Biosignal Processing Fundamentals and Recent Applications

with MATLAB ® 1st Edition

Visit the link below to download the full version of this book:

https://medipdf.com/product/biosignal-processing-fundamentals-and-recent-applica
tions-with-matlab-1st-edition/

Click Download Now

Foreword
The interdisciplinary field of biosignal processing has become an indispensable part
of today’s medical technology applications. In everyday clinical practice, large quant-
ities of image data and signals are continuously evaluated for diagnostic purposes.
In order to be able to offer patients the most efficient therapy for their diseases, re-
liable diagnostic tools are the foundations for proper medical decision-making. In
many situations, the doctor only has minutes – one reason why the data and signals
must be processed in a meaningful way. In addition to desirable real-time processing,
this includes above all reliability and robustness as well as a clear and unambigu-
ous presentation of the diagnostic statements. According to these requirements, some
fundamental rules must be considered in the analogue and digital processing of bi-
osignals.
This book provides a fundamental work on biosignal processing in which, be-
sides a general introduction to the generation, measurement and analogue/digital
processing of biosignals, also recent application examples are discussed. Thus, the
book is on the one hand suitable for students and teachers as an introduction to the
fundamental methodology and on the other hand a compact reference for experienced
users in the field of medical technology and medical informatics.
The subject matter of the book is supplemented by numerous recent research top-
ics in biosignal processing in the form of application-oriented programming examples
and thus also offers researchers from the natural and life sciences such as medicine
and medical technology diverse incitation for the development of novel methods in
the processing of biosignals.
The book introduces the fundamental physiological processes in the formation of
biosignals, the authors discuss the basics of analogue signal amplification and pro-
cessing as well as the digitisation for various biosignals along the measurement chain
by means of sensor and measurement technology. The design of important digital fil-
ters and basic methods for analysing the signals in the time, frequency and composite
domain as well as methods for statistical or model-based evaluation of signals are also
discussed in detail.
The content is equally convenient for students of natural sciences, namely math-
ematics, physics and biology, engineering sciences, especially computer science,
medical technology and physical engineering, as well as life sciences, especially
medicine. The presentation of the mathematical and experimental methods origin-
ated in the authors’ long-standing activity as lecturers in biosignal processing.
Compact introductions to the physical and physiological basics make it easier for
readers from related disciplines to get started and also serve as a refresher on neces-
sary basic concepts. The sections do not claim to be complete and therefore refer to
further literature at the appropriate places.

https://doi.org/10.1515/9783110736298-202
VI | Foreword

Each chapter is supplemented by an example and exercise section, which can be

used by the reader to deepen and prepare for the preparation of the material. The pro-
gramming examples for Matlab, Scilab/COS and LTSpice are available as additional
material and can be downloaded from the publisher’s website.
The content and form of the book originates from the authors’ didactic experience
of lectures held at three different German University of Applied Sciences. The wealth
of experience gained over the years thus also includes influences from the indirect
collaboration of students, doctoral students and colleagues, who cannot all be men-
tioned here by name. The authors would therefore like to express our gratitude for
all the support we have received in the form of suggestions for improvement, criticism
and praise during our time as lecturers at the University! Furthermore, the authors like
to thank Prof. Dr. Jörg Subke and Benedict Schneider for their contribution of sub-
section 6.2.2 and Dr. Urs Hackstein for his contribution in subsection 6.3.4. Finally,
we would like to thank Ms Eva Funk and Franziska Marschall for creating, editing or
translating numerous illustrations.
For the sake of better readability, the book uses the masculine form throughout,
although all readers should of course feel addressed by their respective person.

Stefan Bernhard
Contents
Foreword | V

Image credits | XI

1 Introduction to Biosignal Processing | 1

2 Fundamentals of Information, Signal and System Theory | 5

2.1 Information and Information Transmission | 5
2.2 Connection between Signals and Systems | 15
2.3 Definition and Classification of Signals | 18
2.3.1 Univariate and Multivariate Signals | 18
2.3.2 Periodic, Quasi-Periodic, Aperiodic and Transient Signals | 19
2.3.3 Even and Odd Signals | 26
2.3.4 Causal and Acausal Signals | 27
2.3.5 Energy and Power Signals | 27
2.3.6 Deterministic and Stochastic Signals | 29
2.3.7 Continuous and Discrete Signals | 33
2.4 Signal Processing Transformations | 34
2.4.1 Continuous Fourier-Transformation | 35
2.4.2 Continuous Laplace Transform | 38
2.4.3 Continuous Short-Time Fourier-Transform and Wavelet Transform | 40
2.4.4 Continuous Linear Convolution | 44
2.5 Biosignal Processing and the Derivation of
Diagnostic Information | 45
2.6 Post-Reading and Exercises | 46

3 Fundamentals of the Formation of Biosignals | 51

3.1 Physiology and Electrical Activity of Muscle and Nerve Cells | 53
3.1.1 Formation and Function of Biomembranes | 54
3.1.2 Analogy to Electrical Circuits | 57
3.1.3 Emergence and Propagation of Action Potentials | 59
3.2 Electrophysiology of the Heart | 64
3.2.1 General Excitation of Muscle Cells | 65
3.2.2 Measurement of Electrical Potentials at the Body Surface | 67
3.2.3 Process of Excitation Propagation during a Heart Beat | 73
3.2.4 Modelling the Excitation System | 75
3.3 Taxonomy of Biosignals | 82
3.4 Post-Reading and Exercises | 87
VIII | Contents

4 Measurement of Biosignals and Analog Signal Processing | 91

4.1 Measurement of Electrical Biosignals | 91
4.1.1 Electrodes | 93
4.1.2 Electrical Amplifier | 97
4.2 Signal Interference | 104
4.2.1 Network Disturbances | 104
4.2.2 Transient Disturbances | 109
4.2.3 High-Frequency Interference due to Electromagnetic Radiation | 109
4.3 Transducer for Non-Electrical Biosignals | 110
4.3.1 Sound Transducer | 110
4.3.2 Optical Sensors for Plethysmography and Determination of Oxygen
Saturation | 113
4.4 Interference Suppression and Analog Filtering | 115
4.5 Design of Analogue Filters | 123
4.5.1 Selective Filters to Optimise the Magnitude Frequency Response | 123
4.5.2 Selective filters with Group Delay Optimisation | 143
4.6 Post-Reading and Exercises | 144

5 Methods for Discrete Processing and Analysis of Biosignals | 149

5.1 Discretisation of Continuous Signals | 149
5.2 Discrete Transformations of Signal Processing | 154
5.2.1 The Discrete-Time Fourier Transform | 154
5.2.2 The Discrete Fourier Transform (DFT) | 155
5.2.3 Discrete Laplace Transform and z-Transform | 158
5.3 Methods for Analysis and Processing of Discrete Biosignals | 159
5.3.1 Time Domain Signal Analysis and Matching | 159
5.3.2 Signal Analysis in the Frequency Domain | 175
5.3.3 Signal Analysis in the Time-Frequency Domain | 183
5.3.4 Discrete Linear Time-Invariant Systems and Digital Filters | 190
5.4 Post-Reading and Exercises | 208

6 Applications and Methods in Biosignal Processing | 213

6.1 Signals of the Brain | 213
6.2 Signals of the Muscles and Motions | 220
6.2.1 Spectral Analysis of the One-Channel EMG | 222
6.2.2 Acoustic-Kinetic Analysis of Osteoarthrosis Patients | 224
6.3 Signals of the Cardiovascular System | 241
6.3.1 Electrocardiogram | 241
6.3.2 Phonocardiogram | 264
6.3.3 Determination of Oxygen Saturation and
Photoplethysmography | 274
6.3.4 Signal Classification of Multichannel Photoplethysmography | 277
 Contents | IX

6.4 Post-Reading and Exercises | 285

7 Appendix: Quantity- / Unit Symbols and Important Constants | 289

Bibliography | 295

Index | 299
Image credits
Fig. 1.1: Image library from TMSi, The Netherlands
Fig.2.7: MRI and CT angiographies of the heart/brain from http://www.osirix-viewer.com/resources/
dicom-image-library/
Fig.2.9: MRI and CT angiographies of the heart/brain from http://www.osirix-viewer.com/resources/
dicom-image-library/
Fig.3.4: Cell membrane from © natros – stock.adobe.com, http://stock.adobe.com
Fig.3.12: Cell, axon, muscle from © joshya – stock.adobe.com, http://stock.adobe.com
Fig.3.38: Brain from © bilderzwerg – stock.adobe.com, http://stock.adobe.com
Fig.3.38: Sensory organs from © www.freepik.com
Fig.4.2: Skin from © Neokryuger – stock.adobe.com, http://stock.adobe.com
Fig.6.1: © Dule964 | Dreamstime.com
Fig.6.2: Modified from © Alila07 | Dreamstime.com
Fig.6.4: © Alila07 | Dreamstime.com
Fig.6.6: Motor unit from © Balint Radu – stock.adobe.com, http://stock.adobe.com
Figs.3.1, 3.20, 4.6, 4.13, 4.16, 6.9, 6.10, 6.11, 6.12, 6.13, 6.14: Modified body outline from ©
Aaltazar – istock.de
Figs. 2.1, 2.2, 2.5, 2.6, 2.7, 2.9, 2.10,2.20, 2.21, 2.24, 3.3, 3.5, 3.6, 3.7, 3.34: Generated using TikZ
library and examples [75].
Figs. 3.13, 3.17, 6.11 - 6.13: Adapted from [47].
Post-processing, compilation and creation of visuals © Eva Funk – www.evafunk.com

https://doi.org/10.1515/9783110736298-204
1 Introduction to Biosignal Processing
Biosignal processing is an interdisciplinary field of research that includes medical in-
formatics, signal processing and life sciences. The main objective of the analysis of
human biosignals is to support medical diagnosis with the help of mathematical meth-
ods. Intelligent evaluation of the signals should provide the medical practitioner with
valuable quantitative information in the diagnosis and effectively support them in
medical decision-making. The topics are wide-ranging: On the one hand, they include
signal processing, methods for monitoring and controlling vital functions in intens-
ive care medicine, for example through the automatic classification of signals, and
on the other hand, the formal description of the relationships between signals and
physiological functions in medical research. Another growing discipline is the mod-
elling of physiological phenomena to gain a deeper understanding of the underlying
pathological mechanisms to subsequently improve the technology in medical device
engineering. Modern simulation techniques often make use of statistical methods like
parameter variation and estimation, as well as the quantification of uncertainties in
the model approach – just to develop increasingly sophisticated data evaluation meth-
ods for diagnostic purposes. In many cases, these methods are the basis for optimal
control and regulation or lead to a complete replacement of physiological functions,
as for example in prosthetics.
The acquisition of biosignals for diagnostic purposes has a long history. Begin-
ning with the recognition of the electrical activity of nerve and muscle cells, which en-
gaged Luigi Galvani in his famous frog’s leg experiment in 1787, a chain of far-reaching
findings on the fundamental mechanisms of electrophysiology followed. As early as
1876, E. J. Marey succeeded in graphically depicting these processes for the first time.
Willem Einthoven received the Nobel Prize in Medicine in 1924 for the development
of the string galvanometer and the physiological interpretation of the electrocardio-
gram. Further development through the use of improved sensing and measuring tech-
nology, such as the tube amplifier, transistors and later integrated circuits as well as
the microprocessor technology, led to a considerable improvement in signal quality
over time.
Nowadays, the main focus of research is on increasingly sophisticated evaluation
algorithms, for example based on large amounts of data¹ and the miniaturisation of
measurement and transmission technology as well as data storage on the internet.
Today, the evaluation of information and signals from the human body is the basis
of almost every medical diagnosis. Electrocardiography (ECG) in particular has de-
veloped into one of the most frequently used medical examination methods – mil-
lions of ECGs are recorded in the world every day. Especially in the field of long-term

1 Evaluation of extremely large data sets with the help of computer algorithms with the aim of making
connections such as trends, correlations or patterns in the behaviour of the data visible.

https://doi.org/10.1515/9783110736298-001
2 | 1 Introduction to Biosignal Processing

ECGs (over 24 h), modern methods are indispensable, as a heart beats approximately
80,000 to 100,000 times during this time. In this order of magnitude, the search for ab-
normal heartbeats such as cardiac arrhythmias or unusual changes such as ventricu-
lar fibrillation presents doctors with a practically impossible task. In addition to the
precious time that would be involved in manual review, there would be a lapse of at-
tention and thus a loss of important events – not to mention the compilation of stat-
istics.
However, large amounts of data are not only generated in the analysis of intens-
ive care data, but also in telemedicine. The variety of data in the field of commercial
sports medicine has grown rapidly through the use of mobile technologies and soft-
ware on smartphones, as has the continuous monitoring of older people to monitor
their health. The list of possible vital signs and health parameters besides ECG is long:
blood pressure, heart rate, blood oxygen saturation, body weight and temperature –
all require conscientious and contextual evaluation with mathematical methods to
guide diagnosis. With their help, the essential parameters are to be found automatic-
ally and the physician is to be pointed to relevant contents. Especially when monitor-
ing patients in intensive care units or in telemedicine, threatening conditions should
be detected quickly and reliably and the treating staff should be informed by an alarm.
The problem with this form of automatic detection of important events is to de-
velop a reliable algorithm with high accuracy –an almost impossible task given the
diversity of patients and the variability in the signals. In other words, algorithms must
be applicable to all possible variants of a given signal. This so-called robustness of an
algorithm is an essential feature in the subsequent approval as a medical device. In
addition, the measurement technology and algorithms must be absolutely insensit-
ive to external electromagnetic interference to which a measurement setup may be
exposed.
Mathematical methods and the development of software are therefore indispens-
able components of biosignal analysis. The success of modern monitoring systems
therefore comes only to a small extent from the development of electronic hardware –
the far greater contribution to the success of an innovative product is now made by
intelligent signal evaluation. For example, in addition to exclusive filtering in the fre-
quency or time domain, interfering signals are nowadays increasingly analysed us-
ing the wavelet transformation in the time-frequency domain, since it allows the best
possible time and frequency resolution for a given signal section. The detection of im-
portant signal sections in an ECG signal course, for example, requires the extraction
of statistically robust features, which are then fed to a classifier or a neural network
for analysis. With this method, a large number of anomalies in ECGs can already be
reliably distinguished today.
Model-based techniques such as the "Kalman" filter or "Markov" models are also
used to detect abnormal (pathological) states and the associated changes in state.
Early detection of trends or random fluctuations in a signal course, such as signal vari-
ations shortly before an epileptic seizure in the electroencephalogram (EEG), using
 1 Introduction to Biosignal Processing | 3

machine learning methods can be used to identify an upcoming seizure. These stat-
istical methods are intended to distinguish stochastic fluctuations from deterministic
ones or to recognise significant changes before they occur themselves. The mathem-
atical relationships that can be used to predict the probability of a future change from
current readings, in order to detect possible signs of a developing pathology, often lie
in the theory of complex systems and non-linear dynamics. For example, in normal
heart rhythms one observes stable rhythms (trajectories in phase space), so-called at-
tractors, which, however, can sometimes pass into chaos via so-called bifurcations
and thus into pathological states such as ventricular fibrillation. Perhaps in the not
too distant future, this insight can be used for the early detection and diagnosis of
such incidents to the benefit of the patient.
This book provides an introduction to the theory and principal methodology of
biosignal processing, describes the origins of the most common human biosignals,
and teaches the techniques for measurement and modern information processing us-
ing LTSpice and Matlab/Simulink². After a brief introduction and historical review to
the individual topics of electrophysiology and analogue and digital signal processing,
the reader is introduced to the practice of biosignal processing with Matlab/Simulink
through selected applications of the methodology learned.
The reader is given an overview of the variety of human biosignals (cf. Fig-
ure 1.1) and is introduced to the topic on the basis of selected biosignals, such as
muscle activity in the electromyogram (EMG), the activity of the heart muscle in the
electrocardiogram (ECG), the activity of the nerve cells of the brain in the electro-
encephalogram (EEG) or the measurement of the oxygen saturation of the blood in
the photoplethysmogram (PPG). In this context, the fundamentals of deriving, pre-
processing, recognising and interpreting these signals are taught with the help of the
simulation environment LTSpice and the programming language Matlab/Simulink.
Carrying out the exercises in Matlab/Simulink also teaches the necessary techniques
of practical biosignal processing and offers the opportunity to apply acquired theor-
etical knowledge in practice. For a better overview of the formula symbols, units and
constants used, a table sorted by chapter is provided in chapter 7 for reference.
The current state of research and development is presented in three fields of ap-
plication from the authors’ research topics: (i) mathematical modelling and analysis
of signals from the heart / circulatory system, and analysis of the electrical activity of
(ii) muscles and (iii) the brain. Each chapter includes a series of examples and exer-
cises, as well as a presentation of the future perspectives of the respective field.

2 The MathWorks, Inc.

4 | 1 Introduction to Biosignal Processing

α β α

δ, ϑ
50µV

Fig. 1.1: Possible sources from the fields of electroencephalography (EEG) and electromyography
(EMG) with illustration of the typical temporal signal sequences of the individual rhythms in the EEG
(bottom) and of a repeated muscle contractions in the EMG (top).
2 Fundamentals of Information, Signal and System
Theory
In biosignal processing, a variety of methods from different disciplines are applied
to problems in medicine. The methods of classical information and signal processing
form an important basis. The following chapter introduces methods from these discip-
lines that are commonly used in biosignal processing and discusses the context along
practical examples.

2.1 Information and Information Transmission

The concept of information occupies a crucial place in all scientific disciplines, in-
cluding biosignal processing. However, the concept of information is not uniformly
defined in the various disciplines and has different meanings. The most common is
the concept of information coined by C. E. Shannon¹, which is used in communica-
tions engineering.

Digital Information
However, the Shannonian definition of digital information refers exclusively to statist-
ical aspects of information, i.e. the probability of occurrence of characters of a char-
acter set in a string to be transmitted. The statistical information content I of this se-
quence of n characters with the respective occurrence probability p i of the individual
characters is:
n
1 1 1 1
I = ∑ log2 = log2 + log2 + ⋅ ⋅ ⋅ + log2 ≥0, n∈ℤ. (2.1)
i=1
pi p1 p2 pn

According to this equation, the information increases steadily with the number n of
characters in the chain, as long as the following relation holds for a character from a
character set with N different characters:
N
0 ≤ pj ≤ 1 , ∑ pj = 1 . (2.2)
j=1

Equation 2.2 stands for the existence of a character and at the same time guarantees
that each character appears only once in the character set. From the two equations
it becomes clear that the information content of a character string increases strongly

1 Claude Elwood Shannon (1916–2001), US mathematician and electrical engineer, founder of inform-
ation theory

https://doi.org/10.1515/9783110736298-002