
Gaze testing, 588
Hertz, 103. See also Frequency of signals
Gender differences, 286–87, 315–16
High-frequency potentials, 429
Generalized dystonia, 558
High safety factor, 372
Generalized myoclonus, 559
Hmax/Mmax (ratio of maximal H-reflex to M-response
Generalized slow spike-and-wave, 176, 177f
amplitude), 526
Generalized spike-and-wave, 3-Hz, 176
Hoffman reflexes. See H reflex(es)
Generators
electrophysiologic, 97–101 Holes, 16
and origin of SEPs, 258 Holmes tremor, 557
Glial cells, role of, 78 Homogeneous sphere model, 51
Glioma, brain stem, 291f Horn cell disorders, anterior, 489–90
Global anhidrosis, 654 Hot line, 21
Global syndrome, 608–9 H-reflex latency, normal values for, 524t
G-protein-coupled receptors, 91, 93 Hydrocephalus, 181–82, 182f
Gray matter disease, 179 Hyperalgesia, 680, 680f, 681f
Great auricular sensory nerve conduction studies, 540 Hyperhidrotic disorders, 645–46t, 647–48
Ground, 22, 25 Hyperpolarization, 84, 88, 332. See also Anodal block
Ground potential, 19 Hyperventilation, 120, 122, 123f, 173, 176f
Guillain–Barré syndrome (AIDP), 269, 270f, 343, 353–54 Hypopnea, 712
Hypsarrhythmia, 143–44, 144f, 176, 176f
H reflex(es), 517, 524, 807
clinical applications, 524–26 Ictal discharges, 145, 147, 148, 149f
vs. F wave, 521t Imbalance. See Balance
factors affecting the presence or amplitude of, 521t Immittance unit, 298, 298f
pediatric, 524 Immobilization, 373
physiologic basis, 519–21, 520f Immobilization test, suggested, 716
purpose and role, 519 Impedance, calculation of, 13–14
technique, 521–24 Implanted electrical devices, stimulating near, 30
Handheld electrical surface stimulator, 332 Imprint methods of sweat measurement, 634
Hard of hearing, 296 Inactive reference electrodes, 45, 331
Harmonics, 64 Inching, 349, 816f
Head and neck surgery, intraoperative monitoring during, Inductance, 7
747 Inductive-capacitive (LC) circuits, 10–11, 11f
876 Index
Inductive-resistive-capacitive circuits, 11–12 methods, 778–80

Inductors, 7 SEPs and MEPs, 779–80
rules for seats of, 8 prevention of injury during surgery, 783–84
Infants, 798–99. See also under EEG purpose and role, 777–78
conditions giving rise to abnormal EEG patterns in, Intraoperative spinal cord monitoring, 751–52, 774. See
171–73 also Spinal surgeries
Infection precautions in needle exams, 410 EMG and nerve conduction studies, 765–68, 766f
Infectious diseases, 171 applications and interpretation of findings, 767–68
Infinite impulse response (IIR), 63 physiologic and technical effects, 767
Inflammatory demyelinating polyradiculopathy. See also recording techniques, 766–67
Guillain-Barré syndrome general principles, 752–53
chronic, 354–55, 534f, 814, 815t motor evoked potentials (MEPs)
Inflammatory disorders, 158 application and interpretation of MEP changes,
Inhibitory postsynaptic potential (IPSP), 34, 35, 92 764–63
deep vs. superficial, 35 physiologic/technical considerations, 764
Innervation, 422 recording techniques, 762–64, 763f
anomalous, 831–35 stimulation techniques, 760–62
Innervation ratio, 453 purpose and role, 752
Insertional activity, 407, 424–25, 425f somatosensory evoked potentials (SEPs), 754
Instrument ground, 25 application and interpretation of SEP changes, 759f,
Insufficient sleep syndrome, 719 759–60, 760f
Intensive care unit (ICU) physiologic and technical issues, 756–59
EEG of, 797–98 recording techniques, 755–56, 756f
SEPs recorded in, 276–77 stimulation techniques, 754–55
Intention tremors, 555 Intraventricular hemorrhage, 171
Interaural latency differences, 307 Inverse problem, 47, 210
Interference, 113, 665. See also Artifacts Inward current flow, 34
common forms of, 112t Ion channel blockade, 94, 96
Interference pattern analysis (IPA), 408 Ion channels, 72–75, 90–91
amplitude analysis, 469–71, 470f categories of, 73–75
analysis, 468 examples of, 74t
method, 467–68 Isoelectric EEG, 170
utility, 468 Isoflurane anesthesia, 728f
Interference patterns, 107, 408, 467, 495 Isolated acquired idiopathic anhidrosis, 646t
properties, 467 Isolated voice tremor, 557
Interictal discharges, 138–40, 796 Isolation transformers, 26, 27f
Interictal epileptiform discharges (IEDs), 796 Isometric tremors, 555
Intermittent waveforms, 103
Jaw jerk
Interpeak intervals, 307–8
applications, 537–38
Interpolation, 207–8
methods of eliciting, 537
Interval analysis, 63 neuroanatomy, 537
Intracarotid amobarbital (ICA), 217f, 217–18 Jerk-locked averaging, 230
Intracerebral hemorrhage, EEG and, 155 Jerks, 788
Intracranial monitoring. See under Epilepsy, surgical psychogenic, 568
evaluation of Jiggle, 481
Intracranial processes, focal, 154–59 Jitter, 476–77, 479–81, 484f, 484–86, 485f
Intraepidermal fibers (IEF), 688 Junctional folds, 370
Intraoperative EEG monitoring, 737, 798
during cardiac surgery, 736 K-complex, 128
EEG changes during carotid endarterectomy, 729f, Kinetic tremors, simple, 555
730f, 733–35 Kirchhoff’s first law, 8
for epilepsy surgery, 736–37 Kirchhoff’s second law, 8–9
technical factors in, 727–31, 729f, 730
Intraoperative monitoring (IOM), 292, 393–94, 725, 737. Labyrinthitis, 607
See also Anesthesia Lambda wave, 126, 127f
clinical applications, 733 Lambert-Eaton myasthenic syndrome, 378f, 380, 381, 488
of cranial nerves, 739, 748 Landau-Kleffner syndrome, 182, 183f
applications, 743–44, 744f, 746–47 Laser evoked potentials (LEPs), 684–88
cranial nerve modalities used, 740t Lateral cord, 819, 820t
methods, 740–43 Lateral spread response (LSR), 363, 535–36, 536f,
monitoring plans of common surgical procedures, 744t 745–47, 747f
Intraoperative nerve conduction studies (NCSs), 741–42 Laterocollis, 570
Intraoperative peripheral nervous system monitoring, Leakage current, 24f, 26f, 27f
777–78 origin, 25
applications, 780–81 tests for, 28–29
Index 877
Leakage current reaching patients, 25 Martin-Gruber anastomosis (MGA), 831

methods to reduce, 25–27 type I, 831, 832t, 832f, 834–35
Learning set, 66 type II, 832–34, 833f, 833t, 834f
Lennox-Gastaut syndrome, 177f in carpal tunnel syndrome, 833–34
Lesions type III, 834t, 834–35
brain stem, 274, 291–92 Masseter inhibitory reflex (MIR), 538
chiasmatic, 317–18 applications, 539–40
facial nerve, 532f, 532–33 methods of eliciting, 538–39
focal cerebral, 159 Masseter reflex. See Jaw jerk
focal intracranial, 154 Maximal voluntary contraction (MVC), 467
posterior fossa, 534 Mayo Clinic, v
prechiasmatic, 317–18 brachial plexus outpatient EMG protocol, 823t
retrochiasmatic, 321–22 Thermoregulatory Laboratory, 650, 651f
root vs. plexus, 822 Mean amplitude, 465
spinal cord, 273–74 Mean consecutive difference (MCD), 478, 480–81
supratentorial, 274 Mean interspike interval, 482
trigeminal nerve, 531–32, 532f Mean sorted difference (MSD), 480
trunk, 818–19, 821t, 822 Medial cord, 819, 821t
vascular, 181 Median and ulnar mixed nerve SEPs, 263–65
visual pathway, 317–22 Median nerve action potentials, 42f
visual system, 317–22 Median neuropathies, 357–59, 359f
Lesions effects of VEMPs, 603 Membrane potentials. See also Cell membrane; Resting
Leukodystrophy, metachromatic, 179f potential
Liability to pressure palsies, 349 characteristics of different, 73t
Ligand-gated receptors, 90–91 generated by diffusion across semipermeable
Limits, method of, 678 membrane, 77f
Linear change, 414 variables that determine, 72f
Linear conductors, 6 Meningitis, 179
Linear display (digital EEG), 55 Mental retardation, 177f
Linear translations and accelerations, 579–80 Meralgia paresthetica, 270f
Metabolic disorders, 171
Linked potentials, 441, 459. See also Satellite potentials
Metabolism, inborn errors of, 171–72
Local potentials, 78–79, 80f
Metachromatic leukodystrophy, 179f
characteristics, 79–82
Microneurography (MCNG), 683–84
ionic basis, 76t, 79
Midbrain tremor, 557
temporal course, 81f
Middle cranial fossa, intraoperative monitoring of, 743–44
Localization
Migraine headache, 159, 183, 183f
of disease, 788–89, 802
Miniature end plate potentials (MEPP), 98, 371, 417
paradoxical, 266, 314
Mitten patterns, 136
Localizing nerve with stimulator (sliding), 333
Mononeuropathies
Long latency reflexes (LLRs), 543–44, 548
common, 253–55
cutaneous reflexes, 546 common focal, 808–10
flexor reflex, 546 Monopolar electrode, 423
to mixed nerve stimulation, 545–46 Monopoles, 38, 39f, 50
purpose and role, 543 Montage reformatting, 55, 205–6
to stretch, 544–45 Motor control test (MCT), 604
Low resolution electromagnetic tomography (LORETA), Motor evoked potentials (MEPs), 325, 385–86, 396
211 anesthesia and, 764
Lower motor neurons (LMNs), 453 applications, 393–95
Lower tracing, 254f contraindications and risks, 395
LSU, 579–80 intraoperative, 741, 779–80. See also under
Lumbosacral radiculopathy, 806–7 Intraoperative spinal cord monitoring
Lumbosacral spine disease, 772 pharmacology, 392–93
Lymphedema, needle exams and, 412 purpose and role, 386
technique
M wave, 334, 339f measurements, 392
Magnetic fields. See also Electromagnets recording, 390, 392
evoked, 47 stimulation, 386–87, 387f
Magnetic stimulation, 332 technical aspects of electrical stimulation, 387–89
Magnetoencephalogram (MEG), 47 technical aspects of magnetic stimulation, 389–90
Magnetoencephalography (MEG), 211–12 Motor nerve conduction studies. See Nerve conduction
Magnitude estimation, 678 studies
Maintenance of wakefulness test (MWT), 703–4 Motor neuron diseases, 255–56, 274–75, 830–31
purpose and role, 698 evolution of EMG changes in, 830t
Mapping, topographic. See Topographic displays Motor neurons, lower, 453
878 Index
Motor potential (MP), 229, 230 Multimodality evoked potential (MMEP), 275–76
Motor symptoms of central origin, assessment of, 791–92 Multi-motor unit action potential analysis (multi-MUAP
Motor unit action potential. See Motor unit potentials analysis), 465–66
Motor unit fractions, 459 Multiplanar and multiple sphere models, 51
Motor unit number estimates (MUNEs), 421, 493–94. See Multiple sclerosis (MS)
also Quantitative MUNE BAEPs and, 288f, 290f, 290–91
from all-or-none increments in CMAP, 500–1, 501f, blink reflex and, 535
503–4 somatosensory evoked potentials and, 272–73
from multiple all-or-none increments at one stimulation visual evoked potentials in, 318–20, 320f
point, 501, 502f, 503–4 Multiple sleep latency test (MSLT), 703, 718
purpose and role, 493 purpose and role, 698
by standard EMG, 495 Multiple system atrophy (MSA), 649
by standard motor NCS, 496–97 Multipoint stimulation (MPS), 504–6, 505f, 506f
statistical (STAT), 506–9, 509f, 510f Multivariate statistical methods, 209–10
Motor unit potential (MUP) variation, 831 Mu rhythm, 126, 126f
vs. stability, 459–60 Muscle activity, intraoperative, 757, 764
Motor unit potentials (MUPs), 36, 406–8, 416, 418, 494, Muscle artifact, 267
740. See also Needle EMG; Quantitative EMG Muscle end plate potentials, 98
area, 457–58 Muscle relaxation, importance of, 262f
automated analysis of single, 464 Muscles, 98. See also Compound muscle action potentials;
characteristics, 453–54 Needle EMG
commonly measured variables, 457f data collection from contracting, 407–8
complexity, 459, 479 data collection of resting, 407
doublets and multiplets, 444, 444f peri-pleural, 410
disorders associated with, 444t primary disorders of, 488–89
long-duration, 438–40f, 441 Muscle stiffness, 827
disorders associated with, 439t Myalgias, 827
manual analysis, 462 Myasthenia, 381
measurement, 408–9 familial infantile, 381
mixed patterns (long- and short-duration), 442 Myasthenia gravis, 380, 486–88
myopathy and, 827 Myasthenic syndromes, congenital, 381, 488
phases, 440–42, 458–59 Myelinated fibers. See also Demyelinating disease
properties
mechanisms of conduction in, 345
evaluated using standard electrodes, 456–60
Myelopathy, 320f, 650f
measurable only with special electrodes, 460–62
cervical spondylitic, 271, 272f
recruitment, 418
Myoclonus, 274, 275f, 559–62
short-duration, 439–40
abnormal patterns, 562–63
disorders associated with, 440t
classification by localization and electrophysiologic
terminal component, 457, 458
features, 560–63t
varying/unstable, 442–44
cortical, 544, 545f, 563–65
disorders associated with, 443t
cortical-subcortical, 566
Motor units, 401, 494
palatal, 558
Motor unit territory, 453
proportion functioning distal to the block, 496 peripheral, 568
Movement abnormalities, voluntary, 571 recording techniques, 562
Movement-associated potentials, 62 segmental, 559, 567–68
Movement disorders, clinical neurophysiology of, 552, subcortical-suprasegmental, 567
568. See also Myoclonus; specific disorders Myokymic discharges, 432–33, 432f
abnormal patterns, 570–71 disorders associated with, 433t
elicited responses, 553–54 Myopathic diseases, 493
normal patterns, 554 Myopathy(ies), 255, 823–27
purpose and role, 552 associated with fibrillation potentials, 826t
techniques, 552–53 associated with myotonic discharges, 826t
Movement-related cortical potentials (MRCPs), 229 disorders causing, 825t
abnormalities in disease, 231–32 with low CMAP amplitudes on distal NCS, 824t
contingent negative variation, 232 Myorhythmia, 557
individual variation, 231 Myotonic discharges, 428–29, 429t, 826. See also
normal waveforms, 230–31, 231f Neuromyotonic discharges
purpose and role, 229 diseases associated with, 429t
technique, 230 Myotonic disorders, 381
M-response amplitude, 526
Multichannel surface EMG recordings, 788 Nascent MUPs, 440, 804
Multifocal motor neuropathy (MMN) with conduction Near-field potentials (NFPs), 43, 261, 284
block, 355 Near-infrared spectrophotometry (NIRS), 736
Multifocal myoclonus, 559 Neck surgery, intraoperative monitoring during, 747
Index 879
Needle electrodes, 332, 477f, 477–78 sensory, 787. See also Sensory nerve action potentials
types used in quantitative EMG, 454f, 454–55 great auricular, 540
Needle EMG (electromyography), 404, 445–47. See also unexpected findings, 835–36
specific disorders volume conductor resistive-capacitive (RC) properties
abnormal electrical activity and, 41–42
disorders of central control, 445 Nerve conduction variables. See also specific variables
voluntary MUPs, 437–44 changes after focal nerve injury, 352, 352t
abnormal recruitment, 437–38 Nerve stimulation
abnormal spontaneous electric activity, 424 in SEPs, 258–60, 259f
complex repetitive discharges, 429–30, 430t unilateral vs. bilateral, 259–60
cramp potentials, 435–36, 436f, 436t Neuralgic amyotrophy, 363
fasciculation potentials, 430–31, 431t Neurochemical transmitters
fibrillation potentials, 425–28, 426f, 427t, 427f biosynthesis, storage, release, and reuptake, 89–99
insertional activity, 424–25, 425f postsynaptic effects, 90–91
myokymic discharges, 432–33, 432f, 433t Neurocutaneous disorders, 172
myotonic discharges, 428–29, 429t Neurogenic blocking, 485f, 485–86
neuromyotonic discharges, 433–35, 434t Neurogenic disorders, 490. See also specific disorders
synkinesis, 436–37 primary, 489
clinical evaluation, 405 Neurogenic motor evoked potential (MEP), 390
conducting needle examination, 405 Neurolabyrinthitis, 581, 607
data collection of contracting muscles, 407–8 Neuroma, acoustic, 289f, 289–90
data collection of resting muscles, 407 Neuromodulation, 93
muscle selection, 405–6 vs. classic neurotransmission, 91t
needle insertion, 406 Neuromuscular blockade and motor evoked potentials
needle movement, 406 and, 764
preparing patient, 405 Neuromuscular junction (NMJ), 823–24
recording display during examination, 407 anatomy and physiology, 370f, 370–72
knowledge base of, 404–5 disorders, 255, 828–29
needle size, 411, 411t Neuromuscular transmission, primary disorders of, 486.
normal EMG activity See also specific disorders
duration and amplitude, 423 Neuromyotonic discharges (neuromyotonia), 433f,
firing rate and recruitment of MUPs, 418–19, 420f, 433–35, 434f
421 disorders associated with, 434t
MUP configuration, 421f, 421–22, 422f Neuronal excitability, 75–76
normal spontaneous activity, 417–18 Neuronopathies, sensory, 813t
normal voluntary activity, 418 Neurons, current flow near
phases, 423 caused by synaptic activation, 34–35, 35f
stability, 423 Neuropathies. See also Demyelinating neuropathies;
origin of EMG potentials, 415–17 Peripheral neuropathies; specific neuropathies
pattern recognition, 414f, 414–15 axonal, 351–52, 352t, 813–14, 814t
patterns of abnormalities, 445, 445f, 446t disorders causing, 825t
potential complications during examination, focal, 355–62
409–14 evaluation of, 349–50
purpose and role, 404 median, 357–59, 359f
recording display during examination, 407 primary autonomic, 646t
semiquantitative EMG, 415 Neurophysiology, 96, 836
steps in evaluation process, 405 clinical correlations, 93–96
technique, 405 Neurotonic discharges, 434, 435f, 740–41, 740f
Needle EMG applications of volume conduction, 49 Neurotransmission. See also Neurochemical transmitters
Needle stimulator position, 333–34 classic, 92
Negative afterpotential, 84 vs. neuromodulation, 91t
Negative slope (NS), 230 Newborn, EEG in the, 798–799
Neonatal EEG patterns Nocturnal myoclonus, 569
abnormal, 169–73 Nodes, 8
normal, 168–69 Noise, 248–49. See also Signal-to-noise ratio
Neonatal screening, evoked otoacoustic emissions tests electrical, during intraoperative monitoring, 756–57
for, 301–2 end plate, 407, 417
Nernst equation, 71 Non-rapid eye movement (NREM) sleep, 697–98, 706,
Nerve action potentials (NAPs), 741, 742, 778 709f
Nerve conduction studies (NCSs). See also specific topics Norepinephrine, plasma, 642–43
digital averaging devices for, 60–62 Normal cloud, 470
intraoperative, 741–42, 778–81 Normal deviates, 338
motor, 787–88 Normotension, maintenance postural, 626
purpose and role of, 328–29 NREM sleep, 697–98, 706, 709f
NCS applications of volume conduction, 48 N-type semiconductor, 16
880 Index
Nylen maneuver, 586, 587f Paroxysmal disorders, EEG evaluation of, 796–97
Nystagmus Paroxysmal positioning vertigo, benign, 586, 587, 607
positioning induced, 586–88 Paroxysmal rhythmic fast activity, 143
testing for pathologic, 586–89 Parsonage-Turner syndrome, 363
types of, 583. See also specific types Partial conduction block, 816, 816t
Passive sources (of bioelectric potentials), 33, 34
O waves, 175 Pathophysiologic mechanisms, 93
Obese patients, 410–11 Patient ground, 25
Obstructive sleep apnea (OSA), 697–98, 707–8, 711f, 715, Pattern recognition, 66, 414
719 Pattern recognition algorithm, developing a, 66
Occipital intermittent rhythmic delta activity (OIRDA), Peak latency, 247
152, 178 Peak ratio, 468
Occipital spikes and seizures, 176–78, 178f People with epilepsy (PWE), 215. See also Epilepsy,
Occupational cramp, 569, 571 surgical evaluation of
Oculomotor nerve, stimulation of, 744f Periodic alternating nystagmus, 588
Oddball stimulus, 233 Periodic discharges, 148, 149f
Oddball technique, 233 Periodic lateralized epileptiform discharges (PLEDs),
Ohms, 7 140, 141f, 153, 157f, 170
Ohm’s law, 7 herpes simplex encephalitis and, 158–59
OIRDA. See Occipital intermittent rhythmic delta activity, Periodic limb movements of sleep (PLMS), 568, 702,
152, 178 714–17, 716f
Onset latency, 247, 248f Periodic paralysis, 382, 382f
Open fields, 35 Periodic patterns, 153–54
Operating room (OR). See also Intraoperative monitoring Peripheral nerve damage, diffuse, 350–52
equipment and electrical safety, 753–54 Peripheral nerve disorders. See also Peripheral nervous
Opsoclonus-myoclonus syndrome, 567 system disorders; specific disorders
Optic neuritis, acute, 319 CMAP findings in, 346–52
Optic pathways, 100
Peripheral nerve injury
Optokinetic nystagmus (OKN), 586
CMAP after, 356t, 804t
Oromandibular dystonia, 569, 570
duration of deficit after, 347t
Orthodromic technique, 243, 244f, 245, 253–54
EMG interpretations after, 357t
Orthostatic technique, 253–54. See also Orthodromic
needle examination findings after, 356t, 804t
technique
repair, 781–83
Orthostatic tremor, 558, 558f
Peripheral nerve stimulation methods, 497–98. See also
Otoacoustic emissions, evoked, 295–96, 301f, 301–3
specific methods
Otoacoustic emissions tests, evoked, 301–2
Peripheral nerves, 97–98
Otoconia, 579
Outward current flow, 34 Peripheral nervous system, monitoring. See Intraoperative
Overshoot, 666 peripheral nervous system monitoring
Peripheral nervous system disorders. See also Peripheral
P300, 233 nerve disorders; specific disorders
Pacemakers clinical neurophysiology in assessment of, 802–6
examining patients with, 410 assessment with EMG and NCS, 806
stimulating near, 30 Peripheral neuromuscular disorders, abnormal NCS
Pachygyria, 172f patterns in, 348t
Pain hyperalgesia, thermal, 680f, 681f Peripheral neuropathies, 350–52, 351t, 489, 533–34,
Pain pathways, techniques to evaluate. See also specific 810–13. See also specific neuropathies
techniques defining pathophysiology, 813
purpose and role, 677–78 Peripheral silent period, 547
Pain syndrome, complex regional, 681 Peripheral stimulation failure during spine surgery, 758f
Pain tolerance, low, 411 Peri-pleural muscles, examining, 410
Palatal tremor, 558 Permeability of membranes, 72
Palmar latency, 357 Peroneal motor conduction study, 507f
Palmar technique, 253–54 Peroneal nerve, accessory, 835f, 835, 836t
Panencephalitis, subacute sclerosing, 161, 162f, 180, Peroneal neuropathies, 362, 361f, 809–10
565–66 Perpendicular potentials, 39, 41
Paradoxical localization, 266, 314 Persistent low voltage, 170
Parallel potentials, 39, 41 Persistent slowing, 171
Paramyotonia, 381 Phantom spike-and-wave, 134
Parasite potentials, 459. See also Satellite potentials Phase cancellation, 242, 347
Parasomnias, 719–21, 720f Phase relation, 110
Parkinsonian tremor, 556, 557f Phase spectrum, 207
Parkinson’s disease (PD) Phases of MUP, 440–42, 456–57
movement-related potentials in, 231–32 Phasic dystonia, 569
myoclonus of, 566 Phasic muscle twitches, 708–9
Paroxysm, 176f Photic driving, 123
Index 881
Photic responses, 130, 173 Projected rhythm, 152

Photic stimulation, 122–23 Prolonged video electroencephalography (PVEEG),
Photoconvulsive response, 144, 145f 193–94, 199
Photomyogenic response, 123 clinical application, 195–99
Photoparoxysmal response, 123 equipment, 194–95
Physiologic myoclonus, 559 Propagating generators
Physiologic tremor, 554 effect of volume conduction on potential components
exaggerated, 555 due to, 44
Plasma norepinephrine, 642–43 stationary potentials produced by, 44
Plastic imprints, 634 Propriospinal myoclonus, 567
Plexopathy, 252–53 Proximal conduction, H reflex used to measure, 524–25
brachial, 817–22, 821t, 822 Proximal myotonic myopathy, 428
Polyneuropathy, 686, 687f Pseudofacilitation, 376
Polyphasic MUPs, 423, 440–42, 459 Pseudohypacusis, 302–3
Polyradiculopathy, 829–30. See also Inflammatory Pseudomyotonic discharges, 429
demyelinating polyradiculopathy Psychogenic jerks, 568
Polysomnography, 698–701, 699f, 700f, 701–703, 792 Psychogenic seizures, 196–97
purpose and role, 698 Psychogenic tremor, 558
typical settings, 701t Psychomotor-variant pattern, 132
Poor activation, 419 P-type semiconductor, 16
Pores of membranes, 73 Pupillary size, 315
Portable monitoring, 705–6 Pure-tone and speech audiometric testing, 296, 297f
Positional nystagmus, 581 QSWEAT, 630
Positional testing without fixation, 588–89 purpose and role, 630
Positioning induced nystagmus, 586–88 Quadrupoles, 38, 39f, 50
Positive, arrhythmic, slow (PAS) waves, 731, 732 Quantitative analysis
Positive afterpotential, 84 of single MUPs, 462
Positive occipital sharp transients of sleep (POSTS), of user selected MUP, computer-assisted, 462–64, 463f
127–29, 129f Quantitative EMG (QEMG), 408, 451–53, 469, 473. See
Positive rolandic sharp waves, 170 also Motor unit potentials; Semiquantitative EMG
Positive spike bursts, 14 and 6 Hz, 132, 175 decomposition-based, 466
Positive waveform, 425–26, 426f purpose and role, 453
Positron emission tomography (PET), 217 recording equipment characteristics, 454–55, 456f
Postactivation facilitation, 374 Quantitative MUNE, 497, 509. See also Motor unit
Posterior cord, 819, 820t number estimates
Posterior cranial fossa, intraoperative monitoring of, 745f, clinical applications, 510
744–47 surface averaging of needle EMG MUPs, 498–500
Posterior fossa lesions, 534 underlying assumptions, 497–98
Posterior fossa level, localization of disease at, 788 Quantitative sensory test (QST), 678–81
Posterior slow waves of youth, 175 Quantitative sudomotor axon reflex test (QSART),
Posterior syndrome, 608 629–30, 631f, 632f, 682–83
Posterior visual pathway lesions, 321–22 abnormal patterns, 632–33
Postexercise exhaustion, 380 applications, 633–34
Postictal slowing, 183 normal response, 631–32
POSTS. See Positive occipital sharp transients of sleep patient preparation, 630t
Postsynaptic effects of neurotransmitters, 90–91 purpose and role, 630
Postsynaptic potentials, 92, 92t, 101 significance, 633
inhibitory, 34–35, 92 Quantization (digitization), 57
Postural normotension, maintenance, 626 Quantization parameters, effect of, 58f
Postural tremors, 555 Quantum, 370
Posture and Valsalva response, 668 Quantum size (ADC resolution), 57
Posturography. See Computerized dynamic posturography Quiet sleep, 168
Potential. See Electrical potentials
Power, electrical, 6 Radial motor nerve, 816f
Power spectrum, 65, 471 Radiculopathies, 252, 271, 363, 806–8
Power-spectrum analysis, 471–72, 670–71 Rapid eye movement (REM) sleep. See REM sleep
Prechiasmatic lesions, 318 Rapid firing, 419
Preganglionic sympathetic neurons, 621 Rapid recruitment, 419, 421, 438
Premovement positivity (PMP), 229, 230 Rapid surgical induction, 731
Pressure palsies, liability to, 349 Rare stimulus, 233
Presynaptic inhibition, 90, 91f Rarefaction, 287
Probability distribution, moments of, 66 Rasmussen encephalitis, 180
Prognosis, determining, 805 Rate of rise, 107
Progressive diseases, 805 Reactance, calculation of, 13
882 Index
Reactivity Riche-Cannieu anastomosis, 834–35

altered, 153 Rigidity, 445
asymmetric, 153 Rise time, 107, 415, 423, 458
Reciprocal innervation, 578 Rolandic epilepsy of childhood, benign, 139–40
Recruitment, 418. See also under Needle EMG Rolandic sharp waves, positive, 170
(electromyography) Root-mean-square (rms) potentials or currents, 12
defined, 495 Rotary chair test, 594, 596f, 598. See also Computerized
rapid, 419, 421, 438 rotary chair tests
Recruitment analysis, 495–96 Rotational torticollis, 570
Recruitment pattern, 437 Rubral tremor, 557
Rectification, 17
Reduced numbers, 419 Saccadic eye movement testing, 585
Reduced recruitment, 419, 437 Safety factor, 371, 372. See also Electric safety
Re-emergent tremor, 557 Saltatory conduction, 87, 87f
Reflex sympathetic dystrophy (RSD), 681 Sampling (digitization), 58–59
Reflex withdrawal, 546 Satellite potentials, 423, 441–42, 458
Reflexes, 517, 554. See also specific reflexes Scoliosis surgery, 771
cardiopulmonary, 625 Sedation, 316
startle, 568 Sedative medications, 267
Refractory period, 85 Segmental anhidrosis, 654
Regional anhidrosis, 654 Seizures
Relative refractory period, 85 absence, 141, 176f
REM (rapid eye movement) sleep, 130, 703, 706, 706f classification of types of, 197–98
without muscle atonia, 717, 717f detection, 204–5
REM latency, 710 EEG evaluation of, 798–99
normal lower bounds for, 710t EEG recorded during a, 206f
REM sleep behavior disorder, 720–21 psychogenic, 196–97
Repeater F waves, 352 PVEEG and surgical evaluation, 198–99
Repetitive nerve stimulation (RNS), 369, 383, 788, 825, routine EEG in surgical evaluation, 218
828–29 Semicircular canals (SCCs), 577, 578
clinical correlations, 380–83 superior SCC dehiscence, 607–8
rapid rates of stimulation, 378–79 Semiconductors, 16, 17f
technique, 372–77 Semiquantitation, 414
Repetitive nerve stimulation (RNS) studies, 822–23 Semiquantitative EMG, 415, 452
criteria of abnormality, 377–78 steps in, 415
purpose and role, 370 Semirhythmic patterns, 414, 418
selection of nerve-muscle combinations, 379t, 379–80 Sensorineural disorders, 305
Repetitive transient waveforms, 62 Sensorineural hearing loss, 297
Residual latency (RL), 335 Sensorineural syndromes of the labyrinth, assessing,
Residual of disease, 805 607–9
Resistance, 6 Sensory axons. See Axons
Resistive-capacitive (RC) circuits, 9 Sensory evoked potentials, 237. See also Somatosensory
time constant, 9–10 evoked potentials
Resistive-capacitive (RC) properties in volume Sensory nerve action potential (SNAP) studies
conductors, 41–42 findings in diseases, 252–56
Resistive-inductive (RL) circuits and time constant, 10 methods, 243, 245
Resistivity, 41 planning, 251–52
Resistors, 7, 17 Sensory nerve action potentials (SNAPs), 36, 61f, 239–40,
rules for seats of, 8 256
Resonance, 13 abnormal, 241–43
Respiration, phases of, 669 averaging, 246–47
Respiratory effort and sleep, 712, 13 clinical importance, 243, 245
Respiratory effort-related arousals (RERAs), 712 intraoperative, 781
Respiratory sinus arrhythmia, 624, 627 measurements, 247–48
Respiratory synkinesis, 436–37, 437f nerve stimulation, 245–46
Respiratory variables, 715 normal, 240–41
Resting discharge rate, 578 pathophysiology, 240–43, 245–47
Resting potential, 76–78 purpose and role, 240
Restless leg syndrome (RLS), 716 recording the potential, 246
Reticular reflex myoclonus, 567 technical factors, 248
Retrochiasmatic lesions, 321–22 distance between recording electrode and nerve,
Retrocollis, 570 249–50, 251f
Reverse biasing, 17 interelectrode distance, 250–51
Reverse first-night effect, 710 measurement, 250
Rhythmic temporal theta bursts of drowsiness, 132, 132f noise and shock artifact, 248–49
Index 883
recording and stimulating distance, 249 REM (rapid eye movement) sleep, 130, 700, 706, 706f,
submaximal stimulation, 249 708, 708f, 720
temperature, 249, 250f staging, 706–12
Sensory neuronopathies, 813t Sleep apnea. See also Obstructive sleep apnea
Sensory organization test (SOT), 604–6, 605f, 606f types of, 712, 713f
Sensory receptors, special, 100 Sleep deprivation, voluntary, 719
Sensory symptoms of central origin, assessment of, 792 Sleep disorders. See also Obstructive sleep apnea
Serrated MUPs, 459 assessing, 719–21
Shaky legs syndrome, 558, 558f Sleep efficiency, 710
Sharp wave, 137, 138f Sleep latency, 710
Sharpness criteria, 204 Sleep onset, 710
Shock artifact from stimulator location, 333 Sleep onset REM (SOREM), 703
Short segmental stimulation. See Inching Sleep spikes, benign sporadic, 132–33, 134t
Shut-eye waves, 175 Sleep spindles, 127–28, 128f
Signal analysis, types of, 65f Sleep stages, 706–9
Signal display, 110–11 Sleep studies, performance of, 718–19
Signal-to-noise ratio, 60–61, 113, 248 Sleep time, total, 710
Silent period, 543, 546–48, 547f Sleep variables, 709–12
cutaneous, 548 Sliding, 333
purpose and role, 543 Slow alpha variant, 130
Silicone imprints, 634 Slow component velocity (SCV), 583
Simple kinetic tremors, 555 Slow firing, 419
Single fiber EMG (SFEMG), 455, 475–77, 488 Slow fused transients, 175
clinical applications, 484–88 Slow harmonic acceleration (SHA), 595, 598
NMJ disorders and, 827–29 Slow lambdas of childhood, 175
pitfalls Slow-wave abnormalities, 152, 160
damaged fiber, 485, 485f Small fiber neuropathy, 646–648
false trigger, 483–84, 484f Small sharp spikes (SSS), 132–33
Smooth ocular pursuit testing, 586
general, 483
Snoring sounds, 714
incorrect measurement position, 484
Sodium pump, 78
neurogenic blocking, 485f, 485
Soleus technique, 522, 523f
split fiber or ephaptic activation, 485
Somatosensory evoked potentials (SEPs), 36, 43–44,
unique to stimulated SFEMG, 485–86
257–58, 277, 788
unstable trigger, 483
anesthesia and, 757f, 757–58
clinical applications, 269
stimulated, 478–79 disorders of CNS, 271–76
technique, 477 disorders of peripheral nervous system, 269–71
hardware, 477–78 SEP findings in brain death, 276
measurement, 479–83 SEPs recorded in ICU, 276–77
method of activation, 478–79 factors that affect amplitude and latencies of evoked
software, 478 response, 266–68
voluntary, 478 in infants and children, 276
Single peak, 39 interpretation of, 266
Single potential, 109 intraoperative, 741, 778. See also under Intraoperative
measurable variables of, 109t spinal cord monitoring
Single-photon emission computed tomography (SPECT), localization
199, 216 amplitude reduction, 268
Sink, 34 latency prolongation, 268
Skin, sympathetic innervation of the, 622–23 nerve stimulation variables and, 258–60
Skin blood flow (SBF), 638 neuroanatomic sites of origin of, 258
Skin problems, 647t purpose and role, 258
needle exams and, 410 recording
Skin response, sympathetic, 653, 683 averaging, 261–62
Skin vasomotor reflexes, 637–43 methods and montages, 260t, 260–61
Sleep. See also under EEG peak nomenclature, 262–66
active vs. quiet, 168 volume conduction and near- and far-field potentials,
assessing impairment of, 792 261
assessing movements in, 715–18 Somnolence, disorders of excessive, 719
assessing respiration during, 712–15 Spasm, 568
benign epileptiform transients of, 132–33 hemifacial, 535–36, 536f
EEG and, 123, 127–30, 132–34, 135f, 136, 173–75, 697 Spasmodic dysphonia, 570
NREM, 697, 706, 709f Spasmodic torticollis, 570
periodic limb movements of, 568–69, 702, 715–17, 716f Spasticity, 445
positive occipital sharp transients of, 127–29, 129f Spatial gradient, 43
884 Index
Spatial summation of local potentials, 81f, 82 Subanesthetic concentrations, 731

Spectral analysis. See Fourier (spectral) analysis Subclinical rhythmic electrographic discharge of adults
Speech-recognition thresholds, 297 (SREDA), 130, 131f
Spike (event), 107 characteristics, 131t
Spike and slow wave complex, 137 Subdural effusion and empyemas, 180–81
Spike and wave, 134, 148f, 176, 177f Subdural hematoma, 158
3-Hz, 141, 142f Subjective visual vertical (SVV) assessment, 598–99
6-Hz, 134, 135f purpose and role, 598
wicket, 127, 134, 135f, 136 Subtraction ictal-interictal SPECT coregistered on
Spike discharge, 137, 138f magnetic resonance images (SISCOM), 199
Spike duration, 458 Sudomotor axon reflex, 99f
Spike form, 425, 426f Suggested immobilization test, 716
Spike-triggered averaging (STA) MUNE, 498–500, 499f Superconducting quantum interference device (SQUID),
Spinal and bulbar muscular atrophy (SBMA), 255 47, 211
Spinal cord, 99, 779 Superimposition, 376, 376f
Spinal cord injury, 394–95 Superior nerve neurolabyrinthitis, 607
Spinal cord lesions, 273–74 Superior nerve syndrome, 607
Spinal cord level, localization of disease at, 788–89 Superior SCC dehiscence, 607–8
Spinal cord tumors, 273 Suppression, 153
Spinal surgeries. See also Intraoperative spinal cord Supratentorial lesions, 274
monitoring Supratentorial level, 788
types of, 768–74 localization of disease at, 788
Spindle coma pattern, 163 Surface EMG, 552–53
Spine disease Surface EMG activity, normal patterns of, 554, 554f
cervical, 770–71 Surface-recorded motor unit potential (SMUP), 494, 497
lumbosacral, 772 determining the size of, 498
primary, 768–69 Surgery. See Intraoperative monitoring
primary neural disease, 772 Surgical evaluation, PVEEG and, 198–99
thoracic, 771 Sustained handgrip, 643
Split fiber, 485 Sweat distribution. See also Thermoregulatory sweat
Split syndrome, 608 distribution
Spondees, 297 non-Mayo methods of measuring, 653
Spontaneous activity. See also under Needle EMG Sweat glands, 98–99, 647t
(electromyography) Sweat loss, focal, 654
automated methods of analysis of, 472 Sweat measurement. See also Thermoregulatory sweat
Sputtering, 435 test
Squatting, heart rate response to, 672 imprint methods of, 634
Stable process, 805 Sympathetic activity, muscle, 623
Stacked ABR (auditory brain stem response), 309 Sympathetic function, 621–22
Staggering. See X-shifted fashion assessment of, 622
Stålberg, E., 486 Sympathetic innervation of the skin, 622–23
Standard stimulus, 233 Sympathetic outflow, functional anatomy of, 621
Startle disorders, 568 Sympathetic reflexes, 621–22
Static compensation, 581 Sympathetic skin response (SSR), 653, 683
Static positional test, 588, 589f Symptomatic myoclonus, 559
Stationary (evoked) potentials, 44, 45f Synapses, electrical, 93
Stationary waves, 261 Synaptic cleft, 88
Statistical analysis, 66 Synaptic transmission, 88, 89, 89f. See also Neurochemical
Statistical methods, multivariate, 209–10 transmitters
Statistical (STAT) MUNE, 506–8, 509f, 510f abnormalities in, 95f
Steady state, 77–78 characteristics, 88–89
Step test, 595 Synaptic vesicles, 90
Stiff-man syndrome, 445 Synkinesis, 436–37, 437f. See also Facial synkinesis
Stimulus intensity, 287 Syringomyelia, 273
Stimulus polarity, 287
Stimulus rate, 287 Tailored resection, 223
Stimulus-induced rhythmic, periodic or ictal discharges Task-specific tremors, 559
(SIRPIDS), 148, 149f Temperature, 249, 250f, 343, 718, 764
Straining, 665 Temporal dispersion, abnormal, 815, 815f
Stretch reflex, 544–45 Temporal intermittent rhythmic delta activity (TIRDA),
Stroke, 276, 394 145, 145f, 157f
Structural generators, 97–101 Temporal spikes, 138, 139f
Subacute disorder, 805 Temporal summation of local potentials, 81, 81f
Subacute sclerosing panencephalitis (SSPE), 161, 162f, Terminal bouton, 370
180, 565–66 Terminal component (MUP), 457, 458
Index 885
Terminal negative afterpotential, 458 psychogenic, 558–59

Terminal-two electrode, 331. See also Inactive reference recording techniques, 555
electrodes task- and position-specific, 557–58
Tetanic stimulation, 371 Trigeminal contact heat evoked potential stimulator
Tethered cord syndrome, 773 (CHEPS) method, 539–41
Thalamic tremor, 557 Trigeminal nerve lesions, 531–32, 532f
Thermal hypesthesia, 680–81f Trigger, amplitude, 478
Thermal pain hyperalgesia, 680, 680f, 681f Triphasic wave patterns, 161–62, 554
Thermoregulatory sweat distribution, 653–55 Triphasic waveforms, 47
abnormal distributions, 654, 655f Triplets, 444, 444f
normal distributions, 653–54, 654f Trunk
Thermoregulatory sweat test (TST), 645, 656 lower, 818, 819t, 822
clinical syndromes and problems evaluated, 646–48t, middle, 818, 819t, 822
648–50 muscle innervation according to, 821t
difficulties and pitfalls in interpretation, 657–80 upper, 818, 818t, 822
method, 650–53 Trunk lesions, 818–19, 821t, 822
procedure, 658–59 root vs. plexus lesions, 822
purpose and role, 645 Tuberous sclerosis, 172f
reporting results, 656f, 656–57 Tumors, 155–56. See also Cranial nerve (CN) VIII tumors
Theta activity, 125 brain stem, 292f
Thickness (MUP), 458 in children, 181
Thoracic outlet syndrome, 270, 363 Turns, 458–59
Thoracic radiculopathies, 807–8 Turns/amplitude ratio, 468
Thoracic spine disease, 771 Twitching. See Jerks; Tremor
Threshold, 498
for excitation, 83 Ulnar nerve, 263–65, 381, 381f, 835
Tibial mixed nerve SEPs, 265–66 Ulnar neuropathies, 254–55, 359, 360f, 361, 809
Tibial nerve stimulation, 259f Unilateral weakness (UW), 585
Unstable MUPs, 442, 443t, 444
Tibial scalp potentials, 274f
Upper tracing, 254f
Tics, 571
Tilt-up. See also Head-up tilt V waves, 128, 129f
beat-to-beat BP response to, 640–41 Vagal baroreflex sensitivity, 667–68
Time constant (TC), 9–11 Valsalva maneuver (VM), 638, 662
Tonic pattern, 554 beat-to-beat BP response to, 638–40, 639f
Top tracing, 242f, 467f normal response and physiologic basis, 665–67, 666f
Topographic analysis, EEG, 209–10 phases, 638–40t
Topographic displays (mapping), 208–9, 209f pitfalls in performing, 669
Torticollis, 569, 570 technique, 667
Tracé alternant pattern, 168 Valsalva ratio, 667
Transcortical electrical stimulation (TCES), 761 pitfalls in interpreting, 669–70
Transcortical (reflex) loop, 544 Valsalva response, factors affecting, 668–69
Transcranial Doppler (TCD), 733 Varying MUPs, 442, 443t, 444
Transcranial electric stimulation (TES), 385–87, 387f, 389 Vascular diseases, surgery for, 773
Transcranial electrical magnetic stimulation, 31 Vascular lesions, 181
Transcranial magnetic stimulation (TMS), 385, 386, 388f, Velocity recovery function, 480–81
389 Venoarteriolar reflexes, 99f, 623, 625, 641–42
Transient abnormalities, 183 neural pathway, 641f
Transient disorders, 159 Vertex sharp transients (V waves), 128, 129f
Transient evoked otoacoustic emissions (TEOAEs), 301, Vertigo, 576, 607–8
301f benign paroxysmal positioning, 586, 587, 607
Transistors, 16–19 Vestibular evoked myogenic potentials (VEMPs),
Transmembrane ion gradients, 69–70, 70f 599–603, 607–9
Traveling evoked potentials, 44, 45f example of VEMP waveforms, 601f
Traveling waves, 261 lesions effects, 603
Tremor, 445, 445f, 554 normal values, 601, 602t
abnormal patterns, 555 purpose and role, 599
cerebellar, 557 recording methods, 599
classification by localization and electrophysiologic analysis, 600–1
features, 556t procedure, 600
essential, 555–56 signal averaging, 600
exaggerated physiologic, 555 stimuli, 599
Holmes, 557 Vestibular hair cell systems, 578
orthostatic, 558, 558f Vestibular mechanisms, functional anatomy and
palatal, 558 physiology of, 577–80
Parkinsonian, 556–57, 557f Vestibular nystagmus, 581–82
886 Index
Vestibular rehabilitation, 610 Volume conduction, 34

Vestibular test methods, laboratory, 582 principles of, 33–34
Vestibular testing, 576–77 applications of, 43–49
clinical applications, 607–9 Volume conduction theory, 41–43
purpose and role, 577 Volume conductors, 33–34
Vestibulo-ocular reflex (VOR), 577, 578, 594 distant recordings in, 42–43
laboratory examination electric properties, 37f, 37–38, 41–43
preparation for testing, 584 nerve action potential from nerves on, 36f
recording method, 582–84 potentials in
videonystagmography test battery, 584–94 cortically generated, 34–35
Vestibulopathy, clinical features of, 580–82 peripherally generated, 35–36
Vestibulo-spinal reflex (VSR), 577, 581–82
VSR-based measures, 599–603 Wake after sleep, 710
Videonystagmography (VNG). See ENG/VNG test battery Wakefulness, 706, 706f
Visceral afferents, 619–20 Warm-up phenomenon, 429
Visceral efferents, 620 Waveforms. See also specific topics
Visceral function, assessing impairment of, 792 alteration of, 108
Visual acuity, 315 physiologic, 108–11
Visual evoked potentials (VEPs), 311, 312, 322, 788 artifactual, 111–13
factors affecting VEP response, 314–16 continuous, 103–6, 109–10
interpretation, 317 event recording and, 106–8
intraoperative, 742 Weakness
localization of visual system lesions, 317–22 episodic, 827–28
purpose and role, 312 facial, 831
recording, 314 generalized, 822–24
stimulation, 312, 314 unilateral, 585
Visual pathways, central West’s syndrome, 143
neuroanatomy, 312, 313f White matter disease, 179
Visual system anatomy and physiology, 312, 313f Wicket spikes/wicket waves, 127, 134, 135f, 136
Visually enhanced VOR (VVOR), 594, 596 Wind-up, 686
Visually guided eye movements, inventory of, 585–86 Word recognition, 297–98
Voice tremor, isolated, 557 Writers’ cramp, 569, 571
Voltage-gated cation channels, 73–75 Writing tremor, primary, 557
Voltage(s)
of common signals, 110t
persistent low, 170 X-shifted fashion, display in, 375, 376f

ClinicalNeurophysiology3e2009 PDF

Uploaded by

Copyright:

Available Formats

ClinicalNeurophysiology3e2009 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ClinicalNeurophysiology3e2009 PDF

Uploaded by

Copyright:

Available Formats

CLINICAL NEUROPHYSIOLOGY

Contemporary Neurology Series

53 SLEEP MEDICINE 67 THE CLINICAL SCIENCE OF

Oxford New York

Copyright © 2009 by Mayo Foundation for Medical Education and Research

Oxford is a registered trademark of Oxford University Press

All rights reserved. No part of this publication may be reproduced,

Library of Congress Cataloging-in-Publication Data

Printed in the United States of America

Jasper R. Daube, MD and Devon I. Rubin, MD

SECTION 1 BASIC NEUROPHYSIOLOGY AND AN

1. ELECTRICITY AND ELECTRONICS FOR CLINICAL

2. ELECTRIC SAFETY IN THE LABORATORY AND HOSPITAL 21

CURRENT SOURCES: MONOPOLES, DIPOLES, AND

4. DIGITAL SIGNAL PROCESSING 53

6. ELECTROPHYSIOLOGIC GENERATORS IN CLINICAL

7. WAVEFORMS AND ARTIFACTS 103

PHYSIOLOGIC ALTERATION OF WAVEFORMS 108

SECTION 2 ELECTROPHYSIOLOGIC ASSESSMENT OF

Part A Assessment of Cortical Function

8. ELECTROENCEPHALOGRAPHY: ADULT, NORMAL, AND BENIGN

9. EPILEPTIFORM ACTIVITY 137

10. ADULT EEG: ABNORMAL NONEPILEPTIFORM ACTIVITY 151

11. ELECTROENCEPHALOGRAPHY: ELECTROENCEPHALOGRAMS OF

12. AMBULATORY ELECTROENCEPHALOGRAPHY 187

13. PROLONGED VIDEO ELECTROENCEPHALOGRAPHY 193

14. ELECTROENCEPHALOGRAPHIC SPECIAL STUDIES 203

15. ELECTROENCEPHALOGRAPHY IN THE SURGICAL EVALUATION

PRESURGICAL EVALUATION WITH CONTINUOUS OR CHRONIC

16. MOVEMENT-RELATED CORTICAL POTENTIALS AND

Part B Sensory Pathways

17. SENSORY NERVE ACTION POTENTIALS 239

18. SOMATOSENSORY EVOKED POTENTIALS 257

19. BRAIN STEM AUDITORY EVOKED POTENTIALS IN CENTRAL

INTERPRETATION OF BAEPs 287

20. AUDIOGRAM, ACOUSTIC REFLEXES, AND EVOKED OTOACOUSTIC

21. BRAIN STEM AUDITORY EVOKED POTENTIALS IN PERIPHERAL

22. VISUAL EVOKED POTENTIALS 311

Part C Motor Pathways

23. COMPOUND MUSCLE ACTION POTENTIALS 327

CMAP MEASUREMENTS 334

24. ASSESSING THE NEUROMUSCULAR JUNCTION WITH REPETITIVE

25. MOTOR EVOKED POTENTIALS 385

Part D Assessing the Motor Unit

26. ASSESSING THE MOTOR UNIT WITH NEEDLE

POTENTIAL COMPLICATIONS DURING NEEDLE

27. QUANTITATIVE ELECTROMYOGRAPHY 451

28. SINGLE FIBER ELECTROMYOGRAPHY 475

29. QUANTITATIVE MOTOR UNIT NUMBER ESTIMATES 493

Part E Reﬂexes and Central Motor Control

30. H REFLEXES 519

31. CRANIAL REFLEXES AND RELATED TECHNIQUES 529

32. LONG LATENCY REFLEXES AND THE SILENT PERIOD 543

33. MOVEMENT DISORDERS 551

34. VERTIGO AND BALANCE 575

Part F Autonomic Function

35. AUTONOMIC PHYSIOLOGY 617

AUTONOMIC CONTROL OF HEART RATE 624

36. QUANTITATIVE SUDOMOTOR AXON REFLEX AND