Clinical Doppler Ultrasound PDF

Commissioning Editor: Meghan McAteer
Development Editors: Hilary Hewitt and Louise Cook

Editorial Assistant: Nani Clansey
Project Manager: Gemma Lawson
Design Manager: Jayne Jones
Illustration Manager: Bruce Hogarth
Illustrator: Hardlines
ii Marketing Manager(s) (UK/USA): Deb Watkins and Emily Christie
Clinical Doppler
Ultrasound SECOND EDITION
Paul Allan BSc, DMRD, FRCR, FRCPE

Consultant Radiologist and Clinical Director of Radiology
Department of Medical Radiology
Royal Infirmary of Edinburgh
Edinburgh, UK
Paul A. Dubbins BSc, MBBS, FRCR

Consultant Radiologist
Department of Radiology
Derriford Hospital
Plymouth, UK
W. Norman McDicken PhD, FIPEM

Professor of Medical Physics and Medical Engineering
Medical Physics
The University of Edinburgh
Edinburgh, UK
Myron A. Pozniak, MD
Professor of Medical Imaging
Department of Radiology
University of Wisconsin
Madison, WI, USA
iii
An imprint of Elsevier Limited
© 2006, Elsevier Limited. All rights reserved.
First published 2006
First edition 2000
The right of Paul Allan, Paul A Dubbins, W Norman McDicken and Myron Pozniak to be identified
as authors of this work has been asserted by them in accordance with the Copyright, Designs and
Patents Act 1988.
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 the Publishers. Permissions may be sought directly from Elsevier’s
Health Sciences Rights Department, 1600 John F. Kennedy Boulevard, Suite 1800, Philadelphia,
PA 19103-2899, USA: phone: (+1) 215 239 3804; fax: (+1) 215 239 3805; or, e-mail:
[email protected] may also complete your request on-line via the Elsevier homepage
(http://www.elsevier.com), by selecting ‘Support and contact’ and then ‘Copyright and Permission’.
ISBN-13: 978-0-443-10116-8
ISBN-10: 0-443-10116-7
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data

A catalog record for this book is available from the Library of Congress
Notice
Medical knowledge is constantly changing. Standard safety precautions must be followed, but as
new research and clinical experience broaden our knowledge, changes in treatment and drug
therapy may become necessary or appropriate. Readers are advised to check the most current
product information provided by the manufacturer of each drug to be administered to verify the
recommended dose, the method and duration of administration, and contraindications. It is the
responsibility of the practitioner, relying on experience and knowledge of the patient, to determine
dosages and the best treatment for each individual patient. Neither the Publisher nor the author
assume any liability for any injury and/or damage to persons or property arising from this
publication.
The Publisher
Printed in China
Last digit is the print number: 9 8 7 6 5 4 3 2 1
iv
List of contributors
Paul Allan BSc, DMRD, FRCR, FRCPE Fred T. Lee, Jr, MD

Consultant Radiologist and Clinical Director of Robert A. Turrell Professor of Medical Imaging
Radiology Department of Radiology
Department of Medical Radiology University of Wisconsin
Royal Infirmary of Edinburgh Madison, WI
Edinburgh, UK USA
Jonathan D. Berry BSc, MBBS, FRCR W. Norman McDicken PhD, FIPEM

Specialist Registrar Professor of Medical Physics and Medical Engineering
Department of Radiology Medical Physics
King’s College Hospital The University of Edinburgh
Denmark Hill Edinburgh, UK
London, UK
Imogen Montague MB ChB, MRCOG
Paul A. Dubbins BSc, MBBS, FRCR Consultant Obstetrician and Gynaecologist
Consultant Radiologist Department of Obstetrics and Gynaecology
Department of Radiology Derriford Hospital
Derriford Hospital Plymouth, UK
Plymouth, UK
Myron A. Pozniak, MD
Karen Gallagher AVS, BN Professor of Medical Imaging
Senior Clinical Vascular Scientist Department of Radiology
Vascular Laboratory University of Wisconsin
Royal Infirmary of Edinburgh Madison, WI
Edinburgh, UK USA
Peter R. Hoskins BA, MSc, PhD, FIPEM Paul S. Sidhu BSc, MBBS, MRCP, FRCP, DTM&H
Consultant Medical Physicist Consultant Radiologist and Senior Lecturer
Medical Physics Department of Radiology
The University of Edinburgh King’s College Hospital
Edinburgh, UK Denmark Hill
London, UK
viii
Preface
The first edition of Clinical Doppler Ultrasound brought up to date and one or two small errors
was designed as a practical introduction to the from the first edition have been corrected.
principles and practice of Doppler ultrasound Although other vascular imaging techniques,
in the clinical setting. Since then there have been such as computed tomography angiography
a number of advances which have required a and magnetic resonance angiography have seen
revision of the original text.These include signif- dramatic evolution during the last few years,
icant technical advances in ultrasound equip- the authors remain convinced of the primary
ment; the evolution of echo-enhancing agents role of Doppler ultrasound in the investigation
and their wider availability; and new applications of vascular disorders and this role has developed
of Doppler ultrasound have been developed. and expanded with time, rather than contracted.
In this second edition we have remained true It is important that those performing Doppler
to our initial aim to provide practical information ultrasound examinations have a clear under-
and advice on the practice and role of Doppler standing of the underlying principles of this
ultrasound. The contributors to this book all powerful tool and we have tried to encourage
have significant clinical experience in diagnostic such understanding with this book.
ultrasound and all the chapters have been re-
written in the light of more recent knowledge Paul L. Allan
and developments; a new chapter on echo- Paul A. Dubbins
enhancing agents has been included.The images W. Norman McDicken
have been updated and expanded, the references Myron Pozniak 2006
vii
1
Physics: principles, practice
and artefacts
W. Norman McDicken and Peter R. Hoskins
A number of techniques have been developed a moving structure within the body, it experiences
which exploit the shift in frequency of ultra- a Doppler shift in its frequency and returns to
sound when it is reflected from moving blood. the receiving (detecting) crystal. Reflected ultra-
This frequency shift is known as the ‘Doppler sound is also detected from static surfaces within
effect’.1 Five types of diagnostic Doppler instru- the body but it has not suffered a Doppler shift
ment are usually distinguished: in frequency. After the reflected ultrasound is
1. Continuous wave (CW) Doppler received, the Doppler instrument separates the
2. Pulsed wave (PW) Doppler signals from static and moving structures by
3. Duplex Doppler exploiting their different frequency.
4. Colour Doppler imaging (CDI; colour velocity Motion of the reflector towards the transducer
imaging) produces an increase in the reflected ultrasonic
5. Power Doppler imaging frequency, whereas motion away gives a reduc-
tion. The system electronics note whether the
The characteristics of an ultrasound beam, detected ultrasound has a higher or lower
the propagation of ultrasound in tissue and the frequency than that transmitted, and hence
design of transducers as found in B-mode imaging extracts information on the direction of motion
are all relevant for Doppler techniques.2–6 relative to the transducer.
When the line of movement of the reflector is
at an angle θ to the transducer beam, then the
THE DOPPLER EFFECT AND ITS
Doppler shift, fD, is given by:
APPLICATION
fD = ft – fr = ft .2.u. cosq
For all waves such as sound or light the Doppler
c
effect is a change in the observed frequency of
the wave because of motion of the source or where ft is the transmitted frequency, fr is the
observer. This is due either to the source received frequency, c is the speed of ultrasound
stretching or compressing the wave or the and ucosq (i.e. u × cosine q) is merely the
observer meeting the wave more quickly or component of the velocity of the reflecting agent
slowly as a result of their motion. In basic along the ultrasonic beam direction (Appendix
medical usage of the Doppler effect, the source A1). For a typical case of blood flow in a super-
and observer (receiver) are a transmitting and a ficial vessel:
receiving crystal usually positioned next to each Transmitted frequency, ft = 5 MHz = 5 × 106 Hz
other in a hand-held transducer (Fig. 1.1a). A Velocity of sound in soft tissue, c = 1540 m s-1
continuous cyclic electrical signal is applied to Velocity of blood movement, u = 30 cm s-1
the transmitting crystal and therefore a corre- Angle between ultrasonic beam and direction
sponding CW ultrasound beam is generated. of flow, q = 45°
When the ultrasound is scattered or reflected at The Doppler shift is therefore: 1
1
fD = (5 × 106 × 2 × 30 × cos45)/154 000
Physics: principles, practice and artefacts
the system electronics can extract a Doppler shift
= 1372 Hz signal from the samples. The Doppler equation
again applies to this Doppler shift and can be
The shift in frequency is small and within the used to calculate the speed of the reflector.7
audible range. In an ultrasonic Doppler instru- A bonus of PW Doppler is that since pulsed
ment, the electronics are designed to extract the ultrasound is employed, the range of the moving
difference in frequency, fD = ft – fr (the Doppler target may be measured from the echo-return
shift frequency). The instrument can therefore time, as well as its speed from the Doppler shift.
feed a signal of frequency fD to some output The range can be measured from one echo
device such as a loudspeaker, or frequency signal; however, the calculation of the Doppler
analyser as discussed later. shift and hence speed typically requires 50–100
So far we have considered an ultrasound echoes. As for the CW case, a group of blood
beam being reflected from a structure moving cells moving with different velocities produce a
at a fixed speed and hence generating a Doppler range of Doppler shift frequency components in
shift of one particular frequency. In practice, the output signal.
there are many reflecting blood cells and their It was noted above that the frequency of
speeds are different. The ultrasound signals reflected ultrasound is shifted upward or down-
returned to the detector from the different cells ward depending on whether the motion of the
therefore have suffered different Doppler shifts reflector is toward or away from the transducer.
and add together to give a complex signal con- A numerical example illustrates this point and
taining a range of frequencies. The Doppler shift emphasises the small changes in frequency that
frequencies are extracted from the detected the instrument must distinguish. When 2 MHz
complex signal and can be fed to a loudspeaker ultrasound is reflected from an object travelling
where they can be interpreted by listening. High- at 30 cm s-1 toward the transducer, it returns to
frequency (high-pitch) components in the audible the receiver with a frequency of 2.00078 MHz,
sound are related to high speeds, whereas low- a shift of +0.00078 MHz. If the object moves at
frequency components correspond to low speeds. 30 cm s-1 away from the transducer, the ultra-
Strong signals, that is of loud audible volume, sound returns with a frequency of 1.99922 MHz,
correspond to strong echoes that have received a a shift of -0.00078 MHz. Virtually all Doppler
Doppler shift. Strong signals could be due to the instruments which measure velocity preserve this
detection of many blood cells, say in a large direction information.
vessel, or to echoes from tissue. Later it is noted
that an output display called a spectral display
CONTINUOUS AND PULSED
or spectrogram is often used to portray the
WAVE DOPPLER INSTRUMENTS
frequency content of Doppler signals.
In the PW Doppler technique, the electrical Doppler blood flow instruments are required to
excitation signal is applied to the crystal as pulses, be extremely sensitive and to be capable of
each containing say 10 cycles, at regular intervals detecting weak signals from moving blood in the
and therefore a corresponding train of pulses of presence of much stronger signals from static
ultrasound are transmitted. For example, 10-cycle or moving tissues; the latter give rise to low-
pulses can be transmitted, separated by non- frequency Doppler shift ‘clutter’ signals. The
transmission intervals of duration 20 times that magnitude of the scattered signal from blood is
of each pulse. Regularly spaced echoes are then typically 40 dB below that received from soft
received back from a reflector and they can be tissues, i.e. the blood echo amplitude is typically
regarded as samples of the signal which would one-hundredth of the soft tissue echo ampli-
be received if a continuous wave had been trans- tude. The dB unit is a measure of the size of a
2 mitted as discussed above. If the reflector is moving signal relative to another signal; the second signal
1
Continuous and pulsed wave Doppler instruments

is often a reference signal or perhaps the input repetition frequency (PRF) of 10 000 s-1, i.e.
signal to an amplifier to which the output is 10 kHz. The highest velocity that the instrument
compared. Blood flow signals may be detected can measure is directly proportional to its PRF
even though the vessel is not clearly depicted, for (see aliasing artefact), therefore the PRF is made
instance in the fetal brain, or the renal artery of as high as possible while still avoiding overlap
the neonate. between successive echo trains. Echo signals, i.e.
The transducer of a basic CW Doppler unit trains of echoes, are produced as a transmitted
has two independent piezoelectric crystals. Since pulse passes through reflecting interfaces and
the transmitting crystal is continually driven to regions of scattering targets. After amplification,
generate a continuous wave of ultrasound, a successive echo signals from a specific depth are
second crystal is used to detect the reflected selected by electronic gating and the Doppler
ultrasound.When a CW Doppler mode is imple- shift frequency is extracted as described above.
mented as part of an ultrasound system which Pulsed Doppler devices can be used on their
uses array transducers, separate groups of array own by altering slowly the beam direction or the
crystal elements are used for transmission and gated range depth while listening to the output,
reception. On extraction of the Doppler shift for example in transcranial blood flow studies.
frequency a filter, the ‘wall-thump’ filter, is often Identification of vessels is made easier by com-
used to remove large, low-frequency compo- bining the PW Doppler mode with a real-time
nents from the signal, such as those from slowly B-scan mode to form a duplex system; however,
moving vessel walls. Typically in a Doppler unit this obviously adds to the cost and complexity.
operating at 5 MHz, Doppler shift frequencies Since the ultrasound is pulsed and the excita-
below 100 Hz are removed by filtering. Basic CW tion time is short, a stand-alone PW unit uses a
Doppler instruments are small and inexpensive; single crystal transducer for transmission and
CW Doppler mode facilities are incorporated reception (Fig. 1.1b). On setting the electronic
into some array systems to allow them to detect gate to select a signal from a specific range, reflec-
high velocities (see section on aliasing artefact). tors within a volume, known as the sample volume,
The transmitted ultrasound field and the zone contribute to the signal. The shape and size of
of maximum receiving sensitivity overlap for a the sample volume are determined by a number
particular range in front of the transducer (Fig. of factors: the transmitted pulse length, the beam
1.1a). Any moving structure within this region width, the gated range length, and the charac-
of overlap will contribute a component frequency teristics of the electronics and transducer. The
to the total Doppler signal. The shape of the sample volume is often described as a tear drop
region of overlap (the beam shape) can be in shape (Fig. 1.1b). Sample volume lengths are
considered as having a crude focus which depends usually altered by changing the gated range length.
on the field and zone shapes and on their angle In a blood flow unit for superficial vessels, the
of orientation to each other. In practice, the beam sample volume length may be as short as 1 mm,
shapes are rarely well known for CW Doppler whereas in a transcranial device it can be 1 or 2 cm;
transducers. A 5 MHz blood flow instrument however, the precise lengths are rarely known.
might be focused at a distance of 2 or 3 cm from The ultrasonic output intensity of pulsed
the transducer and a 10 MHz device at a distance Doppler instruments varies considerably from
of 0.5–1 cm. CW Doppler instruments normally unit to unit.The intensity (Ispta -see safety section,
have ultrasonic output intensities (Ispta) of less pp. 23–24) may typically be a few hundred
than 10 mW cm-2 although they may be signifi- mW cm-2 but can be as high as 1000 mW cm-2,
cantly higher when used in conjunction with particularly when they are required to penetrate
duplex systems to measure high velocities. bone, as in transcranial Doppler. At present the
A PW Doppler instrument, operating with most common use of stand-alone PW units is in
5 MHz ultrasonic pulses, may have a pulse transcranial examinations of cerebral vessels. 3
1
Fig. 1.1 Sample volumes in
Doppler techniques. (a) For dual
crystal continuous wave Doppler
Extended sample volume
unit. (b) For pulsed wave Doppler
unit. (c) Neighbouring sample
volumes along a beam for imaging
(a) Doppler units.
Tear-drop sample volume
(b)
Consecutive sample volumes
(c)
Technical factors in the use of CW and signal from vessel walls and any other
PW Doppler moving tissue.
1. Doppler beams are subject to the same 7. The final result in many cases should be a
physical processes in tissue as B-mode distinct display, called a ‘spectrogram’ or
beams, i.e. attenuation, refraction, speed of ‘sonogram’ (see section on the spectrum
sound variation, defocusing, etc. analyser), with a clearly defined maximum-
2. Since the stand-alone CW and PW units are velocity trace.
used blind, the beam direction and also the 8. Since the beam–vessel angle is unlikely to be
sample volume in the PW case must be known, the sonogram cannot be calibrated in
systematically moved through the region of velocity and the vertical axis remains as
interest to maximise both the volume and Doppler shift frequency.
pitch of the audible Doppler signal. 9. Care should be taken to ensure good acoustic
3. PW Doppler is subject to the aliasing arte- coupling between the transducer and the
fact in the measurement of high velocities, patient. Since there is no associated image it
CW Doppler is not. is not always apparent that a weak signal
4. The sensitivity (gain, transmit power) of the may be due to a lack of coupling agent.
Doppler unit should not be so high that 10. If possible, information should be obtained
noise detracts from the signal quality. on the shape of the sample volume for both
5. The instrument should be assessed on CW and PW beams.The sample volume size
normal vessels where the blood flow pattern can then be related to the size of the vessel
is known and the expected Doppler signal under study.With CW Doppler there is very
well understood. little depth discrimination.With PW Doppler
6. The wall-thump filter should just be high the sample volume depth and size are set by
enough to remove the strong low-frequency the user.
4
1
Imaging and Doppler

measured without the aliasing artefact, for
IMAGING AND DOPPLER
example in the estimation of the velocity of a jet
There are three types of imaging used with through stenosed heart valves. The direction of
Doppler techniques. The first, known as ‘duplex the CW beam is shown as a line across the
Doppler’, uses a real-time B-scanner to locate B-scan image. In the case of PW Doppler, a
the site at which blood flow is to be examined marker on the beam line shows the position of
then a Doppler beam interrogates that site. The the sample volume. The same transducer is
second type creates an image from Doppler usually employed for both imaging and Doppler
information, i.e. an image of velocities in regions but two separate ones may be linked together.
of blood flow.8 Known as ‘colour Doppler’, The Doppler beam is often directed across the
‘colour Doppler imaging’ or ‘colour velocity field of view so that it does not intersect the
imaging’ (CVI), it is normally combined with a blood flow at 90°.
conventional real-time B-scan so that both tissue In practice it is difficult to run the B-mode
structure and areas of flow are displayed. The display and the Doppler facility simultaneously,
third type of Doppler imaging is similar to colour as pulses from one unit are picked up by the
Doppler but generates an image of the power of other. A time-share solution is employed by many
the Doppler signal from pixel locations through- manufacturers, whereby most time is spent in
out the field of view and is known as ‘power the Doppler mode and the B-mode image is only
Doppler imaging’ (power Doppler).9 A power refreshed at short time intervals. The operator
Doppler image depicts the amount of blood can then check that the ultrasound beam is still
moving in each region, i.e. an image of the intersecting the site of interest. Another option
detected blood pool. is to switch off the imaging mode once the
Doppler beam direction has been fixed and to
Duplex instruments maximise the PRF for the PW Doppler.
Duplex systems link CW or PW Doppler features In duplex systems, the transmitted ultrasound
and real-time B-scanners so that the Doppler frequency in the Doppler mode is often lower
beam can interrogate specific locations in the than that for the B-mode.The low Doppler beam
B-scan image (Fig. 1.2). CW duplex is normally frequency is to enable higher velocities to be
only used where very high velocities have to be handled before aliasing occurs, while the high
Fig. 1.2 Duplex system combining

real-time B-mode and a Doppler
beam of variable position across
the B-mode field of view.
Imaging and
Doppler transducer
Field of view
Doppler beam
5
1
B-scan frequency is to optimise resolution in the direction of flow is obtained by examining the
image. An example could be 5 MHz for Doppler signals for the direction of the shift as for CW
and 7 MHz for B-mode when studies of super- and PW Doppler devices. Each image pixel is
ficial vessels are undertaken. then colour-coded for direction and mean
Doppler shift e.g. shades of red for flow towards
Technical factors in the use of duplex the transducer and shades of blue for flow away
Doppler (Fig. 1.3a).
1. The factors quoted for CW and PW Doppler B-scanning and Doppler imaging are carried
may also be relevant. out with the common types of real-time trans-
2. A spectrogram can be obtained from a ducer. Echo signals from the blood and tissues
known vessel and a known location within are processed along two signal paths in the
the vessel. system electronics (Fig. 1.4). Going along one
3. The Doppler beam may be refracted and not path, the signals produce the real-time B-scan
pass along the line shown in the B-mode image. image; going along the other path, autocorrela-
4. The beam–vessel angle may be measured tion function processing and direction flow
manually, allowing blood velocity to be sensing are employed to give a colour flow image.
estimated. An important exclusion circuit in the autocorre-
5. Simple estimates of blood flow may be lation path separates large-amplitude signals
obtained from the measured diameter and which arise from tissue and excludes them from
mean velocity. the blood velocity processing. The B-mode and
6. Spectral broadening due to the use of wide mean velocity images are then superimposed in
aperture transducers to give good focusing the final display. Strictly speaking, the flow image
can cause large errors in the measurement of is of the mean Doppler shift frequency and not
maximum velocity. the mean velocity, since the beam–vessel angles
7. Since the Doppler beam is held fixed and the throughout the field of view are not measured.
PRF is high, particular attention should be Colour shades in the image can indicate the
paid to the outputs of PW units in relation to magnitude of the velocity, for example light red
safety. for high velocity and dark red for low velocity.
Turbulence, related to the range of velocities in
Colour Doppler imaging each sample volume, may be presented as a
Pulsed Doppler techniques require between 50 different colour or as a mosaic of colours.
and 100 ultrasonic pulses to be transmitted in Doppler images typically contain about 64
each beam direction for the determination of genuine lines of information and 128 consecu-
velocities of blood in a sample volume. It is tive sample volumes along each line. The frame
therefore not possible to move the beam rapidly rate varies from 5 to 40 frames s-1, depending on
through the scan plane to build up real-time the depth of penetration and the width of the
Doppler images of velocity of flow. Such imag- field of view. As in B-scanning, the appearance
ing became possible when signal processing was of the image is usually improved by inserting
developed which could quickly produce a measure additional lines or frames whose data are calcu-
of mean blood velocity at each sample volume lated from the genuine lines, a process known
from a small number of ultrasonic echo pulses. as interpolation. Alteration in flow can occur
A technique called ‘autocorrelation processing’ rapidly over the cardiac cycle, therefore a cine-
of the signals from blood quickly gives the mean loop store of say the last 128 frames is of value
velocity in each small sample volume along the for review purposes. Doppler spectrograms can
beam (Fig. 1.1c). This real-time colour Doppler be made by selecting the appropriate beam direc-
imaging processes between 2 and 16 echo signals tion and sample volume location in the image
6 from each sample volume. In addition, the and then switching to the PW or CW mode. PW
a
Fig. 1.3 (a) Colour flow image

1
Imaging and Doppler

of flow in a straight tube.
When the flow velocity
component is along the beam
towards the transducer it is
colour-coded red, when the
component along the beam is
away from the transducer it is
coded blue. (b) Power Doppler
image of flow in a straight tube.
Direction of flow is not
measured so it is not colour-
coded.
and CW Doppler techniques provide more colours depicted in the image are therefore
detailed information on blood velocities than heavily angle dependent.
colour Doppler, so spectral information is still 3. The flow pattern on the colour Doppler
of value. display can be related to the structures shown
in the B-mode image.
Technical factors in the use of colour 4. CDI is a pulsed technique so aliasing is a
Doppler imaging problem.
1. The mean Doppler frequency is the quantity 5. A good CDI machine is one which discrimi-
which is presented in a colour-coded form in nates well between signals from tissue and
each pixel. When the colour bar is labelled in those from blood.
velocity the beam–vessel angle has been 6. The CDI field of view box should be adjusted
assumed to be zero throughout the image. to cover only the region of interest and there-
2. The velocity component along the Doppler fore maximise the frame rate.
beam is heavily dependent on the angle 7. The velocity range covered by the colour scale
between the direction of flow and the beam should be carefully matched to the velocities
direction (the cosine q dependence). The expected in the study. 7
1
Echo
image
processor
Ultrasound
signals
from body Doppler Display
direction system
detector
Doppler
velocity
estimator
Fig. 1.4 B-mode and Doppler imaging signal processing paths in a scanning machine.
8. A cine-loop is useful for the review of fast- power of the background noise plus Doppler
changing blood flow patterns. signal when blood flow is present. The back-
9. A change in the direction of flow, say from ground noise may be used to set a threshold
toward to away from the transducer, and above which signals are accepted for Doppler
hence a colour change, need not mean a flow. Noise from sample volume regions lacking
change in the direction of flow along a vessel. blood flow is therefore reduced in the power
It may merely mean the beam–flow angle has image by a threshold detector. However, when
changed from less than 90° to more than 90°. the same signal is used in the velocity imaging
mode, the noise will produce a mean velocity
value which the machine will treat as a genuine
Power Doppler imaging blood velocity and which will therefore appear in
The power of the Doppler signal from each the image.The power Doppler mode is therefore
small sample volume in the field of view may be less prone to noise and hence more sensitive and
displayed rather than the mean frequency shift can be used to detect small vessels. Further
(Fig. 1.3b). The power of the signal from each sensitivity can be obtained by averaging power
point relates to the number of moving blood images over several frames to reduce spuriously
cells in that sample volume. The power Doppler distributed noise even more. In velocity imaging
image may be considered to be an image of the there is interest in showing quick changes in
blood pool. The power mode does not measure blood flow and hence less averaging is used.
velocity or direction and therefore the image Power Doppler imaging is fairly easy to use
shows little angle dependence, neither does it and often provides a more complete image of
suffer from aliasing; however, it obviously the vasculature than velocity imaging. This has
presents less information about blood flow. The made it popular in clinical use and it is
attraction of power Doppler images is that they commonly used initially to locate regions of
suffer less from noise than velocity images, as the interest prior to investigation by colour Doppler
power of the background noise for any sample or duplex methods. It is also possible with some
8 volume with no blood flow signal is less than the equipment to use the direction information in
1
Ultrasonic microbubble contrast agents

the colour Doppler image to colour-code the range of medical ultrasound systems. Bubbles of
corresponding power image. these dimensions are important since, even with
only very thin wall encapsulation, they are able
Technical factors in the use of power to pass through the capillaries of the lung into
Doppler imaging the systemic circulation. An investigation by a
1. In the power Doppler image, the power of the committee of the American Society of Echo-
Doppler signal at each pixel is colour-coded. cardiography concluded that contrast echocardio-
2. No velocity information is displayed. graphy carried a minimal risk for patients and
3. Most power Doppler systems are direction that there were few residual or complicating
insensitive; i.e. the display is the same regard- side-effects.11 More recent individual studies
less of whether the blood is flowing toward or have confirmed these conclusions; however, more
away from the transducer. However, some work is required on new agents as they become
machines do display different colours depend- available.12
ing on the direction of blood flow by includ- The development around 1990 of contrast
ing some directional Doppler information. agent microbubbles that can be used by percu-
4. The power Doppler image is insensitive to taneous venous injection was the breakthrough
angle except near 90° where the Doppler which gave rise to the current high level of
signal may fall below the clutter filter, which activity in this field. Table 1.1 gives examples of
cuts out low Doppler shifts, and no signal is agents which are currently under commercial
displayed. development. The ‘Echogen’ agent is different
5. There is no aliasing because the frequency from the others in that it undergoes a phase
information (i.e. velocity) is not estimated change from liquid to gas as a result of the
from the Doppler signal. increase in temperature following injection into
6. The power Doppler image is very sensitive to the bloodstream. Large gas molecules are
movement of the tissue or probe (the flash encapsulated in some agents to reduce the rate
artefact). Some machines incorporate a flash of diffusion and so increase the lifetime of the
filter to try to reduce this effect. bubbles. Typically the lifetime in blood ranges
from 2 to 3 min up to 20–30 min. An attraction
of contrast agents is the ability to increase the
ULTRASONIC MICROBUBBLE
signal obtained from small blood vessels which
CONTRAST AGENTS
are difficult to detect by conventional Doppler
A number of agents have been considered which methods, e.g. cerebral or renal vessels. There is
can enhance the scattering of ultrasound from also interest in perfusion studies, for example to
blood and hence could be employed as echo- observe and measure the wash-in and wash-out
enhancing or contrast agents. Ophir and Parker10 of agent in the myocardium in a manner
have reviewed contrast agents. From this review analogous to nuclear medicine studies.
and more recent experience it has become Enhanced scattering is obtained if a bubble is
obvious that agents in the form of encapsulated insonated with ultrasound of a frequency equal
microbubbles are by far the most likely to be to that of the fundamental resonance frequency
successful echo-enhancing agents in the imme- of the bubble. At low power (strictly – low ultra-
diate future.This is due to the large difference in sonic wave pressure amplitude) the oscillations
acoustic impedance between the gas in the of the bubble are about the centre of the bubble
bubbles and the surrounding blood. In addition, and are directly in proportion to the size of the
bubbles of a few microns in diameter have a pressure fluctuations in the ultrasound wave.
fundamental resonance frequency of a few However, at higher powers the oscillations
megahertz. For example, 4 μm diameter bubbles become distorted and ultrasound at frequencies
resonate at 4 MHz, which is well within the different from that of the incident wave is 9
10
1
Table 1.1 Properties of some common intravenous, lung-crossing contrast agents
Agent Manufacturer Type of Agent Capsule Gas Bubble size Dose (Concentration)
DEFINITY Bristol-Myers Squibb Microsphere Bi-layer Lipid Octafluoropropane Mean diameter 10 × 108 μbubbles/ml
Medical Imaging 1.1–3.3 μm
(98% <10 μm)
OPTISON Amersham Health Microsphere Albumin Octafluoropropane Mean diameter 5–8 × 108 μbubbles/ml
range
3–4.5 μm
(95% <10 μm)
SONOVUE™ Bracco Stabilised Phospholipids SF6 Mean diameter 2–5 × 108 μbubbles/ml
microbubble 2.5 μm
(90% <6 μm)
CARDIOSPHERE Point Biomedical Bi-layer Albumin – outer Nitrogen but gas Variable Variable
Corporation microbubble Caprolactone – inner can be varied
Courtesy of C. Moran.
1
Information from Doppler signals

generated by the bubbles. These frequencies are backscatter due to bubble destruction. This
known as harmonics and are simply related to phenomenon is also known as ‘acoustically
the fundamental resonance frequency of the stimulated emission’. Aspects of its safety still
bubble, e.g. the second harmonic frequency is need to be fully evaluated.
twice the fundamental frequency. There is 5. Determine the physical phenomena which
considerable interest in detecting and using the are expected to be present under the operating
second harmonic, since tissue does not produce conditions and which will enhance the echo
this effect to any great extent and the second signal, e.g. harmonic scattering, free bubble
harmonic signal comes predominantly from the creation and bubble destruction.
echo-enhancing agent in the blood vessels. Both
pulse-echo and Doppler systems have been
INFORMATION FROM DOPPLER
designed to pick out the second harmonic com-
SIGNALS
ponent in the ultrasound returned to the trans-
ducer and use it to enhance the signal from When the Doppler signal which contains infor-
the agent in blood, possibly by as much as mation on blood velocities and haemodynamics
20–30 dB. These systems are being evaluated in is being produced it is necessary to interpret it.
clinical practice.13 It is essential to obtain good-quality signals in
The scattering from contrast agents can also order to be able to detect disease.
be enhanced if the acoustic pressure fluctuations
in the beam are large enough to damage the The spectrum analyser
microbubbles, causing them to leak. An unencap- A Doppler signal may be analysed into its
sulated gas bubble then forms next to the original frequency components in order to give a display
one; however, since it has no outer shell, the of the velocities of the blood cells at each instant
scattering from it is undamped and can be around (Fig. 1.5).2 Short time intervals of the Doppler
1000 times higher than that from the encapsu- signal are analysed, for example a segment of
lated bubble. This effect has been exploited in a 5 ms duration. This produces an instantaneous
technique known as ‘intermittent’ or ‘transient’ spectrum of the frequencies in the sample
imaging, which allows time for the damaged volume. If an angle correction is then applied,
bubbles to be replaced between sweeps of the this spectrum will represent the range of velocities
ultrasound beam.14 in the sample volume. The range of frequencies
in each spectrum is then displayed on the vertical
Technical factors in the use of axis and the power of each frequency component
microbubble contrast agents is presented as a shade of grey. The consecutive
Contrast agents are still very much in the develop- velocity spectra are displayed as side-by-side
mental stage. Nevertheless, some technical points grey-shade lines. In this way a spectral display,
of importance are becoming apparent. or spectrogram, is built up (Fig. 1.6). Note the
1. Check the age and shelf-life of the agent. difference between the instantaneous velocity
2. Microbubble contrast agents are fairly fragile. spectrum which tells us about the pattern of
The handling instructions should be carefully velocities in the sample volume at that instant
followed and indeed with experience some and the spectral display or spectrogram, which
additional precautions may be identified. shows how the velocity pattern varies with time.
3. The ultrasonic beam may be capable of Spectrograms are generated in real time during
destroying some microbubbles; higher transmit the clinical examination and it is usually possible
power will therefore not necessarily produce to store a few seconds of the trace in an analyser
a stronger echo signal. for subsequent review.
4. Several contrast agents require to be insonated The temporal resolution in a spectrogram,
at high acoustic pressures to give enhanced that is the smallest discernible time interval, 11
1
Volts
Audio
Doppler
signal
5 ms
Instantaneous
spectra
Amplitude
Frequency
Sonogram
Frequency
Time
Components of sonogram
Fig. 1.5 Schematic representation of the analysis of a Doppler signal to form a sonogram (spectrogram).
From McDicken2 with permission.
equals the length of the portion of Doppler

signal used to produce each instantaneous spec-
trum and is typically 5–10 ms.The frequency (or
velocity) axis of the spectrogram usually has
about 100 scale intervals each of 100 Hz. The
Doppler signal is therefore resolved into frequency
components separated by 100 Hz, the frequency
resolution of the spectrogram.
It is often desirable to make measurements on
a spectrogram, for example an estimate of the
time between two events or of the maximum Fig. 1.6 A sonogram illustrating pixel structure and
velocity during systole or diastole. Measurement speckle pattern.
12
1
Spectrograms (Sonograms) and indices

is performed by placing a cursor on the relevant
SPECTROGRAMS (SONOGRAMS)
points of interest then a variety of calculations
AND INDICES
can be performed by the system computer.
Indices related to the spectrogram shape and A good-quality audible Doppler signal will result
hence to normality or abnormality of flow in a good quality sonogram. Degradation of a
velocities can be calculated within the analyser sonogram is due to electronic and acoustic noise
and displayed on its screen. These are discussed which should only be a problem when the gain
later in this chapter. The oscillatory shape of a of the Doppler unit or analyser is very high for
spectrum or a trace derived from it is often the detection of weak signals. Noise in sonograms
referred to as a waveform. creates considerable difficulties for the automatic
calculation of quantities and they should there-
Technical factors in relation to spectral fore be used with caution. It is sometimes possible
analysis to check traces drawn automatically on a sono-
1. Arrange the frequency and time scales to best gram to see if they have been corrupted by
display the detailed information in the noise. If the automatic mode cannot deal with
sonogram by adjusting velocity scale, base- the noise, then the traces should be drawn
line and sweep speed. manually making use of the eye’s ability to
2. Arrange the grey tones, or colours, of the distinguish noise from true signal. It should also
spectral display to give the best ‘image’ quality be remembered that the vertical axis of a sono-
by adjusting the Doppler gain. gram can only be labelled as velocity after the
3. Treat with caution information from weak beam–vessel angle has been measured.
signals in a noisy background.
4. A clear spectrogram is required before the Waveform indices
maximum frequency can be traced. Waveform indices are derived from a combina-
5. It is usually best to make measurements over tion of a few dominant features of the wave-
at least five heart cycles. However, if this is form.3 Indices that have the same or similar
not possible, perhaps due to respiratory move- names in the literature are occasionally defined
ment, useful information can be obtained from differently, so a first step is to check the
one or two cardiac cycles. definition of any index to be used. In practice
6. Make use of the total storage capacity of the only two classes of index are used to any great
analyser to make a full recording from which extent, those related to the degree of diastolic
the most suitable part of the spectrogram can flow and others related to spectral broadening.
be selected, using the scroll facility, in order The variation in time of the maximum velocity
to make measurements. displayed in a spectrogram is commonly used as
7. Check thoroughly the reliability of automatic a source of data for the derivation of an index
tracing techniques for maximum and mini- (Fig. 1.7a). Since the maximum velocity is not
mum velocity. always clearly apparent in a spectrogram, some
8. Note whether the vertical axis has been analysers produce a trace which is closely related
calibrated in velocity by allowing for the to the maximum velocity trace. One example is
beam–vessel angle. a trace showing the upper velocity boundary
9. To obtain reproducible results, try to use the below which the velocity components contain
same beam–vessel angle for all examinations, seven-eighths of the power of the Doppler signal.
e.g. 60°. The mean velocity waveform (average velocity
waveform) is also employed (Fig. 1.7b).To calcu-
late the mean velocity at each instant, the values
13
1
a Resistance index (RI)
In Fig. 1.8:
S–D
RI =
S
High resistance in the distal vessels produces

low diastolic flow in the supplying artery and
results in a high value for this index; a low
b resistance results in a low value as there is higher
diastolic flow. It is also known as the Pourcelot
index.
Pulsatility index (PI)

The pulsatility index (PI) is defined as:
maximum velocity excursion

PI =
mean of the velocity
Fig. 1.7 (a) A maximum-velocity trace on a
sonogram. (b) A mean velocity on a sonogram. It is used in vessels where reverse flow may
occur, for example in the lower limbs (Fig. 1.8).
The PI may typically have a value of 10 for the
of velocity and the intensities of the signal for normal common femoral artery but be around 2
each velocity component in the instantaneous when proximal disease severely dampens the
spectrum are used. The mean velocity is used waveform.
together with the vessel cross-sectional area to As defined above, the PI index is heart rate
calculate blood flow rate. However, it is difficult dependent. To avoid this, PI can be calculated
to measure mean velocity accurately and there over a specified time from the start of systole,
are several other problems associated with calcu- e.g. for the first 500 ms. The pulsatility index is
lating flow rates; these are discussed further in then labelled ‘PI(500)’.
Chapter 2.
Since the beam–vessel angle may not always Damping factor
be known, the waveforms or spectrograms will The damping factor is defined as the ratio of
not be corrected for angle. Indices are therefore pulsatility indices at two sites along an artery. It
defined involving ratios of velocities. In such a quantifies the damping of the waveform down-
ratio, the angle factors appear on both the top stream along a diseased vessel.
and bottom and hence cancel each other out, so
PI (proximal site)
that the index is independent of beam angle. Damping factor =
PI (distal site)
Errors are also reduced by averaging the calcu-
lated indices over several heart cycles. The numerical value of this index increases as
A number of the most commonly encoun- disease becomes more severe, a value of 2 being
tered indices are briefly discussed below. typical of a high degree of damping. This index
is mentioned for completeness, it is not widely
A/B ratio used.
The A/B ratio is defined as the ratio of two
specified velocities, e.g. maximum velocities, at Spectral broadening
two points in the cardiac cycle (Fig. 1.8). It is Turbulence increases the range of blood cell
usually employed where there is no reverse flow velocities in a vessel. One index to quantify this
14 in the waveform. broadening of the spectrum of velocities is:
Fig. 1.8 Waveform indices,

1
Spectrogram (Sonograms) and indices

(a) normally calculated from a
maximum velocity waveform,
but the mean velocity waveform
can also be used. From McDicken2
Velocity
with permission.
M
(mean velocity
of waveform)
A or S B
D
Time
T
(b)
S
Velocity
M
(mean velocity
of waveform)
D
T Time
D¢
S-D
RI =
S
S/ ratio (umbilical and uterine arteries)
D
A/ ratio (carotid arteries)

B
PI = Maximum velocity excursion

Mean velocity
= S-D in (a)
M
S+D¢
= in (b)
M
Spectral broadening index (at systole) can cause spectral broadening; for example
‘geometrical spectral broadening’, which results
maximum velocity
= from the range of Doppler angles with which
mean velocity
ultrasound can insonate a sample volume in the
Conclusions with regard to the presence of vessel from points on the face of the transducer.
turbulence should be made with caution and only
after familiarity has been gained with the patterns Transit time
for laminar and plug flow for the particular The time for the pulse pressure wave to travel
instrument being used. Other technical factors along a length of artery can be measured by 15
1
placing a Doppler probe at either end of it. any more information than can be deduced
From this time and knowing the length, the by direct observation of the whole sonogram.
pulse wave velocity (PWV) is calculated. Alter-
natively, using the electrocardiogram (ECG) and
ARTEFACTS IN DOPPLER
one Doppler unit the transit time can also be
TECHNIQUES
measured, the QRS of the ECG giving the time
when the pressure pulse leaves the heart. The The most important artefacts are mentioned
time is first measured for the pulse to travel from here and methods of dealing with them are
the heart to the proximal artery site; a second suggested. Further details can be found in other
time is then measured for the pulse to go from texts.2, 5 Artefacts are usually dealt with by
the heart to the distal site. Subtracting these two explaining their origin or by recognising that
times gives the transit time and, if the length of they occur fairly frequently and are not of
the artery is known, the PWV can be calculated. significance.
Any error associated with the assumption that
the QRS represents the time at which the pulse Attenuation
leaves the heart is removed by the subtraction. The reduction in echo signal size due to
A normal aorta has a PWV of around 10 m s -1. attenuation of the beam in tissue is very familiar
The transit time along 0.5 m is then 50 ms. Pulse from B-mode imaging. The same attenuation
wave velocity depends on disease states of the processes occur for Doppler techniques, hence
artery wall and blood pressure.This index is also stronger signals are detected from superficial
mentioned for completeness, it is not widely vessels than from deep ones. With CW Doppler
used. units this imbalance cannot be compensated.
With PW stand-alone and duplex Doppler
Technical factors in the application of devices, the signals are from a sample volume at
waveform indices a selected depth and the gain can therefore be
1. Use good-quality spectral data to calculate adjusted to optimise the signal. In colour
indices. Doppler and power Doppler imaging, time gain
2. Most indices are calculated using the maxi- compensation could help to compensate for
mum Doppler frequency as this is relatively attenuation but it is more usual just to process
insensitive to beam–vessel alignment and all signals that are above the noise level;
sample volume size. obviously those from deep vessels will be closer
3. Mean Doppler frequency shift is very sensi- to the noise level and hence will be more likely
tive to alignment and sample volume size and to be affected by noise.
is not widely used to estimate indices.
4. Indices of waveform shape are insensitive to Refraction
the beam–vessel angle except near 90° where Refraction deviates a beam as it crosses at an
the end-diastolic component may disappear angle the interface between two tissues in which
below the high-pass filter. the speed of sound is different. The direction of
5. For obstetrics use, the wall-thump filter must the transducer axis may not, therefore, coin-
be set low, typically 50–100 Hz, in order that cide with the actual beam path. With duplex
absent end-diastolic flow may be ascertained systems a weaker signal than expected from a
unambiguously. well-imaged vessel is probably due to refraction
6. The most suitable index or measurement for of the Doppler beam. This is less of a problem
each application can be determined from the with colour or power Doppler imaging instru-
literature. Indices are useful in that they allow ments, since the presence of a signal is noted
velocity patterns to be classified and related first in the image before any spectral analysis
16 to disease state. However, they do not extract is attempted.
1
Shadowing and enhancement Inadequate coupling
Artefacts in Doppler techniques

Attenuation of a Doppler beam at a structure Weak Doppler signals can be attributed to a lack
may be so large that blood flow behind it cannot of coupling liquid between the transducer and
be detected, for example at a calcified plaque on the skin. In CW and PW Doppler applications
a vessel wall. Microbubble contrast agents can this artefact is not as evident as in imaging tech-
also give rise to shadowing problems behind niques, where poor penetration is readily seen.
them. Signal enhancement occurs where the
beam passes through a medium of low attenu- Electrical pick-up
ation to reach the vessel, such as a collection of Doppler devices are required to be highly sensi-
amniotic fluid, or the full bladder. tive and are therefore prone to electrical pick-up
of stray signals. Some such signals can be recog-
Beam width nised by their pattern in the sonogram. The
A wide beam can cause contributions to the spectrum analyser unit may allow the operator
Doppler signal from moving structures well off to attempt to clean up the sonogram by deleting
the central axis. This is most likely to be due to spurious signals.
a strong reflector such as a heart valve leaflet,
but it could also be due to a large blood vessel. Compression of the spectral display
A narrow beam may result in only partial When a sonogram is compressed, either in its
insonation of a vessel with the related errors in grey shades or in its scale of presentation, infor-
the Doppler signal; for example, overemphasis mation is lost. A sonogram should be treated as
of the high-velocity flow at the centre of a vessel a type of image and thus presented with
occurs when a narrow beam is directed into a optimum grey-shade contrast and spatial detail.
vessel but does not encompass the slower- The latter means that the velocity and time axis
moving blood at the side. scales should be selected to show detailed struc-
ture in the sonogram. The accuracy of measure-
Spectral broadening ments from the sonogram will also be affected
Spectral broadening is another artefact resulting by poor presentation.
from beam shape. As noted in the discussion on
the spectral broadening index, this arises due to Erroneous direction sensing
ultrasound in the beam insonating a sample The direction-sensing circuitry does not always
volume over a range of angles. function correctly since its design and imple-
mentation are difficult. Flow will then be
Speckle and the spectral display presented in the wrong direction, either in a
The speckled appearance of a sonogram results sonogram or in a colour Doppler image; if this is
from fluctuations in the power levels of the suspected then the set-up of the system can be
velocity components in neighbouring pixels checked by examining a normal artery in which
(Fig. 1.6). These fluctuations are due to fluctu- the direction of flow is known.
ations in the ultrasonic signal received from the
random distribution of blood cells. Due to this Filtering
speckle noise the power level in a pixel cannot be Filters are used to greatly reduce low frequen-
directly related to the number of cells moving cies, such as those obtained from arterial walls.
with a particular velocity. Averaging the power Filters also remove information on slow-moving
levels in neighbouring pixels gives a more accu- blood but this is not usually a serious problem
rate measure of the number of cells moving with unless it is desired to measure mean velocity
each velocity. accurately or slow flow specifically (Fig. 1.9).
17
1
a
b
Fig. 1.9 Filtering. Raising the high-pass (wall-thump)
filter as in the lower sonogram removes low-velocity
signals.
Harmonic generation by large signal

distortion
The harmonics of a frequency are higher multiples
of that frequency; for example, harmonics of
100 Hz are 200 Hz, 300 Hz, etc. If a signal is too
large to be handled by the electronics it becomes
distorted and then contains additional harmonic
frequency components. When such a distorted
signal is analysed, the harmonics appear at
c
regular frequency intervals in the spectral display.
Strong blood flow signals exhibit harmonic
components as a higher frequency part of the
sonogram above that which would probably be
obtained if the gain were reduced (Fig. 1.10).
High or low sensitivity

In colour Doppler and power Doppler imaging,
setting up the system with too low a sensitivity
causes blood flow signal to be lost. Too high a
sensitivity causes spurious echoes to be colour-
coded as blood (Fig. 1.11).
Aliasing
Pulsed Doppler and colour Doppler units have
to reconstruct the Doppler shift signal from
regularly timed samples of information, rather
than the complete signal as used in CW units. Fig. 1.10 Increasing gain from (a) to (c). Very high
The sampling rate is equal to the PRF of the gain distorts the signal and introduces harmonic
18 Doppler unit. If the sampling rate is too low, the frequencies.
a
1

frequency of the reconstructed Doppler signal is
in error and the direction of flow is presented
wrongly. In a spectral display or flow image,
this is known as an ‘aliasing’ artefact (Figs 1.12
and 1.13).
Fig. 1.11 Variation of gain in a Doppler image.

The image content is very sensitive to gain setting:
(a) shows too much gain, (b) too little.
Fig. 1.12 Aliasing in a sonogram. The highest
velocities in (c) appear in the reverse channel.
19
1
Physics: principles, practice and artefacts a
The aliasing artefact is encountered when high-
frequency Doppler shift signals are produced,
usually by high-velocity flow. It also occurs when
sampling deeper vessels, as the PRF is reduced
to allow time for the echoes from a pulse to
return before the next pulse is transmitted. If the
PRF is too low for the Doppler shift frequencies
from the blood in the vessel, aliasing will occur.
An approach to raising the velocity level at which
aliasing becomes a problem is to use a high PRF
(the high-PRF mode), even although the echoes
from deep structures have not died out before
the next pulse is transmitted. If the deep echoes
are strong enough, Doppler signals from deep
b
vessels may then be superimposed on those from
a more superficial site. This uncertainty with
regard to the source of the signal is referred to as
‘range ambiguity’. The high-PRF mode must
therefore be used with care. Although this mode
can be useful, it increases the intensity of the
beam, which is another reason for using it only
when necessary.
Figures 1.12 and 1.13 show the aliasing arte-
fact.The high-velocity forward flow components,
above the upper limit, appear as high-velocity
reverse flow. Aliasing can be of value in colour
Doppler imaging since it allows high-velocity
jets to be identified. Power Doppler does not
c
suffer from aliasing.
Effect of beam angle to flow

direction
The quality of a Doppler signal depends on
beam–vessel angle and, in particular, above 70°
it degrades quite rapidly (Fig. 1.14). If the direc-
tion of an ultrasound beam is at 90° to the
direction of the flowing blood, no Doppler signal
is expected since, in the Doppler equation,
cos90° = 0. However, a poor-quality Doppler
signal is usually obtained for two reasons. First,
the ultrasound beam may converge or diverge
slightly from the beam axis, so all of it is never at
Fig. 1.13 Aliasing in a colour Doppler velocity
90° to the flow. Second, there may be some
image. Note that there is no black space between
the red and blue pixels in the lateral colour displays,
turbulence in the flow, in which case the blood
indicating that the pixel colour-coding has gone off cells are not all travelling in parallel paths at 90°
the top of the red colours and wrapped round to the to the beam.
blue colours. However, there is a black space between Colour Doppler images obtained at 90° to the
20 the colour displays centrally, indicating reversal of direction of flow appear dark or noisy, corre-
flow direction in relation to the transducer. sponding to absent or slow flow (Fig. 1.3).
a
1

angle which is particularly common in sector
scan imaging.
Effect of velocity scale

The choice of velocity scale can dramatically
change the appearance of a colour flow image
(Fig. 1.15).The scale should be chosen to accom-
modate the range of velocities thought to be
present. Too low a scale will cause aliasing and
too high a scale results in the flow being depicted
as a few dark colours in colour Doppler imaging.
b
Unexpected machine artefacts
Doppler technology is developing rapidly and
can still have gremlins in it. The operator must
therefore check the performance and calibration
of the instrument. This is most readily done in a
situation where the flow pattern is considered to
be well understood, such as in a clearly seen
normal blood vessel or a flow test-object. Figure
1.16 shows an unexpected artefact in which the
maximum velocity measured varies with the
c beam position in the field of view. This has
arisen because the transducer aperture used is a
different size in the different positions, resulting
in a different amount of spectral broadening.
Interference from neighbouring

vessels
If part or all of a neighbouring vessel in addition
to the vessel of interest is within the sample
volume of a CW or PW instrument, the Doppler
signal will contain a contribution from the extra
Fig. 1.14 Variation of quality of signal with vessel. Moving the sample volume or redirecting
beam–vessel angle: (a) at 60°; (b) at 70°; (c) at 80°. the ultrasound beam to try to interrogate only
the vessel of interest may reduce this artefact.
Power Doppler images are relatively insensitive Vessel compression

to angle, except near 90°, where the low Doppler It is easy to compress superficial vessels by
frequencies may fall below the clutter filter and transducer pressure. Increased velocity of flow
no power signal is displayed (Fig. 1.3). through the restriction in the compressed vessel
In a vessel in which the direction of flow alters results in a higher-pitched Doppler sound or a
with respect to the ultrasound beam, different colour change in an image.
regions of the vessel will be colour-coded differ-
ently. Note that this may be due to a genuine Factors affecting the patient
change in flow direction as seen in the normal It is necessary to have as complete a knowledge
carotid bulb, or merely due to the changing beam as possible of the patient’s physiological status 21
1
a when undertaking blood flow studies, since
many factors affect the cardiovascular system.
Examples of these factors are exercise, heart
rate, temperature, anxiety, posture, food, smoking
and other drugs.
Patient or vessel movement

If movement causes the sample volume of a CW
or PW beam to interrogate a different region,
the blood flow signal will obviously be altered. It
can be difficult to eliminate this factor, especially
in abdominal examinations, and it is not always
clear whether respiration has actually affected
the flow or just moved the vessel.
b
Flash artefact
Movement of the patient, an organ, or the trans-
ducer during Doppler imaging gives tissues a
velocity relative to the transducer and hence
scattered ultrasound is Doppler shifted; a large
area of the image is therefore colour-coded for
the duration of the movement. This artefact is
more severe with power Doppler imaging due to
its increased sensitivity.
Beam position within the vessel

lumen
c The Doppler signal obtained from a vessel
depends on how the beam insonates the vessel.
The effect is less marked if the beam is wider
than the vessel. A narrow beam through the
centre of a vessel, however, overemphasises the
high velocities while a beam through the side of
the vessel detects lower velocities.
One-dimensional scan
Blood flow often occurs in an unknown direc-
tion in space in relation to the transducer, for
example in the heart or at a vessel bifurcation.
Detecting the flow from one direction only
measures the velocity component along that
Fig. 1.15 Effect of changing velocity scale. beam direction. Measurement of the actual
(a) Maximum velocity 77 cm s-1, (b) maximum velocity in these situations requires components
velocity 32 cm s-1, (c) maximum velocity 8 cm s-1. to be measured in three directions not in the
22
Fig. 1.16 Variation of sonogram

1
Safety and prudent use of Doppler instruments

for different beam positions in the
field of view. A wider aperture is
used near the centre of the array
to increase focusing. This increases
the range of angles of approach of
ultrasound and hence increases
spectral broadening.
a b c
same plane. However, when laminar flow occurs reported but it is not confirmed by further work.
in a vessel lying in an imaging scan plane, There is a need for well-controlled studies but
measurement of one velocity component and these are increasingly difficult to conduct since
the beam–vessel angle permits measurement of unscanned control populations are almost non-
the actual velocity. existent in the developed world. Although no
harmful effects have been confirmed, there is
some concern that the outputs of machines have
SAFETY AND PRUDENT USE
been increasing by factors of as much as three
OF DOPPLER INSTRUMENTS
or five since 1991, as manufacturers seek to
Ultrasound beams transmit energy into tissue so produce better B-mode images and more sensi-
the possibility of hazard has to be considered. tive Doppler units.17 The situation is summed
The most likely mechanisms for harmful effects up in a commentary by ter Haar.18
are thought to be tissue heating as the ultra- Until the early 1990s, attempts were made to
sound energy is absorbed, or cavitation in which specify the maximum intensities permissible for
microbubbles in the tissue react violently under different clinical applications. This proved to be
the influence of the pressure fluctuations of limiting and impractical, so the approach now is
the ultrasound field. The most sensitive struc- to use the ALARA principle (as low as reason-
tures are considered to be the developing fetus, ably achievable) borrowed from the field of
the brain, the eye, the lung and bone–tissue ionising radiations. The user is now informed of
interfaces. the output of the machine and has the respon-
There is a considerable amount of literature sibility to keep the exposure to a low value which
on bioeffects and safety of diagnostic ultra- will still give a diagnosis. Some systems will
sound. The literature is scrutinised by several display the output on the screen in terms of a
national and international bodies who produce thermal index (TI), related to tissue heating and
statements on safety and the prudent use of a mechanical index (MI), related to the possi-
ultrasound. Organisations actively monitoring bility of producing cavitation. These indices are
the safety of ultrasound are the World Federa- defined in the Output Display Standard (ODS)19
tion for Ultrasound in Medicine (WFUM), the developed in the USA.
European Federation of Societies for Ultra- EFSUMB puts out an annual statement on
sound in Medicine and Biology (EFSUMB),15 safety and in its most recent statement it says
the British Medical Ultrasound Society16 and that the use of B-mode is not contraindicated in
the American Institute of Ultrasound in Medicine routine scanning during pregnancy. However, it
(AIUM). It is still true to say that there are no is more cautious with regard to pulsed Doppler,
confirmed harmful effects of diagnostic ultra- saying ‘routine examination of the developing
sound. Often the possibility of an effect is embryo during the particularly sensitive period 23
1
of organogenesis using pulsed Doppler devices a beam through a 3D volume. At present 3D
is considered to be inadvisable at present’. images are often produced by stacking 2D
images next to each other, i.e. a series of parallel
Technical factors affecting prudent use scans. These images are proving useful for
and safety example in the study of flow through a cardiac
1. Use the lowest transmit power which will valve or complex vascular bed (Fig. 1.17).
give a diagnostic result. This involves keeping However true 3D flow imaging would involve
the MI and TI less than one if possible. measurement of the three velocity components
2. Use high receiver gain rather than high of flow at each sample volume, i.e. at each voxel
output power to achieve high sensitivity. in the scanned volume. True 3D flow imaging is
3. Use the minimum scan time possible. still at the laboratory development phase and
4. Check that the transducer ceases to transmit quite far from clinical application.
when the imaging mode is frozen.
5. Take particular care when the fixed beam Tissue Doppler imaging
direction PW Doppler mode is being used All Doppler instruments can be adapted to
near sensitive tissues. study tissue motion rather than blood flow. The
6. Compare the maximum output values (inten- echo signals from tissue are larger than from
sity, power, pressure amplitudes) for your blood and the velocities do not reach the high
machine to those quoted in the published values encountered in blood flow. Nevertheless
data of equipment surveys. the signal processing techniques remain valid
7. Safety considerations related to contrast agents for Doppler tissue motion.20, 21 Most commonly
need to be studied at frequent intervals during it is the myocardium that is studied by both
their developmental stage.15 PW Doppler and colour Doppler imaging (Fig.
1.18). The velocity information in 2D tissue
Doppler images can be further analysed to give
FUTURE INSTRUMENTATION
images of strain and strain rate in the
After the basic principle of a new type of myocardium.22 Doppler tissue techniques have
Doppler instrument has been established there been incorporated in most cardiac imaging
then usually follows several years of techno- instruments.
logical and clinical development. During this
time the technical performance is improved often
by the introduction of new transducers and
computer processing of ultrasonic echo signals.
The clinical performance is also improved by
increased operator experience and the identi-
fication of new applications. The introduction
of completely new technology typically occurs
at the rate of one or two new instruments per
decade. Three recent examples are mentioned
briefly below.
Three-dimensional Doppler flow

imaging
Just as two-dimensional (2D) colour Doppler Fig. 1.17 A 2D projection of a 3D colour Doppler
flow imaging can be performed by scanning an image onto a display screen. The image can be rotated
ultrasound beam through a 2D plane, a 3D to provide a 3D impression of the flow in the 3D
24 colour flow image can be produced by scanning volume scanned. Courtesy of Philips Medical.
1
Appendix A1
the ultrasound beam is interrogating the blood
flow but high quality low noise signals may be
obtained due to the catheter being immersed in
the blood. Experimental systems have been made
which combine Doppler imaging and grey-shade
B-mode imaging.
APPENDIX A1
The velocity of a structure, say in a plane, can
be regarded as the result of combining two other
velocities in two other directions in the plane.
Each of these two velocities is called the velocity
Fig. 1.18 A longitudinal Doppler tissue image of
a section through the left ventricle of the heart.
component along its direction. A Doppler ultra-
sound beam intersecting a moving structure at
an angle to its direction of motion will measure
Catheter Doppler the velocity component of the structure velocity
High frequency transducers can be miniaturised along the beam axis (Fig. 1.19a). Figure 1.19b
to dimensions of less than 1 mm making them illustrates the point that to fully determine the
suitable for insertion into arteries. PW Doppler velocity in 3D space, it is necessary to measure
catheters operating at 20–40 MHz have been three velocity components in three directions in
commercially available for a number of years. space. Figure 1.19c shows that if flow is confined
The small transducer crystal is designed to to motion parallel to the walls of a straight
transmit ultrasound along the artery in the vessel, velocity along the vessel can be deter-
direction of the axis of the catheter wire.23 In mined with one component i.e. with one ultra-
practice it can be difficult to know exactly how sound beam.
1 1 1
Velocity component Velocity components in
2 V1 = Vcosα
in direction 1 3 directions in space
V1 = Vcosα V1 = Vcosα
V1 V2=Vcosβ
V1 V2 V1
V V = Vcosγ V
α β 3
α
α
γ
V3
V (blood velocity) (Flow in vessel in scan plane)
3
(a) (b) (c)
Fig. 1.19 Velocity components along beam directions.
25
1
REFERENCES
1. White DN. Johann Christian Doppler and his effect 13. Burns PN, Powers JE, Fritzsch T. Harmonic imaging;
– a brief history. Ultrasound Med Biol 1982; new imaging and Doppler method for contrast-
8:583–591. enhanced ultrasound. Radiology 1992; 182:142.
2. McDicken WN. Diagnostic ultrasonics: principles 14. Porter TA, Xie F. Transient myocardial contrast after
and use of instruments. London: Churchill initial exposure to diagnostic ultrasound pressures
Livingstone; 1991. with minute doses of intravenously injected
3. Evans DH, McDicken WN. Doppler ultrasound: microbubbles. Circulation 1995; 92:2391–2395.
physics, instrumentation and clinical applications. 15. EFSUMB. Guidelines for the use of contrast agents
Chichester: Wiley; 2000. in ultrasound. Ultraschall in der Medizin 2004;
4. Hoskins PR. Measurement of arterial blood flow by 25:249–256.
Doppler ultrasound. Clin Phys Physiol Meas 1990; 16. http://www.bmus.org
1:1–26. 17. Henderson J, Willson K, Jago JR, et al. A survey
5. Taylor KJW, Burns PN, Wells PNT. Clinical of the acoustic outputs of diagnostic ultrasound
applications of Doppler ultrasound. New York: equipment in current clinical use. Ultrasound Med
Raven Press; 1988. Biol 1995; 21:669–705.
6. Fish P. Physics and instrumentation of diagnostic 18. ter Haar G. Commentary: safety of diagnostic
medical ultrasound. Chichester: Wiley; 1990. ultrasound. Br J Radiol 1996; 69:1083–1085.
7. Wells PNT. A range-gated ultrasonic Doppler 19. American Institute of Ultrasound in Medicine/
system. Med Biol Eng 1969; 7:641–652. National Electrical Manufacturers Association.
8. Kasai C, Namekawa K, Koyano A, et al. Real-time Standard for real-time display of thermal and
two-dimensional blood flow imaging using an mechanical acoustic output indices on diagnostic
autocorrelation technique. Inst Electric Electron Eng ultrasound equipment. Rockville: American Institute
Trans Sonogr Ultrasonogr 1985; 32:458–464. of Ultrasound in Medicine; 1992.
9. Rubin JM, Bude RO, Carson PL, et al. Power 20. Anderson T, McDicken WN. Measurement of tissue
Doppler US: a potentially useful alternative to motion. Proc Inst Mech Eng 1999; 213 Part
mean-frequency based colour Doppler US. H:81–191.
Radiology 1994; 190:853–856. 21. McDicken WN, Sutherland GR, Moran CM, et al.
10. Ophir J, Parker KJ. Contrast agents in diagnostic (1992) Colour Doppler velocity imaging of the
ultrasound. Ultrasound Med Biol 1989; 15:319–333. myocardium. Ultrasound Med Biol 1992;
11. Bommer WJ, Shah P, Allen H, et al. The safety of 18:651–654.
contrast echocardiography – report of the 22. Heimdal A, Stoylen A, Torp H, et al. Real-time strain
Committee on Contrast Echocardiography for the rate imaging of the left ventricle by ultrasound.
American Society of Echocardiography. J Am Coll J Am Soc Echocardiogr 1998; 11:1013–1019.
Cardiol 1984; 3:6–13. 23. Doucette JW, Corl PD, Payne HM, et al. Validation
12. Williams AR, Kubowicz G, Cramer E, et al. of Doppler guide wire for intravascular
The effects of the microbubble suspension SHU measurement of coronary artery flow velocity.
454 (Echovist) on ultrasound-induced cell lysis in a Circulation 1992; 85:1899–1911.
rotating tube exposure system. Echocardiography
1991; 8:423–433.
26
2
Haemodynamics and blood
flow
Peter R. Hoskins, W. Norman McDicken

and Paul L. Allan
Re = rLV
PRINCIPLES OF BLOOD FLOW μ
This section describes the simple principles of where r is the fluid density, L is the vessel
blood flow which are of value in understanding diameter, V is the mean velocity and μ is the
the role of Doppler and for performing vascular fluid viscosity. For a wide variety of fluids the
ultrasound examinations. The underlying prin- transition to turbulence takes place at a value of
ciples of fluid mechanics applied to the flow of Re of about 2000. For flow in which Re is about
blood are complex, and discussed in detail in a 2000 the fluid flow will alternate between
number of texts including those by McDonald,1 turbulent and laminar. When velocity is
Caro et al,2 Strackee & Westerhof,3 and chapters increased so that the Re is above the critical
in the Doppler ultrasound books by Evans et al4 value, turbulence will take a small amount of
and Taylor et al.5 time to develop. During pulsatile flow it is
The blood vessels carry blood from the heart therefore possible for the flow to be laminar at
through the pulmonary and systemic arterial values of Re higher than the critical value,
circulations and back to the heart through the because turbulence does not have time to
venous network. Atheroma develops in the arteries develop before the blood velocity has decreased.
and impedes the flow of blood to a greater, or There is also a third flow state called disturbed
lesser, extent depending on the degree of flow, which refers to variations in velocity magni-
obstruction that results from its presence. tude and direction which occur at low values of
Re. The most important phenomenon is that of
Types of flow vortices, which are regions of circulating flow often
The two essential flow states are laminar and produced when there is some obvious change in
turbulent. At low velocity, fluid flow is laminar vessel geometry such as a stenosis, or the normal
(Fig. 2.1a). This is characterised by the motion carotid bulb. The pattern of vortex production
of fluid along well-defined paths called streamlines. will change as the degree of stenosis and blood
At very high velocities, fluid flow is turbulent velocity increase. During steady flow at low Re
(Fig. 2.1b); particular elements of the fluid no the vortex will be stable and limited to the region
longer travel along well-defined paths, and there is immediately behind the stenosis. At low velocity
a random component to the motion of the fluid. the fluid flow within the vortex is actually laminar.
Concepts of laminar and turbulent flow first At higher Re there will be vortex shedding at
arose from consideration of flow in long straight regular time intervals. Again this is not strictly
tubes. It was found that a dimensionless number turbulence, as the velocity magnitude and direction
called the Reynolds number (Re) was useful in at any location is not random, but follows a regular
characterising the fluid flow.The Reynolds number pattern. At even higher Re the vortex shedding
is defined as: will be combined with the random flow patterns 27
2
Haemodynamics and blood flow
Fig. 2.1 (a) Laminar flow consists
(a) Laminar flow of flow along well-defined
streamlines; the velocity profile
in a long straight tube under
conditions of steady flow is
parabolic. (b) The velocity vector
magnitude and direction in
turbulent flow have random
(b) Turbulent flow
components; the time-averaged
profile is blunt.
of true turbulence.Vortices which are shed travel ultrasound, very little practical distinction is
a few diameters downstream and eventually die made between disturbed and turbulent flow.
out as their energy is absorbed through viscous The presence of spectral broadening is often
losses. During pulsatile flow, vortex shedding may indicative of pathological change in the vessel.
occur for only a portion of the cardiac cycle. A summary of points concerning flow state is
The effect of the flow state on the Doppler given below:
waveform is illustrated in Figure 2.2. Doppler 1. In the normal circulation, flow is mostly
spectra are shown from the normal femoral artery laminar.
in Figure 2.2a; in this case flow is laminar.Within 2. Disturbed flow may occur in particular vessels,
the sample volume the blood velocity magnitude e.g. in the region of the carotid bulb.
and direction is similar for all of the red cells, hence 3. Disturbed and turbulent flow occur in the
the spectral width is low and the waveform poststenotic region.
outline is smooth. In the poststenotic region of a 4. Disturbed and turbulent flow both give rise
diseased artery the Doppler waveform is more to spectral broadening.
complex (Fig. 2.2b). The blood which was at
rest in the poststenotic region during diastole is Pressure and energy
accelerated through the sample volume. For this In the circulation the essential principle is that a
blood, flow is laminar and the initial up-slope of pressure gradient must be created in order for
the waveform has a smooth outline with low blood to flow; this is produced by the contraction
spectral width, whereas blood which was in the of the heart and the resultant ejection of blood
prestenotic region during diastole has to pass into the aorta and systemic vessels.
through the stenosis, producing disturbed and Energy is a useful concept in fluid mechanics.
turbulent flow within the sample volume. The When there is steady flow of an incompressible
variation in velocity magnitude and direction frictionless fluid, the principles of conservation
which this produces results in an increase in the of energy can be used to give Bernoulli’s
spectral width (Fig. 2.2b), and the waveform equation.
outline is no longer smooth.
Energy associated with blood
In the normal circulation, flow is mostly laminar;
pressure +
however, disturbed flow may occur in particular
vessels such as the carotid arteries. Flow recircu- Kinetic energy of moving blood +
lation is commonly seen in the region of the bulb Potential energy associated
and there may be disturbed flow in the distal with the height of
28 region. For the purposes of clinical Doppler the fluid above ground = Constant
a
2
Principles of blood flow

accelerate the fluid.4 Energy is conserved in this
simple ideal lossless system. In the circulation,
energy is lost in the form of heat through viscous
effects, manifested through friction of the blood
at the vessel wall and between adjacent layers of
blood. Energy losses are highest in the region of
a stenosis, as there is considerable friction
during turbulent flow and vortex motion.
Within a stenosis there will be an increase in
blood kinetic energy associated with increase in
blood velocity, and according to Bernoulli’s
equation there is a corresponding fall in blood
pressure. If there was no energy loss within the
system then the decrease in velocity (and hence
kinetic energy) in the poststenotic region would
b be compensated by a return of the pressure to the
prestenotic level. In practice the loss of energy
through turbulence and vortex shedding gives
rise to a pressure drop across the stenosis whose
magnitude is dependent on the degree of
stenosis.
The most common application of Bernoulli’s
equation is in the prediction of pressure drop
across a stenosed cardiac valve.6 The equation
may be simplified to:
P = 4V 2
where V is the measured velocity in m s-1, and P

is the pressure drop in mmHg. Points concerning
Fig. 2.2 Femoral artery Doppler waveforms.
pressure and energy are summarised below:
(a) From a normal segment; the waveform has a
smooth outline and the spectral width is low. 1. The pressure drop across a stenosis is high as
(b) From the poststenotic region; in the early systolic a result of energy loss in the poststenotic
phase the waveform has a clearly defined outline region.
associated with the passage of blood which was at 2. For the estimation of the pressure drop across
rest in the poststenotic region during diastole through a cardiac valve stenosis, Bernoulli’s equation
the insonation site. In the later part of the waveform, may be simplified to P = 4V 2.
blood which has passed through the stenosis has
developed turbulent, disturbed flow with increased Velocity profiles
velocity. This appears as a region in which there is The Doppler ultrasound spectrum is critically
spectral broadening and high-frequency spikes.
related to the detailed variation of velocity within
the vessel of interest. The velocity of the blood
This is a simple expression which demon- will vary as a function of its position within the
strates that there will be an interchange between vessel; this is called the velocity profile.The most
the different types of energy within the circulation. commonly known velocity profile is called the
In the human body, however, the flow is non- parabolic velocity profile.
steady and the above equation must be modified Strictly speaking, a parabolic velocity profile
slightly to account for the energy required to only applies to steady laminar flow in a long 29
2
straight tube, when there is maximum velocity in and the direction of motion near the wall will
the centre of the vessel and zero velocity at the reverse if there is a sufficiently rapid increase in
edge of the vessel (Fig. 2.1a).The profile is radially vessel cross-sectional area with distance.The phrase
symmetric, which means that it is the same ‘flow separation’ is often used to describe this
regardless of which diameter is considered. The phenomenon; that is, the high-velocity central
shape of the profile is an exact mathematical jet is located next to a region in which the flow
equation, that of a parabola. is of low velocity and recirculating.The production
Velocity profiles in vessels in the body are of vortices was noted above; both the central jet
generally more complex; they may not be even and vortices die out after a length equivalent to
approximately radially symmetric and they vary a few diameters and laminar flow is re-estab-
with time during the cardiac cycle. It is worth lished. Figure 2.4 shows the velocity profiles in
exploring the various effects that will influence the region of a small stenosis. When the expan-
true velocity profiles in the circulation. sion is less severe, such as a gradually widening
taper, the velocity profile simply becomes more
Entrance effect blunted.
The velocity profile in a vessel is strongly influ-
enced by the distance of the region of interest Curved vessels
from the entrance to the vessel. For a long straight Figure 2.5 shows that the velocity profile for
vessel, when there is steady flow, the profile is steady flow in a curved vessel is skewed towards
initially flat at the entrance to the vessel. With the outer wall when the entrance profile is para-
increasing distance from the entrance the profile bolic, and skewed towards the inner wall when
will change, becoming parabolic at a distance the entrance profile is flat.
called the inlet length (Fig. 2.3).
Y-shaped junction
Vessel narrowing Figure 2.6 shows the velocity profiles from a Y-
For steady flow, a gradual narrowing taper will shaped junction and it can be seen that the profiles
tend to sharpen the velocity profile. are skewed within the two branches, so that the
higher velocities occur on the inner aspects of
Vessel expansion the two branches.
At regions where the cross-sectional area of the
vessel increases, an adverse pressure gradient in Pulsatile flow
the direction of flow is created; that is, there is a During pulsatile flow the velocity profile will vary
pressure decrease in the direction of flow, which throughout the cardiac cycle. Figure 2.7 shows
tends to retard the flow. For the central high- the profiles from a long straight tube with a
velocity region, the high momentum opposes this, velocity waveform similar to that found in the
but at the edge of the vessel the velocities are low femoral artery.
Fig. 2.3 Velocity profiles during

steady flow at different distances
from the entrance to a long straight
tube from a reservoir. The parabolic
velocity profile is restored at a
distance from the entrance,
the ‘inlet length’. After Caro et al,2
Velocity with permission.
30
Fig. 2.4 Velocity profiles from a

2

stenosis model; the region of
recirculation in the poststenotic
region can be seen on the lower
aspect.
(a) (b)
Entrance Outside wall

to pipe
Inside wall
Axial velocity profile
Fig. 2.5 Velocity profiles within a curved vessel during steady flow. (a) A parabolic velocity profile at the
entrance results in higher velocities on the outer aspect of the curve. (b) A blunt velocity profile at the entrance
results in the higher velocities occurring on the inner aspect. From Caro et al,2 with permission.
Turbulence
As discussed above, the velocities during turbu-
lence have a random component, so that it is
necessary to take an average value over time. If
this is done, then the averaged velocity profile
during steady turbulent flow is found to be blunted,
with high-velocity gradients near to the vessel
wall (Fig. 2.1b).
Secondary flow motions

In many of the geometrical situations described
above the components of flow are three-
Fig. 2.6 Velocity profiles in a Y-shaped junction with dimensional, which means that there will be
skewing of the velocity profile and higher velocities some secondary flow motion in the plane perpen-
on the inner aspects of the two branches. dicular to the vessel axis. These motions may 31
2
Fig. 2.7 Changes in the velocity
profiles at various moments during
pulsatile flow for a velocity
waveform similar to that found in
the femoral artery. From Evans
et al,4 with permission.
easily be demonstrated using flow models and steady flow conditions; it is therefore useful mainly
dye-injection techniques, and there are a few in as an aid in understanding general concepts of
vivo studies which claim to have demonstrated flow in arteries, and a more complex version of
this.7, 8 A summary of points concerning velocity this equation must be used for pulsatile flow. For
profiles is given below. a long straight vessel the resistance to flow depends
1. Velocity profiles are influenced by a large on the fourth power of the diameter. A segment
number of factors, and it generally cannot be of vessel 2 mm in diameter will therefore have a
assumed that the profile is parabolic. resistance 16 times that of a similar segment of
2. The displayed Doppler spectrum will be 4 mm diameter.
critically related to the velocity profile present A simple model of the flow to an organ is
within the sample volume of the Doppler shown in Figure 2.8. The net flow is controlled
system. by a combination of the small vessel (arteriolar)
resistance and the large vessel (arterial) resistance.
A simple flow model In the non-diseased circulation, the main arteries
The creation of a pressure gradient within the have relatively large diameters and their resistance
arterial system is performed by ejection of blood to flow is small; the main resistance vessels are
into the arterial tree by the heart. The resistance the arterioles.The essential clinical manifestations
R to flow of a vessel segment may be defined as: of atherosclerosis may be understood with the
aid of this model; an increase in resistance in a
R = (P1 - P2)
large distributing artery because of atheroma must
Q
be compensated by a decrease in the resistance
where Q is the flow through the vessel, and P1 of the small arteries and arterioles in order to
and P2 are the pressures at the entrance and exit preserve flow to the capillary bed. As the disease
points of the vessel. One way of expressing this progresses, flow is maintained by arteriolar dilata-
equation is to say that in order to maintain flow tion until the point is reached where the arteriolar
at a constant level, the pressure difference must network is fully dilated. Further progression of
be greater when the resistance to flow increases. the proximal disease results in a reduction in flow
32 Strictly speaking this formula only applies for to the organ and the development of ischaemia
Fig. 2.8 A simple model of flow

2

from the heart to an organ through
Large Small a large vessel (arterial) and small
Heart vessels vessels Capillary bed vessels (arterioles) to the capillary
(arteries) (arterioles)
bed.
because no further compensatory dilatation is The concept of a critical stenosis is useful

achievable. In patients with lower limb claudi- but its application to atherosclerosis should not
cation, the presence of severe proximal disease be taken too far. As atherosclerosis develops,
results in the distal arteriolar network being fully various other compensatory mechanisms come
dilated at rest in order to maintain flow to the into play in an attempt to preserve perfusion.
lower limb muscles. Whilst this is sufficient in These include the development of a collateral
the resting limb, the combination of the proximal circulation and a degree of local dilatation of the
stenosis and the maximum arteriolar dilatation affected arterial segment. In addition, there is an
means that no further increase in flow can be increase in the extraction efficiency of oxygen
obtained to cope with the increased metabolic from blood. A summary for this section is given
demands of limb exercise. below.
The concept of a critical stenosis follows from 1. The degree of constriction of the distal arte-
the above model. As the degree of narrowing at riolar bed is one factor used to control the
a single, isolated stenosis increases, a point is flow rate to the organ.
reached at which the distal arteriolar dilatation 2. As the resistance to flow of diseased arteries
is at maximum. Consequently, any increase in increases, flow rate is maintained within
the degree of stenosis beyond this point leads to normal levels as a result of distal arteriolar
a reduction in flow. Experiments performed on dilatation.
animals suggest that this point of critical stenosis 3. Very high degrees of stenosis are accompanied
is reached with an area reduction of about 75%, by a low flow rate and low velocities.
which corresponds to a diameter reduction of
approximately 50%.Two quantities of interest in
Doppler ultrasound are the volume flow rate
and the velocity of the blood; the relationship
Flow (mL/min) or velocity (cm/s-1)
600
between these two parameters, according to the Velocity
500
model developed above, is shown in Figure 2.9.9
As the calibre of the vessel is reduced, the volume 400
of blood flowing along the vessel is maintained Blood flow
300
by increasing the velocity. However, above the
point of critical stenosis (75% area stenosis), the 200
volume of blood starts to reduce. It should also
100
be noted that the velocity peaks at about 85%
diameter stenosis, subsequently tailing off, so that 0
in very tight stenoses the velocity is relatively low. 100 80 60 40 20 0
Two stenoses in series have a larger overall % Stenosis by diameter
resistance compared with either stenosis considered
individually. In practice the combined resistance Fig. 2.9 Flow rate and velocity based upon a single
of stenoses in series is dominated by the one arterial stenosis inserted into an otherwise normal
with the smallest luminal diameter. artery. After Spencer & Reid,9 with permission. 33
2
Pulsatile flow and distal resistance
and flow waves. As noted above in a long straight
Doppler ultrasound is commonly used to assess pipe there are no reflected waves. However the
distal resistance to flow. The origin of pulsatile arterial system is a branching network and the
waveforms and their relation to distal resistance site of the branches will cause a portion of
are considered here. the pressure wave to be reflected and travel back
For a particular element of blood it is the upstream. The major source of reflected waves
pressure gradient, not the actual pressure, which occurs at the arteriolar junctions, which are the
accelerates the blood. The pressure gradient is major resistance vessels in the body (Fig. 2.13).
related to the difference between the pressures on When the arterioles are tightly constricted as in
either side of the element of blood (Fig. 2.10). the case in resting muscle the amplitude of reflected
When the pressure gradient is positive the blood waves is high leading to reverse flow. On the
will be accelerated along the vessel; when the other hand during exercise or reactive hyperaemia
gradient is negative the blood will be decelerated. the arterioles are dilated, the amplitude of reflected
The corresponding flow waveform is found by waves is low and the reverse flow component is
detailed calculation of the pressure gradient at lost. This relationship between the degree of
the site of interest. diastolic flow and the downstream resistance is
The blood ejected by the heart passes into the used as a diagnostic tool, for example in obstetrics
aorta. This causes local expansion of the aorta where umbilical artery Doppler waveforms are
and distal arteries due to the local high pressure. used to provide an indicator of placental resist-
The expanded region passes down the arterial ance to flow. The normal placenta has a low
tree in the form of a pressure wave (Fig. 2.11). resistance to flow and umbilical waveforms show
If the artery were long and straight then, at a flow throughout the cardiac cycle. Absent end-
particular location along the vessel, the pressure diastolic flow is associated with increased resist-
would reach a maximum and then decline to the ance to flow and abnormal placental development;
baseline value (Fig. 2.12a), resulting in a flow and there is an increased incidence of fetal
waveform with forward flow only (Fig. 2.12b).10 compromise. Studies in sheep have provided
In practice it is known that flow waveforms in good evidence for the basis for this work.11
arteries can exhibit periods of reverse flow, and Before considering the detail of how reflected
this section is concerned with understanding the waves give rise to reverse flow, it is worth
origin of this reverse flow. considering the simple phenomenon of water
Periods of reverse flow typically occur in waves to illustrate some of the concepts. If a
arteries which supply muscle at rest, for example buoy is placed in the ocean it will move up and
the brachial artery or the femoral arteries. However down with the waves, but the height of the buoy
on exercise, or during periods of reactive does not provide any information on the direction
hyperaemia, the reverse flow disappears, and of travel of the wave. If two waves are travelling
there is forward flow throughout the cardiac towards each other from opposite directions and
cycle. This difference in flow waveform is due to cross each other at the location of buoy, then the
differences in the amplitude of reflected pressure height of the buoy will double as the waves cross
Fig. 2.10 The force acting on an

element of blood is related to the
difference in pressures on either
P1 P2
side of that element.
Pressure difference = P1 – P2
34
2
Quantitative flow measurement

Forward-going
pressure wave
Backward-going
pressure wave
Large Arteriolar
arteries bed
Fig. 2.13 Reflected pressure and flow waves are

Fig. 2.11 The blood ejected by the heart causes produced from the distal arteriolar bed.
expansion of the elastic arteries. This expansion
passes down the arteries in the form of pressure and with the forward-going pressure wave in an additive
flow waves. manner, hence the combined pressure will be
greater (Fig. 2.14). There is also a reflected
(reverse-going) flow wave which will also travel
(a) Pressure back up the arterial tree. However in the case of
the flow wave the direction of motion is important.
Because the reverse-going flow wave is travelling
in the opposite direction to the forward-going
flow wave, the result will be a subtraction of the
reverse-going from the forward-going wave, which
then results in reverse flow (Fig. 2.15).
A summary of this section is given below:
1. Reflected pressure waves from the distal arte-
riolar bed interact with the forward-going waves
(b) Flow
to produce the waveform causing increases in
pulsatility of the flow waveform.
2. For some arteries the waveform pulsatility as
measured using Doppler ultrasound may be
a clinically useful indicator of disease.
QUANTITATIVE FLOW
MEASUREMENT
Whilst it is possible to measure blood flow
Fig. 2.12 (a) and (b) Pressure and flow waves in the quantitatively, the error is usually rather large,
absence of distal reflection. From Murgo et al,10 with probably between 20 and 100%.12,13 In some
permission. applications such errors may be tolerable, for
instance when a large change in flow in the order
due to the additive effect arising from two waves. of 300% from normal to abnormal flow exists.
Returning to the arterial system, there will be a However, one factor which mitigates against
reflected (reverse going) pressure wave which blood flow being a sensitive indicator of disease
will travel back up the arterial tree and combine affecting an organ or limb is that circulatory 35
2
Pressure
Time
Forward-going Reverse-going Resultant

+ =
pressure wave pressure wave pressure wave
Fig. 2.14 The resultant pressure wave is a combination of the forward-going wave and the reverse-going wave.
Flow
Time
Forward-going Reverse-going Resultant

+ =
flow wave flow wave flow wave
Fig. 2.15 The resultant flow wave can be considered to be a combination of a forward-going flow wave and a
reverse-going flow wave.
regulatory mechanisms can often maintain the 2. In practice it is difficult to ensure uniform
level of blood supply until the disease is quite insonation of a blood vessel, and the resultant
advanced. error in instantaneous average velocity can be
A variety of factors must be borne in mind very large, perhaps greater than 50%.
when considering volume flow calculations in Maximum velocity, for use in the calculation
relation to the circulation of blood: of instantaneous average velocity (see below),
1. Flow is pulsatile, therefore the velocity varies can be measured to around 5% by ensuring
over the cardiac cycle. The velocity also varies that the ultrasound beam passes through the
across the vessel lumen. In addition, turbu- centre of the vessel. It should always be borne
lence and disturbed flow will result in vectors in mind that spectral broadening errors can
of flow off the central axis of the vessel, be large, up to 50%, when maximum velocity
leading to inaccurate angle estimation and is measured by wide-aperture arrays.14
correction. Complex flow patterns can occur 3. The calibre of compliant arteries varies with
in curved vessels and near bifurcations, so the cardiac cycle. Pulsation of an artery can
that the assumption of laminar or plug flow change its cross-section by up to 20%, hence
should be carefully examined for each appli- an instantaneous diameter measurement (see
cation. Turbulence observed in a sonogram below) should be used if possible.
or colour Doppler image makes accurate 4. The accuracy of the measurement of the vessel
36 measurement of volume flow impossible. diameter, or cross-sectional area, is inversely
2
Quantitative flow measurement

related to the size of the vessel and measurement bolic flow the instantaneous average velocity equals
errors are usually quite large, up to 20% for half of the maximum velocity. The maximum
a 10 mm diameter vessel, rendering the tech- velocity is readily measured from a spectral display,
nique of doubtful value for small vessels. The provided that part of the beam passes through
cross-sectional area and the velocity should the centre of the vessel.
be measured at the same site; in addition, The instantaneous cross-section is most easily
errors will occur if the scan plane is not exactly calculated from the instantaneous diameter
at right angles to the axis of the vessel. Special measured from an M-scan generated simulta-
efforts should be made to reduce the error in neously with the sonogram; the assumption of a
diameter measurement, since diameter is circular cross-section for a given diameter appears
squared in the formula for cross-section (area to be reasonable for arteries and introduces little
= πr2) and this increases the error. error. The diameter can also be obtained from a
5. For laminar flow in a straight vessel the beam- B-scan provided that it is recorded at the same
vessel angle can be determined to within 2° time and the same location as the spectrum
or 3°. However, even with this accuracy, if the being used from the sonogram.
beam–vessel angle is made greater than 60°, Averaging the instantaneous flow rate over
the error in the calculated velocity rapidly rises the cardiac cycle gives the average flow rate.
above 10%, so that the calculated velocities In practice, these measurements may not be
will have significant errors associated with easily achieved on some systems and the following
them in this situation. is a more practical formula to use for volume
flow rate, in mL min-1.
Knowing the sources of error, steps may be
Flow rate (mL min-1) = TAVmean (cm s-1) ×
taken to minimise them. A simple approach to
area (cm2) × 60
measurement is based on the formula:
In these cases the time-averaged maximum
Instantaneous flow rate =
velocity (TAVmax) can be calculated over several
instantaneous cross-section × instantaneous
cardiac cycles and, assuming parabolic flow, divided
average velocity
by two to give a calculated time-averaged mean
‘Instantaneous’ means the value measured at the velocity (TAVmean). Some systems will calculate
point in the cardiac cycle being considered. the time-averaged mean velocity directly from
The instantaneous average velocity may be the spectral display. Similarly, only a few systems
calculated by taking the average of the velocities will provide a time-averaged diameter measure-
in the spectral display. Both the power level of ment and therefore an instantaneous diameter
each velocity in the spectrum and the velocity measurement must be used to calculate area,
itself are used in this calculation.The power level usually from an image frozen in systole.
for each velocity is a measure of the number of Alternatively, the area of the vessel can be
cells moving with that velocity. For this calcu- measured directly from a transverse image of the
lation to be accurate, the vessel must be uniformly vessel, providing care is taken to ensure that the
insonated, which is not easy to achieve due to section is a true cross-section of the vessel made
beam misalignment and distortion. at systole.
Another way to obtain the instantaneous average Measurement of volumetric flow is difficult and
velocity is to measure the maximum velocity and not generally attempted, as it is easy to produce
assume a profile for the velocities across the errors of 100%,12 but there is little in the literature
vessel, for example a plug or a parabolic profile. that gives accurate data on errors in practice. In
With plug flow all of the blood cells are travelling general terms it should be considered as a research
with the same velocity, so the instantaneous tool and the results treated with suitable caution
average velocity equals the maximum. For para- with regard to the potential errors. 37
2
on the thickness of the arterial wall. It would be very
ARTERIAL MOTION
desirable to measure the bulk modulus of the
In the section on pressure and energy above, it artery, also called the Young’s modulus; however
was described how the blood ejected from the this requires knowledge of the wall thickness, which
heart travels down the elastic arteries causing is difficult to measure using ultrasound imaging.
them to distend. Measurement of wall motion on commercially
It is possible to examine arterial distension available systems is slowly being introduced.
using ultrasound systems. Typically the artery is These are aimed mainly at an assessment of flow
imaged along the longitudinal plane and the mediated dilatation.This is the phenomena whereby
distension at a selected location examined. The the mean arterial diameter will increase slightly
artery increases in diameter during systole, and following a brief period of absent blood flow.
reduces in diameter during diastole. Typically Usually the brachial artery is considered and flow
the overall distension is 10% of the diameter, or is stopped by an inflated arm cuff.The degree of
about 0.5 mm for a 5 mm diameter vessel. diameter increase following cuff release is used
The degree of arterial distension is related to as a measure of endothelial function. Studies over
the elasticity of the artery; stiff arteries will not many years have shown that the overall diameter
stretch much, whereas elastic arteries will stretch increase may be measured using ultrasound.17, 18
more. It is possible to estimate an index of
elasticity, called pressure strain elastic modulus,
CONCLUSIONS
from distension and blood pressure.15
The study of haemodynamics involves the appli-
Pressure strain elastic modulus
cation of the principles of fluid mechanics to blood
change in blood pressure
= flow in the circulation and provides an insight
fractional change in diameter
into the events which occur in both normal and
This provides an easy estimate of arterial diseased vessels. It is essential to have an under-
stiffness, allowing applications in patients with standing of these haemodynamic principles in
for example abdominal aortic aneurysm.16 The order to be able to carry out Doppler exami-
pressure strain elastic modulus is a measure of nations and to understand the findings which
the structural stiffness of the artery and is dependent are obtained during the procedures.
REFERENCES
1. McDonald DA. Blood flow in arteries. London: 7. Hoskins PR, Fleming A, Stonebridge P, et al.
Edward Arnold; 1974. Scan-plane vector maps and secondary flow
2. Caro CG, Pedley TJ, Schroter RC, et al. The mechanics motions. Eur J Ultrasound 1994; 1:159–169.
of the circulation. Oxford: Oxford University Press; 8. Stonebridge PA, Hoskins PR, Allan PL, et al. Spiral
1978. laminar flow in vivo. Clin Sci 1996; 91:17–21.
3. Strackee J, Westerhof N. The physics of heart and 9. Spencer MP, Reid JM. Quantitation of carotid
circulation. Bristol: Institute of Physics; 1993. stenosis with continuous wave (CW) Doppler
4. Evans DH, McDicken WN, Skidmore R, et al. ultrasound. Stroke 1979; 10:326–330.
Doppler ultrasound: physics, instrumentation and 10. Murgo JP, Col MC, Westerhof N, et al. Manipulation
clinical applications. Chichester: Wiley; 1989. of ascending aortic pressure and flow waveform
5. Taylor KJW, Burns PN, Wells PNT. Clinical applications reflections with the Valsalva manoeuvre:
of Doppler ultrasound. New York: Raven Press; 1995. relationship to input impedance. Circulation 1981;
6. Holen J, Aaslid R, Landmark K, et al. Determination 63:122–132.
of pressure gradient in mitral stenoses with a non- 11. Adamson SL, Morrow RJ, Langille BL, et al.
invasive ultrasound Doppler technique. Acta Med Site-dependent effects of increases in placental
38 Scand 1976; 199:455–460. vascular resistance on the umbilical arterial velocity
waveform in fetal sheep. Ultrasound Med Biol
16. Wilson KA, Lee AJ, Lee AJ, et al. The relationship
2
References
1990; 16:19–27. between aortic wall distensibility and rupture of
12. Evans DH. Can ultrasonic duplex scanners really infrarenal abdominal aortic aneurysms. J Vasc Surg
measure volumetric flow? In: Evans JA, ed. 2003; 37:112–117.
Physics in medical ultrasound. York: Institute of 17. Celermajer DS, Sorensen KE, Gooch VM, et al.
Physical Sciences in Medicine; 1986. Noninvasive detection of endothelial dysfunction
13. Gill RW. Measurement of blood flow by ultrasound: in children and adults at risk of atherosclerosis.
accuracy and sources of error. Ultrasound Med Biol Lancet 1992; 340:1111–1115.
1985; 11:625–641. 18. Sidhu JS, Newey VR, Nassiri DK, et al. A rapid and
14. Hoskins PR. Accuracy of maximum velocity reproducible on line automated technique to
estimates made using Doppler ultrasound systems. determine endothelial function. Heart 2002;
Br J Radiol 1996; 69:172–177. 88:289–292.
15. Peterson LH, Jensen RE, Parnell J. Mechanical
properties of aneurysms in vivo. Circ Res 1960;
8:622–639.
39
3
The carotid and vertebral
arteries; Transcranial colour
Doppler
Paul L. Allan and Karen Gallagher
Table 3.1 Indications for carotid ultrasound

THE CAROTID AND VERTEBRAL
ARTERIES • Transient ischaemic attacks
• Reversible ischaemic neurological deficits
Indications • Mild resolving strokes in younger patients
Ultrasound of the extracranial cerebral circu- • Atypical, non-focal symptoms which may have a
lation is used predominantly in the assessment vascular aetiology
of patients with symptoms which might arise • Arteriopaths/high-risk patients prior to surgery
from disease in the carotid arteries, such as • Postendarterectomy
amaurosis fugax and transient ischaemic attacks • Pulsatile neck masses
(TIA), in order to identify those patients with • Trauma, or dissection
significant changes who will benefit from • Screening for disease
surgery. Two major trials have shown that

endarterectomy for symptomatic patients with
significant stenoses [>80% diameter reduction Cerebral ischaemic symptoms
in the European Carotid Surgery Trial (ECST), There are many causes of cerebral ischaemic
>70% in the North American Symptomatic symptoms apart from disease at the carotid
Carotid Endarterectomy Trial (NASCET)] bifurcation. These include cardiac arrhythmias,
confers a significant advantage over medical hypotensive episodes, emboli and atheromatous
management in terms of reducing morbidity disease elsewhere in the circulation between the
and mortality.1,2 The ECST data showed that, heart and the intracerebral arterioles. Many of
during the follow-up period, the percentage of these can be treated with medical therapy but it
subjects having ischaemic symptoms lasting is only the extracranial section of the carotid
more than 7 days was 16.8% for those on artery which is amenable to surgery, and it is for
medical management, compared with 2.8% if this reason that so much effort is devoted to the
surgery was performed. There is no value in assessment of this area. The main aim is to
having surgery with symptomatic stenoses less classify patients into five main groups.
than 50% and it is only of marginal benefit for 1. Those without significant disease.
stenoses of 50–70%.3 A Cochrane Review of 2. Those with mild disease (<50% diameter
the topic4 concluded that surgery was only reduction), who will benefit from medical
beneficial for symptomatic patients with a therapy if they are symptomatic.
stenosis of more than 70% on ECST criteria, 3. Those with more severe disease (50–70%
or 50% on NASCET criteria. The main indica- diameter stenosis), who will be treated medically
tions for ultrasound of the carotids are shown and may be followed to assess progression of
in Table 3.1. disease, particularly if they are symptomatic.
41
3
The carotid and vertebral arteries; Transcranial colour Doppler
4. Those patients with severe disease (>70% for endarterectomy, in those centres which offer
diameter reduction) who will benefit from surgery. However, if there is a policy not to offer
surgery if they are symptomatic. surgery to asymptomatic patients, then it might
5. Those patients with a complete occlusion, be argued that an ultrasound examination is
who are therefore not candidates for surgery. unnecessary.
The policy for ultrasound in patients with
The relationship between the presence of carotid strokes needs some consideration. Patients with
artery disease and the development of cerebral a significant, persistent, established neurological
ischaemic symptoms is not straightforward and deficit will not benefit from carotid surgery
detailed discussion of this subject is beyond the whereas individuals with a mild, resolving
scope of this book. However, patients who have deficit, who are otherwise candidates for carotid
suffered from temporary ischaemic symptoms, endarterectomy, may be considered for surgery
such as TIA, reversible ischaemic neurological in order to reduce the risk of a subsequent event.
deficits, or amaurosis fugax, are significantly It is therefore reasonable to suggest that an
more likely to suffer a stroke than asymptomatic ultrasound examination is of benefit in younger
subjects: 36% of patients who have a TIA will patients with milder, resolving neurological
have an infarct within 5 years of the TIA, com- deficits but it is of little value in older patients
pared with an annual stroke rate of 1% for with more permanent neurological signs.
asymptomatic, elderly individuals.5 Therefore it
is reasonable to investigate patients with reversible Atypical symptoms
ischaemic cerebral symptoms in order to iden- Some patients have unusual symptoms which
tify those with a 70% or greater stenosis who will may or may not be related to carotid disease.
benefit from endarterectomy. Those with lesser Atypical migraine, hyperventilation attacks and
degrees of stenosis can be treated medically and temporal lobe epilepsy may sometimes be diffi-
followed up; those who progress to more than cult to diagnose and, in some patients, the possi-
70% diameter stenosis can then be considered bility of carotid disease might be considered.
for surgery. Ultrasound is of value in excluding carotid disease
The situation regarding the examination of as a cause of the symptoms in this group of
patients with asymptomatic carotid bruits is also patients, although some care must be given to
complex. The authors, along with many people, patient selection to prevent large numbers of
would wish to know the status of their arteries unnecessary examinations.
if they were found to have an asymptomatic
carotid bruit. Although the risk of ipsilateral Patients at risk of perioperative stroke
stroke from an asymptomatic carotid stenosis is Arterial disease is usually a generalised process,
low, the Asymptomatic Carotid Atherosclerosis although it affects different arterial territories to
Study (ACAS) reported that surgery is bene- varying degrees. Therefore, patients undergoing
ficial in reducing the risk of subsequent ipsi- surgery for conditions such as coronary artery
lateral stroke by some 5% in asymptomatic disease, peripheral arterial disease and aortic
patients with a stenosis of more than 60% aneurysms may also have significant carotid
diameter reduction, providing that the centre disease; there is concern that perioperative
has a perioperative morbidity/mortality rate of morbidity from strokes can be increased in these
less than 3%.6 A Cochrane Review of surgery for patients as a result of emboli or inadequate
asymptomatic stenosis concluded that the value perfusion. Diabetics can also have severe arterial
of surgery was barely significant and extremely disease and are at risk from perioperative strokes
small in terms of absolute risk reduction.7 Ultra- when undergoing major surgery. A review of
sound will therefore have a role in the identifi- carotid artery disease and stroke during coronary
42 cation of those patients who might be considered artery bypass surgery8 showed that patients with-
3
The carotid and vertebral arteries

out carotid disease had a <2% stroke rate and
Table 3.2 Causes of pulsatile neck masses
this only rose to 3% in patients with asymp-
tomatic disease; it also drew attention to the role • Normal but prominent carotid artery and bulb
of aortic arch disease in the aetiology of strokes • Ectatic carotid, brachiocephalic or subclavian
artery
in coronary artery bypass graft patients. Ivey et
• Aneurysm of the carotid artery
al9 showed that there was no increase in risk
• Carotid body tumour
associated with asymptomatic bruits in patients
• Enlarged lymph node adjacent to carotid sheath
undergoing cardiopulmonary bypass procedures,
even if there was a haemodynamically significant
stenosis; although they felt that patients with a
bruit and a history of ischaemic symptoms Carotid dissection
should be considered for a staged, or simulta- Dissection of the carotid artery may develop
neous, endarterectomy if they were shown to from a variety of causes.11
have a stenosis of >70% diameter reduction. 1. It may occur spontaneously, usually conse-
However, this decision would depend on the quent upon atheromatous change.
relative urgency of the primary condition and 2. It may result from the extension of an aortic
many centres, whilst taking note of the carotid arch dissection.
disease, will proceed with the main operation 3. It may develop following trauma to the neck,
and consider subsequent endarterectomy in such as occurs in whiplash injuries.
symptomatic patients. 4. As a result of iatrogenic causes, such as carotid
catheterisation.
Postendarterectomy patients
Complications following endarterectomy can be Colour Doppler can be used to identify
divided into three groups based on the timing of different flow patterns on either side of the flap,
the events. or the presence of a thrombosed channel, and
1. Early occlusion due to thrombosis, occurring monitor subsequent progress.
within the first 24–48 h after the operation.
2. Stenosis developing over 12–18 months due Epidemiological studies and monitoring
to neointimal hyperplasia. of therapy
3. Recurrence of atheroma over a period of The carotids provide a convenient window for
several years resulting in restenosis. the assessment of the whole arterial system.
Patterns of development of atheroma vary for
Colour Doppler ultrasound provides a rapid different arterial areas. Nevertheless it could be
and straightforward method for diagnosis of these expected that changes in the carotids might be
complications. related to disease in other vessels, such as the
Routine follow-up of asymptomatic patients coronary arteries, and that the changes in the
is not justified by the pick-up rate for developing carotids might allow some prediction of severity
significant recurrent stenoses,10 but any patient of this disease. In addition their examination
suffering symptoms related to the operated side could also provide a method for assessing the
should be examined by colour Doppler in the rate of progression, or regression of disease, if
first instance. treatment regimens are being investigated, or
epidemiological studies are being performed.12
Pulsatile masses By far the largest group of patients will fall
Colour Doppler ultrasound provides a rapid into the first group of indications relating to the
technique for the assessment of pulsatile neck diagnosis of carotid atheroma and stenosis as a
lumps. There are a variety of causes for these; cause of cerebral ischaemic symptoms. In the
the main ones are listed in Table 3.2. end the aim of the sonographer is to identify 43
3
patients with carotid disease which may be the means that the couch and machine have to be
cause of their symptoms, and to assess the severity moved around in the middle of the list, which is
of the disease so that appropriate management disruptive and time consuming.
decisions in relation to surgery or medical A high-frequency transducer (7–12 MHz) is
management can be taken. used and the examination starts with a trans-
verse scan of the carotid artery from as low in
Anatomy and scanning technique the neck as possible, to as high in the neck as
The main steps in the examination are given in possible behind the angle of the mandible. This
Table 3.3. The patient lies supine, with their approach will allow the depth and course of the
neck a little extended by placing a pillow under vessels to be ascertained, together with the level
their shoulders. The patient should be comfort- of the bifurcation and the orientation of its
able and excessive extension of the neck should branches (Fig. 3.1). In addition, areas of major
be avoided. In addition, some patients with carotid disease will be identified and can be noted for
or veretebral disease may find that neck exten- further assessment.
sion compromises the flow of blood to the cerebral Colour Doppler is then activated and the
circulation, so if the patient appears to be asleep vessels are examined in the longitudinal plane,
it is worth checking that they have not lost again from the lower neck upwards. Areas of
consciousness. Some patients may not be able to abnormal flow are identified with colour Doppler,
lie supine; if this is the case they can usually be an initial assessment of their significance is made
examined adequately in a sitting position. and the need to undertake a spectral examina-
The examiner can sit beside the patient’s tion can be considered. Just as importantly, areas
thorax and scan the neck from this position, or of normal flow are seen so that the normal
sit at the patient’s head and scan the neck from segments of the vessel can be identified rapidly
this location; this latter arrangement was and excluded from further investigation. It may
favoured in the early days of carotid ultrasound be necessary to try a variety of scan planes in
as it was easier to obtain a standardised probe order to see the bifurcation in some subjects; the
position, but it is no longer necessary with modern normal approach is from an anterolateral or
colour Doppler equipment. Furthermore, using lateral direction but more posterior planes may
this position at the head of the patient for
carotid scanning during a general ultrasound list
Table 3.3 Basic steps in the examination
1. Transverse scan from low in the neck up to

behind the angle of the mandible to locate
bifurcation
2. Longitudinal colour scan to identify areas of
abnormal flow and disease
3. Positive identification of the external carotid
artery and internal carotid artery
4. Spectral Doppler
(a) in normal vessels take readings from
common carotid artery, internal carotid
artery, external carotid artery
(b) in abnormal vessels take readings from Fig. 3.1 The left carotid bifurcation on transverse
areas of disease in addition to standard scanning from a lateral approach. The external carotid
readings from common carotid artery,
internal carotid artery, external carotid artery artery (ECA) and a small branch vessel lie anteriorly
5. Examine the vertebral arteries with the internal jugular vein (IJV) lying laterally, the
44 internal carotid artery (ICA) lies behind the ECA.
3

be required and, in a few individuals, the approach allows access to the carotids and other cervical
may be from under the mastoid process and structures, although there is some impairment
behind the sternomastoid muscle. In patients who of resolution.
have undergone recent carotid surgery, access
can be problematical due to the skin incision Identification of the internal and external
and oedema of the soft tissues, so that a variety carotid arteries
of approaches may need to be tried, or a lower The common carotid artery on the right arises
frequency transducer may be successful; suffi- from the brachiocephalic artery behind the right
cient information can usually be obtained to sternoclavicular joint (Fig. 3.2), where the origin
confirm flow in the vessel, or the absence of can usually be seen on ultrasound. On the left it
flow. Beards are not usually a problem but if usually arises directly from the aorta, so that its
they are particularly extensive and luxuriant origin on the left cannot be seen on scanning
they may interfere with access; liberal applica- from the neck. The level of the carotid bifurca-
tion of gel to exclude air between the hairs usually tion is usually at about the level of the upper
Basilar artery
External
carotid artery
Internal
carotid artery
Common
Superior
carotid artery
thyroid artery
Common Vertebral artery

carotid artery
Vertebral artery
Subclavian artery
Brachiocephalic
trunk
Fig. 3.2 Carotid and vertebral arteries. 45

3
border of the laryngeal cartilage but it may vary
Table 3.4 Identification of the external and
considerably. The two branches of the common internal carotid arteries
carotid artery are the internal carotid artery and
The external carotid artery
the external carotid artery. It is essential that they
Branches present
are identified positively, otherwise there is the possi-
Anterior position
bility that disease in one vessel will be mistakenly
Waveform characteristics:
attributed to the other, which may lead to further
High resistance pattern with relatively little
inappropriate investigations.The external carotid diastolic flow
artery is usually the easier of the two branches at Appears more pulsatile on colour Doppler
the bifurcation to recognise positively and the Dichrotic notch is more prominent
criteria to look for are listed in Table 3.4. Positive ‘temporal tap’
The external carotid artery has branches just The internal carotid artery
above the bifurcation (Fig. 3.3); the superior The other branch of the bifurcation
thyroid, ascending pharyngeal and lingual Bulb at origin
arteries may all arise from the external carotid Posterior position and course angled posteriorly
artery below, or around, the level of the angle of Less pulsatile waveform on colour Doppler with
the mandible. relatively high diastolic flow
The external carotid artery is nearly always

the more anterior of the two branches. In one
study it lay anteromedial to the internal carotid teristic waveform on spectral Doppler with rela-
artery in 48.5% of bifurcations studied, anterior tively low diastolic flow (Fig. 3.4a). In addition
in 34.5% and anterolateral in 13%; other posi- the dichrotic notch of the pulse wave is usually
tions accounted for only 4% of vessels.13 more prominent in the external carotid artery
The external carotid artery supplies the rela- spectrum than in the internal carotid artery
tively high-resistance vascular bed of the facial spectrum.
muscles, pharynx, tongue and scalp. Therefore The superficial temporal artery is one of the
the external carotid artery has relatively less terminal branches of the external carotid artery,
diastolic flow, which makes it appear more and if this is tapped by a finger as it passes over
pulsatile on colour Doppler and to have a charac- the zygoma it will produce rapid, clear fluctua-
Fig. 3.3 The external carotid artery

showing a branch arising just above
the bifurcation and fluctuations
induced in the spectrum by tapping
the superficial temporal artery at
the level of the zygoma.
46
3

tions in the waveform in the external carotid at this level but the bulge of the carotid bulb
artery, whereas there is generally little or no is usually apparent in subjects without severe
effect in the ipsilateral common carotid artery or disease. Colour Doppler will show the normal
internal carotid artery14 (Fig. 3.4a). area of reversed flow in the carotid bulb, some-
Once the external carotid artery has been times referred to as the boundary layer separa-
identified positively then it can be assumed that tion zone.The spectrum from the internal carotid
the other large vessel arising from the carotid artery is less pulsatile and more sustained than
bifurcation is the internal carotid artery. This that of the external carotid artery, with relatively
vessel is nearly always the more posterior of the high diastolic flow (Fig. 3.4b).
two branches and tends to run deeply and more Diseased vessels may be more difficult to
posteriorly. It does not have visible branches distinguish as plaques can obscure visual details;
a Fig. 3.4 (a) The external carotid

artery and the characteristic
spectrum seen in normal vessels.
There is relatively low diastolic
flow and a prominent dichrotic
notch compared with the
spectrum from the internal
carotid artery. (b) The internal
carotid artery and its characteristic
spectrum with more flow
throughout diastole. The normal
area of reversed flow in the
carotid bulb is clearly visible.
47
3
local and remote disease can lead to alterations cations may be very difficult to see well enough
in the normal patterns of flow, so that distinction to allow reliable assessment; in this situation scan-
on the basis of the appearance of the waveform ning transversely with colour Doppler switched
may be impossible. In addition some high bifur- on may allow localisation of the internal and
a Fig. 3.5 (a) The angle

correction has been made
aligning the cursor to the vessel
wall resulting in a velocity
estimate of approximately
2 m s–1. (b) Using colour
Doppler, the cursor is positioned
along the line of the flow and
results in a velocity of
approximately 3.5 m s–1.
48
3

external carotid arteries, so that some spectral from the intima-media complex are most easily
information can be acquired. distinguished. The machine settings should be
set to give a clear, uncluttered image of the vessel
Standard velocity measurements wall and the position of the transducer adjusted
Once the bifurcation and its branches have been to show the characteristic double-line appearance
identified and assuming that no areas of signi- of the vessel wall (Fig. 3.6). The image should
ficant disease are present, it is good practice to be magnified as much as possible to make the
take peak systolic velocity measurements from measurement easier to perform. The intima-
the common carotid artery, the internal carotid medial thickness (IMT) is best demonstrated in
artery and the external carotid artery in order to the upper common carotid artery on the poste-
have a record of the examination. These are rior wall where the vessel is usually at right angles
obtained using spectral Doppler from the upper to the ultrasound beam.The internal carotid artery
common carotid artery 2–3 cm below the is more difficult to assess as the vessel slopes
bifurcation; the internal carotid artery from 1 to obliquely away from the transducer face in many
2 cm above the bulb, or as high as possible, in cases.With careful attention to detail it is possible
order to allow the normal bulbar turbulence to measure the IMT with satisfactory, reproducible
to settle; and from the lower external carotid accuracy. The precise upper limit of the normal
artery. For routine measurements the sample range is a matter of some discussion. It does
volume is set at about one-third of the total increase with age but values of less than 0.8 mm
diameter and placed in the centre of the vessel in correlate well with lack of coronary artery disease,
order to avoid the natural turbulence at the edge whereas an increasing thickness above this level
of the lumen and ‘wall thump’ from inclusion (Fig. 3.6) is associated with increasingly severe
of the vessel wall in the sample volume. The coronary artery disease,15, 16 an increased risk of
Doppler angle is kept as low as possible, ideally myocardial infarction17 and also stroke.18
less than 60°; it is good practice to try and keep
to a specific angle, such as 55° or 60°, in order The vertebral arteries
to improve the reproducibility of results between Once both carotids have been examined the
examinations. For tight stenoses it is better to vertebral arteries are assessed. The vertebral
reduce the size of the sample volume, as this artery on each side is the first branch of the
allows the area of the peak systolic flow signal to subclavian artery (Fig. 3.2). It passes posteriorly
be better defined. Colour Doppler allows more and upwards to the vertebral foramen in the
accurate assessment of the direction of flow in a transverse process of the sixth cervical vertebra
stenosis as this may not always be parallel to the (V1 segment), and from there it passes upwards
walls of the vessel (Fig. 3.5). The precise final in the vertebral canal to the level of the axis (C2)
location of the sample volume is chosen using a (V2 segment). It emerges from the vertebral
combination of the audible Doppler signal and canal at C2, passing behind the lateral mass of
the spectrum so that the clearest, highest-frequency the atlas (C1) to enter the skull through the
audible signal and the best spectral trace are foramen magnum (V3 segment) and runs ante-
obtained; in stenoses, the sample volume should rior to the brain stem (V4 segment) to join with
be moved through the length of the stenotic the vessel from the other side in front of the
segment in order to locate the peak signal. brain stem to form the basilar artery.19
The vertebral arteries are most easily located
Measuring the intima-medial thickness by placing the transducer longitudinally over the
This is not always required but should it need to common carotid artery and angling it medially
be measured, for instance as part of a population so that the vertebral bodies are identified; the
survey, then it can be measured on an image of transducer is then rotated laterally so that the
the distal common carotid wall where the echoes transverse processes of the vertebrae and the 49
3
a Fig. 3.6 (a) A normal intima-
medial thickness (IMT) of
0.7 mm measured in the
upper common carotid artery.
(b) Moderate thickening of the
IMT at 1.2 mm (c) More marked
intimal thickening of 1.4 mm.
spaces between them are visualised, the vertebral in the lower neck as it passes backwards from the
artery and vein may then be seen in these gaps subclavian artery towards C6; or in the upper neck
(Fig. 3.7). If the vertebral artery cannot be iden- behind the mastoid process as it passes around
50 tified in the vertebral canal, it may be looked for the atlas (C1) and into the foramen magnum.
Fig. 3.7 The vertebral artery seen

3

between the transverse processes
of the cervical vertebrae.
There may be marked variation in the size of Direct visualisation and measurement
the vertebral arteries and their relative contri- If the stenosis and plaque can be seen clearly
bution to basilar artery flow; when there is then it is possible to measure the calibre of the
disparity of size, the left artery is usually the residual lumen and the original calibre of the
larger of the two and in 7–10% of individuals vessel. Diameter reduction ratios, or area reduc-
there are significant segments of hypoplasia, tion ratios, are the usual methods for describing
which result in the artery not being visible.20 the reduction in vessel calibre; percentage residual
Clear visualisation of the vein but not the artery lumen can also be used. Diameter measurements
suggests that the artery may be either throm- are generally a little quicker to perform but are
bosed or congenitally absent. slightly less representative of the stenosis, as they
Colour Doppler makes assessment of flow do not take account of variations in plaque
direction in the vertebral arteries straightforward thickness around the circumference of the vessel
(Fig. 3.8). They should have the same colour as and there is the potential to underestimate, or
the common carotid artery in front of them. overestimate, the degree of stenosis (Fig. 3.9).
Care needs to be taken in the diagnosis of reversed Care must be taken to examine a diseased
flow, particularly if the spectral Doppler trace segment of vessel in both transverse and longi-
has been inverted during the examination: if sub- tudinal views so that the distribution of plaque
clavian steal is suspected it is worth confirming can be clearly assessed and the most appropriate
that the machine is set up appropriately in order diameter measurement can be made; this is usually
to avoid making an error. the shortest diameter. Measuring stenoses by
area reduction, although more time consuming,
Assessment of disease overcomes this problem with the eccentricity of
Measurement of the degree of stenosis the plaque being taken into account as the
Two types of information can be used to assess luminal and vessel areas are measured.
the degree of stenosis: direct measurement using It is important that the type of measurement
the calipers on the machine; and velocity criteria used is clearly defined as either a diameter
derived from spectral Doppler. reduction or an area reduction, because signifi- 51
3
a Fig. 3.8 (a) Reversed flow in
the vertebral artery in a patient
with a proximal left subclavian
artery stenosis. The vertebral
artery is the opposite colour
(blue) from the common carotid
artery. (b) Biphasic flow in the
vertebral artery of a patient with
a developing subclavian steal
syndrome.
cant misunderstandings may occur in the inter- essential if the measurement is in a different
pretation of the results. For a given stenosis, a form from that normally used.
50% diameter reduction corresponds to a 70% Most stenoses are relatively short in longi-
area reduction, so that misinterpretation of a tudinal extent, usually no more than a centi-
70% area stenosis as a diameter reduction will metre for the maximum degree of narrowing.
result in a significant overestimation in the assess- Some patients, however, have longer segments
ment of calibre reduction, possibly leading to of varying calibre reduction and it is important
unwarranted surgery (Fig. 3.10). It is good prac- to remember that, although the degree of pres-
tice always to define the value of a stenosis as sure drop across a stenosis is related primarily to
52 either an area or a diameter reduction and this is the reduction in radius, it is also related to the
3

of the mandible), high distal extent of the
stenotic segment (>2 cm above the bifurcation)
and a small internal carotid artery (<0.5 cm), or
a kinked internal carotid artery, can complicate
surgery and prior warning will allow the appro-
priate surgical technique to be used. Ultrasound
can predict these features satisfactorily and supple-
mental arteriography is not usually necessary.21
Doppler criteria
Fig. 3.9 Transverse view of a plaque in the common
In many cases the region of the stenosis is not
carotid artery with the diameter reduction calculated
seen clearly due to complex plaque structure and
from the diameter measurements. Note that an
inappropriate longitudinal scan plane (x–x) could result calcification. In these cases direct measurement
in a significant underestimation of the degree of cannot be used to quantify the degree of stenosis
stenosis. and Doppler criteria must be used. Over the years,
much work has been done correlating Doppler
findings with degrees of stenosis found on arterio-
length of the stenosis. However, length has a graphy, or at surgery. It has been shown that
much smaller effect, as the resistance is related carefully obtained Doppler criteria correspond
to the first power of the length, rather than the well with the degree of stenosis, and values which
fourth power of the radius (Poiseuille’s Law): allow the severity of internal carotid artery stenoses
to be predicted have been developed. However,
R = 8lη
the literature can be confusing, with apparently
π r4
widely varying velocities being quoted for specific
If the patient is being considered for endar- levels of stenosis. One of the first studies sug-
terectomy, in addition to measuring the degree gested that a peak systolic velocity of >1.3 m s-1
of stenosis, it is important to assess the level of was appropriate for diagnosis of a diameter
the bifurcation, the length of the stenosed stenosis of 60% or greater.22 However, other
segment and the diameter of the internal carotid workers have reported velocities of 1.7 m s-1 for
artery above the stenosis. The reason for this is a 60% stenosis,23 2.25 m s-1 for a 70% stenosis24
that high bifurcations (<1.5 cm from the angle and 1.3 m s-1 for 70% diameter stenosis.25 This
a b
Fig. 3.10 A stenosis measured with (a) an area-reduction calculation (measured as 62%) and (b) a diameter-
reduction calculation (measured as 40%). 53
3
apparent lack of consensus emphasises the fact flow if they are present. The IC/CC diastolic
that there is some variation from one department ratio can also be measured but this does not
to another depending on the equipment and tech- usually add to the information obtained from
nique used. Each department must therefore the three main criteria. The main levels which
develop and audit criteria which they find work need to be distinguished are 50% diameter
in their environment and complement the clinical reduction, where blood flow starts to decline,
criteria and practices used in their institution. and 70% diameter reduction, which is the level
The peak systolic velocity, end-diastolic velocity strongly associated with clinical symptoms and
and the ratio of peak systolic velocities in the for which surgery will be considered. The values
internal and common carotid arteries (IC/CC for the criteria which the authors have found to
systolic ratio) are the most useful measurements be useful in their practice to predict these levels
in general practice24,26 (Fig. 3.11). Spectral of stenosis in the internal carotid artery are
broadening and filling in of the window under shown in Table 3.5 and are based on those
the spectrum are subjective, difficult to quantify reported by Robinson et al24 and Grant et al.26
and can be affected significantly by gain control It is important to remember that the peak
settings; however, they do indicate abnormal systolic and diastolic values refer only to the internal
Fig. 3.11 A stenosis of the

internal carotid artery showing a
peak systolic velocity in excess of
2.6 m s-1 and an end-diastolic
velocity of 1.9 m s-1. There is aliasing
of both the colour and spectral
Doppler displays.
Table 3.5 Diagnostic criteria for Doppler diagnosis of stenoses

of 50% and 70%
Diameter Peak systolic End-diastolic IC/CC systolic

stenosis velocity ICAa velocity ICA ratio
(%) (m s–1) (m s–1)
50 >1.25 >0.4 >2
70 >2.3 >1.0 >4
ICA, internal carotid artery; IC/CC, internal carotid/common carotid. From Robinson
et al24 and Grant et al.25
54
3

carotid artery, not to the common carotid artery In addition to the direct measurement of the
or external carotid artery. Furthermore, it should residual lumen demonstrated by the colour map,
be borne in mind that physiological variations a cursor with angle correction can be placed over
due to heart rate, cardiac output and contralateral the colour map and used to provide an estimate
occlusion may affect the velocities in a vessel, of the mean velocity in the underlying pixel.This
potentially leading to a false diagnosis of a patho- allows the mean velocities in a stenotic segment
logically high velocity, although these cases should to be estimated. However, these correlate less
be clarified by the use of the velocity ratios. It well with the degree of stenosis than peak systolic
should also be remembered that peak velocities or diastolic velocities. Although experienced
decline with very high degrees of stenosis (>90% operators can often get a good idea of the severity
diameter stenosis)27 as discussed in Chapter 2. of a stenosis from the overall appearances and
Attempts have been made to use colour or the colour map changes, it is always better to use
power Doppler criteria to assess the severity of the colour map to identify areas of abnormal
stenoses.28, 29 Direct measurement of the residual flow and use this to position the sample volume
lumen based on power or colour Doppler can be for spectral Doppler analysis.
made but it is essential that care is taken with
the gain settings to ensure that there is not any Effects of disease elsewhere
over- or underestimation of the residual lumen; The velocities and flow characteristics seen in
this type of direct measurement will always be any given section of a vessel depend not only on
less accurate than measurements made on a good local conditions but also on conditions elsewhere
B-mode image but, for more severe degrees of along the vessel, in other vessels connected to
stenosis (>50% diameter reduction), they can that vascular territory (Fig. 3.12) and to other
provide additional confirmatory information. For factors, such as heart rate, cardiac output and
cases where a diagnosis must be made between blood pressure (Table 3.6). The best example
‘subtotal’, or ‘near’ occlusion and complete of this is vertebral or subclavian steal, where a
occlusion, then careful setting up of the colour proximal occlusion of the subclavian artery
and power Doppler modes is essential for reliable results in reversed flow in the ipsilateral verte-
distinction (see below). bral artery.
Fig. 3.12 Common carotid

waveform in a patient with aortic
valve incompetence showing
reversed diastolic flow.
55
3
Carotid occlusion
Table 3.6 Factors affecting the waveform
Occlusion can affect the internal carotid artery
Local Atheroma and plaques (Fig. 3.13) or common carotid artery separately,
Tortuosity
or together. Occlusion of the common carotid
Proximal Common carotid artery origin
disease artery does not always result in occlusion of the
Aortic valve disease internal carotid artery, as sufficient blood flow
Distal Carotid siphon disease may be provided by retrograde flow down the
Intracranial vessel disease
ipsilateral external carotid artery to maintain
Remote Contralateral carotid occlusion
patency of the internal carotid artery.This pattern
Physiological High cardiac output states
of abnormal flow may be quite confusing if it is
not recognised but it is of clinical importance, as
these patients can still suffer ischaemic events in
the relevant internal carotid artery territory.31
Carotid occlusion can lead to an increase in The opposite pattern of flow may be seen in a
the volume of blood flowing in the contralateral small number of patients with common carotid
carotid artery. This is achieved primarily by the artery occlusion with reverse flow in the internal
blood flowing more quickly, therefore there is carotid artery on the side of the common carotid
the potential to mistakenly suggest that an artery occlusion (a carotid steal phenomenon)
increased velocity measurement is compatible and antegrade flow in the ipsilateral external
with a degree of stenosis but review of the colour carotid artery. If the lower margin of the
Doppler image should refute this impression in occluded segment is above the level of the
a non-diseased carotid. The use of ratios, mainly bifurcation then a characteristic pattern of
the IC/CC peak systolic ratio, helps identify the forward and reverse flow (‘stump thump’) is
nature of increased velocities in both normal seen in the patent residual lumen of the internal
vessels and those with minor atheroma, as in carotid artery.
these circumstances the velocity in both the If an occlusion of the internal carotid artery is
common carotid artery and the internal carotid suspected it is essential that care is taken to
artery is increased and the ratio does not alter; ensure that the Doppler settings on the machine
whereas if there is significant local disease are appropriate for locating any low-velocity,
affecting the internal carotid artery, the velocity small-volume flow that may be present in a
in this vessel is increased but the common
carotid artery velocity is unchanged, so that the
ratio is also increased.
In patients with bilateral severe carotid
stenoses, when one side is stented or operated
upon, the improvement in flow up the operated
vessel will reduce flow up the contralateral
carotid and this may result in a decrease in peak
velocities on the untreated side that might lead
to this stenosis being down-graded in severity. In
one series of patients with bilateral significant
stenoses on duplex scanning,30 the stenosis on
the non-operated side was reclassified as non-
haemodynamically significant in 20% of cases. It
is therefore necessary for these patients to be
reassessed prior to any management decisions Fig. 3.13 An occluded internal carotid artery.
56 relating to the untreated artery. The occluded lumen is indicated by the arrows.
3

narrow residual lumen in the segment under the external carotid artery; this results in a
investigation. Colour and power Doppler assess- mistaken impression of a patent internal carotid
ments of the vessel should be carried out with artery, if it is not recognised. The ‘temporal tap’
the system set for maximum sensitivity in order manoeuvre can usually clarify the situation.
to identify any small residual lumen (‘string sign’ Occlusion of the carotid on one side may result
or pseudo-occlusion) (Fig. 3.14) or adjacent in increased flow through the contralateral carotid
vessels that might confuse the situation.32 Spectral vessels, as discussed earlier.
Doppler should also be performed carefully at Recanalisation of an occluded carotid artery
maximum sensitivity, although care must be is not as rare as might be imagined.35 One series
taken to identify and exclude any waveforms followed eight patients with internal carotid artery
arising from adjacent vessels.33 Conversely, colour occlusion with serial 6-monthly ultrasounds for
Doppler or power Doppler may show the loca- a period of up to 8 years, all of the eight patients
tion of a residual channel in a vessel that was showed evidence of spontaneous recanalisation
otherwise thought to be occluded; echo-enhancing occurring between 6 and 96 months.36
agents are of value if there is any persisting If there is occlusion of both the internal
uncertainty. carotid artery and external carotid artery then
Occlusion of the internal carotid artery results the ipsilateral common carotid artery also usually
in the reduction of diastolic flow in the ipsilateral occludes. Before thrombosis occurs, however, a
common carotid artery, so that the common to-and-fro pattern of flow may be seen in the
carotid artery waveform becomes more like that common carotid artery, which signifies that there
seen in the external carotid artery. Reduction of is no net forward flow of blood up the vessel.
common carotid artery end-diastolic flow may
therefore be an initial clue to the presence of an Plaque characterisation
internal carotid artery occlusion, or a very severe Much effort has been expended in attempting to
stenosis.34 Care should be taken, however, as the classify atheroma and plaques on ultrasound,
development of collateral channels between the particularly in the carotids, where high-resolution
external carotid artery and internal carotid ultrasound gives good images of many plaques.
artery circulations in the orbit and meninges can Steffen et al37 proposed a classification of plaque
result in ‘internalisation’ of the external carotid which takes account of the different types of
artery flow with relatively high diastolic flow in plaque and its components (Fig. 3.15). Types 1
a
b
Fig. 3.14 A thin residual channel on colour Doppler (a) and on power Doppler (b) which is more sensitive for
weaker, slower signals and shows the tortuous residual lumen more clearly. 57
3
a
Fig. 3.15 Different types of plaque seen on ultrasound. (a) Type 1, dominantly echolucent with a thin
echogenic cap. (b) Type 2, substantially echolucent lesions with small areas of echogenicity. (c) Type 3,
dominantly echogenic lesions with small areas of echolucency of <25%. (d) Type 4, uniformly echogenic lesions.
and 2 were predominant in symptomatic arteries, of hypoechoic areas within the plaque. However,
whereas types 3 and 4 were more common in it is also possible that many of these areas are
asymptomatic patients; these findings have been aggregates of lipid or necrosis, rather than areas
supported by other groups.38,39 This supports of haemorrhage. Sometimes an ulcer in the
the suggestion that the more friable, lipid- plaque can be clearly seen, but many plaques are
containing, soft plaques are more likely to result irregular without being ulcerated and, conversely,
in plaque disruption and produce symptoms an ulcerated plaque may not be identified on
than firmer, more fibrous and coherent plaques. ultrasound. Thrombus adherent to the surface is
The value of ultrasound in predicting the suggested by an anechoic or hypoechoic area
complications associated with plaques is more adjacent to the plaque surface on colour or
difficult to define. These complications include power Doppler. It is important that the system is
intraplaque haemorrhage, surface ulceration and set up appropriately, otherwise the lack of colour
adherent thrombus. The presence of intraplaque on the image may be due to technical factors,
58 haemorrhage has been inferred from the presence rather than the presence of thrombus.
3

Some studies have shown a good correlation
Table 3.7 Classification of carotid plaques
between the ultrasound appearances and those
found at operation,40,41 but the results from Haemodynamic classification
others have been less satisfactory, with poor H1 0–20% Diameter reduction Normal to mild
prediction of ulceration or haemorrhage.42,43 H2 20–60% Diameter reduction Moderate
Another study from the Seattle group showed H3 60–80% Diameter reduction Severe
that there was no significant difference in plaque H4 80–99% Diameter reduction Critical
constituents between the endarterectomy speci- H5 Occluded
mens removed from symptomatic and asymp- Morphological components
tomatic patients with high grade stenoses.44 P1 Homogeneous
Visualisation of plaque type in high grade P2 Heterogeneous
(>60% diameter reduction) stenoses is usually Surface characteristics
inadequate to identify reliably plaque charac- S1 Smooth
teristics and the Seattle study would suggest that S2 Irregular (defect <2 mm)
the degree of stenosis is the more important S3 Ulcerated (defect >2 mm)
factor in this group of patients. However, if there From Thiele et al,45 with permission.
is a good view of the diseased segment, the
plaque can be described in terms of its type
(1–4), extent (focal, diffuse, circumferential) and clearly seen, otherwise plaques should be
any obvious associated complications (ulceration, described as smooth or irregular. It should be
thrombus, haemorrhage) (Fig. 3.16). If visuali- remembered that many diseased segments are
sation is moderate or poor then discretion is not clearly seen due to the presence of calcifi-
necessary and only those features which are cation, which makes any attempt at plaque
clearly seen should be noted. For example, it characterisation very difficult, or impossible.
may be difficult to distinguish between a plaque In practical terms, a smooth, homogeneous,
ulcer and a gap between two adjacent plaques, predominantly echogenic plaque is less likely to be
or to decide if a colour void associated with the associated with symptoms, whereas an irregular,
plaque surface is really due to adherent thrombus heterogeneous or hypoechoic lesion is of greater
or to technical factors. Ulceration should only concern. A stenosis of only 50% (diameter reduc-
be diagnosed if the plaque and the ulcer are tion) but with an unstable plaque type in a symp-
b
a
Fig. 3.16 An ulcerated plaque. A thin rim is seen on the B-scan image (a) but colour Doppler (b) shows flow
within the plaque. 59
3
tomatic patient might well be a matter of clinical
concern and considered for surgery.
An attempt has been made to standardise these
descriptions for carotid disease in relation to ultra-
sound and other non-invasive modalities.45 This
proposes that lesions can be described in terms of
the degree of stenosis, the morphological plaque
components and the surface characteristics, where
these can be clearly visualised. The suggested
classification is given in Table 3.7. A lesion listed
as H4, S2, P2 therefore represents a lesion which
is producing a stenosis of more than 80% diameter
reduction (H4), which has an irregular surface
(S2) and is heterogeneous (P2). In practice, this
type of classification does not add significantly to
the information that can be provided by a descrip-
tion of the diseased segment as discussed above. Fig. 3.17 Tumour recurrence from an oropharyngeal
In addition, the original classification for the malignancy closely applied to the carotid sheath.
degree of stenosis does not allow for distinction
between <70% and >70% diameter reduction, Lymph nodes and other masses adjacent to
as H3 covers the range of 60–80%. the carotids will transmit pulsations and require
distinction from intrinsic vascular lesions.This is
Pulsatile masses not usually a problem, but occasionally a deposit
The main causes of pulsatile neck masses are will surround the carotid artery (Fig. 3.17) and,
given in Table 3.2. Normal but prominent carotids unless there is a previous history of malignancy,
and ectatic carotid or subclavian arteries are easily diagnosis can be difficult. Adherence to the carotid
identified using colour Doppler and do not sheath can be assessed by gentle palpation and
usually require any further investigation. getting the patient to swallow so that relative
Aneurysms of the carotid arteries can also be movement between the mass and the carotid can
identified as they are in continuity with the be assessed.47 Colour Doppler ultrasound also
artery. The majority arises following surgery but has an important role in defining the solid nature
they may also occur following trauma, including of a neck lump prior to biopsy and excluding a
whiplash neck injuries and biopsy of cervical vascular lesion, such as an aneurysm.
masses. The flow in the aneurysm may be seen Carotid body tumours are rare but can be
with colour Doppler, unless there is thrombosis diagnosed easily with ultrasound. Character-
of the lumen of the dilated segment. In some istically there is a hypoechoic mass between the
cases of aneurysm of the common carotid artery, two branches at the bifurcation, spreading them
it may be difficult to identify the internal carotid apart in a ‘wine glass’ deformity (Fig. 3.18).
artery above the dilatation and care must be Colour Doppler shows a highly vascular lesion
taken to establish whether it is patent or not; and the external carotid usually shows a low
flow in the ipsilateral ophthalmic artery is not resistance pattern of flow on spectral Doppler.48
necessarily evidence of patency, as this may come
from collateral filling via the circle of Willis. Dissection of the carotid arteries
Aneurysms of the upper internal carotid artery The ultrasound findings in this condition can
may be difficult to identify with certainty, or the vary considerably. The vessel may be occluded
findings may be misinterpreted as a straight- completely; show a smooth tapering stenosis, with
60 forward stenosis or dissection.46 or without a recognisable haematoma/thrombosed
3

problem, when a more detailed examination of
the vessels is required.19 As noted previously,
there are several variations of anatomy and size
of the vertebral arteries which may affect the
ultrasound findings; a further issue to be
considered is that disease affecting one artery
may be compensated by the other side, or
collaterals from the circle of Willis, so that blood
flow to the posterior fossa circulation is main-
tained and local, unilateral vertebral artery disease
is of little clinical significance. Nevertheless,
some basic patterns of flow can be identified:
1. No flow is detected in the region of the
Fig. 3.18 A carotid body tumour splaying apart the
relevant vertebral artery means that the artery
internal carotid artery and external carotid artery. is either hypoplastic, absent, or thrombosed.
Colour Doppler shows the abnormal tumour circulation. Visualisation of the artery on B-mode with-
out demonstrable flow within it on colour/
false lumen being visible (Fig. 3.19); or a mem- power and spectral Doppler with the system
brane with a double lumen may be seen with set to maximum sensitivity suggests that it is
variable flow patterns in the two channels on thrombosed, or dissected.
either side.49,50 Recanalisation of the occluded 2. Demonstration of a focal area of increased
vessel is a recognised occurrence and occurs in velocity at the origin of the vertebral artery
up to 60% of cases.51 Dissection of the vertebral (Fig. 3.20), or at some point along its course
arteries may also occur and is usually manifest between the subclavian artery and the foramen
as absent flow in the affected artery.50 magnum, is consistent with a focal stenosis, the
significance of which will depend on the clinical
Vertebral arteries situation and the state of the contralateral verte-
Frequently only cursory attention is paid to the bral artery.Visualisation of the vertebral artery
vertebral arteries, unless there are specific signs is normally inadequate to make a reliable direct
and symptoms pointing to a posterior fossa diameter reduction measurement feasible.
Fig. 3.19 Dissection of the

common carotid artery extending up
from a dissection of the aortic arch.
One channel has thrombosed,
producing a haematoma in the wall
of the vessel posteriorly which is
narrowing the residual lumen.
61
3
occlusion or tight stenosis of the proximal
subclavian artery at its origin and blood
supply to the arm is maintained by reversal of
blood flow down the ipsilateral vertebral artery.
6. A biphasic vertebral artery waveform may
be seen in patients with a developing steal
situation from slightly less severe subclavian
stenoses and, in some patients, reversed flow
may only occur with the arm in certain posi-
tions, or after a period of exercise; therefore
scanning after a period of arm exercise, such
as elbow flexions holding a book, or some
other relatively heavy object, should be
considered if a steal syndrome is suspected.
Fig. 3.20 A stenosis at the origin of the vertebral Alternatively a pressure cuff can be inflated
artery with marked turbulence shown on colour
to occlude blood flow to the arm and then
Doppler.
released after 2–3 min. The resulting reactive
3. A ‘tardus parvus’ waveform suggests a proximal hyperaemia in the arm produces an increased
stenosis; this should be sought by careful exam- demand for blood and reversal of flow in the
ination of the artery from the subclavian artery relevant vertebral artery.
up to the foramen magnum (Fig. 3.21).
4. The proximal vertebral artery waveform shows Reporting carotid ultrasound
reduced or absent diastolic flow, implying a examinations
distal stenosis/occlusion (Fig. 3.22). Unlike an arteriogram, or magnetic resonance
5. Reversed flow is consistent with subclavian angiography (MRA) examination, there are no
steal syndrome. This occurs when there is an easily interpretable images from a Doppler
Fig. 3.21 Vertebral artery with a

damped ‘tardus parvus’ waveform
indicating a proximal stenosis.
62
Fig. 3.22 A difficult vertebral

3

artery to image but spectral
Doppler shows absent diastolic
flow, this together with the low
velocities suggest a distal
obstruction. In addition, there is
significant interference and noise
resulting from the patient going
to sleep and snoring during the
examination.
examination as the outcome is a mixture of image of the radius; however, as noted previously, it is
assessment, Doppler parameters and clinical judg- also related to the length of the stenosis. This is
ment. It is therefore sensible to standardise the not relevant for many stenoses, as they are rela-
information given in the report of a Doppler tively short, but some stenoses, particularly in
examination, to ensure that all necessary data the common carotid artery, may be longer.There-
are recorded. The easiest way to do this is to
record the data onto a standardised form, which
Table 3.8 Carotid examination report
can then be used in reaching management deci- information
sions about the patient, or comparing the Doppler
Patient demographic data etc.
examination with any associated MRA, computed
Patency of CCA, ICA, ECA
tomography angiography (CTA), or arteriogram.
Any variations of standard anatomy
The precise details will vary from centre to centre,
PSV, EDV for each CCA, ICA and ECA
depending on local preferences. The form used
IC/CC peak systolic ratio for each side
in our institution is shown in Figure 3.23. The
Estimate of degree of stenosis
main information which should be noted is
Description of any plaque clearly seen
given in Table 3.8.
Plaque type: lucent/echogenic/calcified
Plaque surface: smooth/irregular/ulcerated/
Problems and pitfalls in carotid thrombus
ultrasound Length of any stenosis
Problems and pitfalls can arise from a variety of Diameter of ICA above any stenosis
sources. These can be divided into those result- Diagram of any stenosis giving an estimate of
ing from poor or faulty technique and those plaque disposition
arising from pathological or physiological causes Level of the bifurcation in relation to angle of
mandible on each side
(Table 3.9). Technical aspects of setting up the
Vertebral arteries
system are discussed elsewhere (see Appendix) Visualised/not visualised
but other aspects which can lead to problems Direction of flow
should be considered. Any specific abnormalities identified in relation
to the vertebral arteries
Long and eccentric lesions CCA, common carotid artery; ECA, external carotid artery;
The amount of pressure reduction across a ICA, internal carotid artery; IC/CC, internal carotid/common
carotid.
stenosis is related primarily to the fourth power 63
3
Vascular Laboratory
Royal Infirmary of Edinburgh - Little France

51 Little France Crescent
Edinburgh
EH16 4SA
Please attach patient label Tel:
Summary
Duplex ultrasound carotid and vertebral assessment
Right
CCA ICA ECA
PSV (m/s) 0.7 m/s 3 m/s 1.4 m/s
R L
EDV (m/s) 1 m/s
ICA/CCA = 4.3
I E I E % stenosis <50% > 70% < 50%
Plaque mixed mixed mixed
appearance smooth smooth smooth
Vertebral
PSV (m/s) 0.5 m/s
Direction Antegrade
Left
CCA ICA ECA
PSV (m/s) 0.8 m/s 1.7 m/s 1.6 m/s
EDV (m/s) 0.4 m/s
ICA/CCA = 2.1
% stenosis <50% 50-69 % < 50%
Plaque mixed mixed mixed
appearance smooth smooth smooth
Vertebral
PSV (m/s) 0.6 m/s
Direction Antegrade
Signature
Please print name and designation
Date
Fig. 3.23 A specimen carotid examination report form.
fore, whilst a 40% diameter stenosis extending Eccentric lesions may cause problems if the
over 5–8 mm length is not haemodynamically exact disposition of plaque around the circum-
significant, a 40% diameter stenosis extending ference of the vessel is not appreciated (Fig. 3.9).
over 5–8 cm may well result in some reduction Care should be taken to examine areas of disease
in pressure and a decrease in flow, resulting in transversely, as well as longitudinally, as discussed
cerebral perfusion problems in some circum- earlier. Another problem with eccentric lesions
stances, and this should be taken into account in is that the high-velocity jet may emerge from the
64 the assessment of the patient. stenosis at an unusual angle that is not parallel
3

tivity of power Doppler and the increase in signal
Table 3.9 Problems and pitfalls
strength obtained with echo-enhancing agents
Technique related are two developments which both individually
Incorrect Doppler sample volume position or size and together improve the success of ultrasound
Doppler angle too large (>60–65°) in discriminating between an occlusion and a
Doppler settings too high for low-velocity, small residual lumen.52
low-volume flow
Good Doppler angles give poor images and vice
versa Lesions at the bifurcation
Disease related These can cause problems as a significant
Long lesions amount of plaque may be present in the bulb
Eccentric lesions but, because of the relative dilatation of the
Tortuous vessels lumen which is normally present here, this does
High-grade stenosis not result in any narrowing of the flow channel
Lesions at the bifurcation between the common carotid artery and the
Disease elsewhere in the vascular tree internal carotid artery. Although this will not
have any effect on the overall flow in the vessel,
this disease may provide a source of emboli and
to the vessel wall, as many people would assume. may not be recognised if spectral criteria alone
Colour Doppler is useful in identifying these are used to identify stenotic segments, as there
oblique jets and allows for a more accurate angle will not be a significant increase in velocity.
correction to be achieved (Fig. 3.5).
Disease elsewhere in the vascular tree
Tortuous vessels Vascular ultrasound allows direct assessment of
Sharp twists in the course of a vessel result in much of the extracranial cerebral circulation from
changes to the pattern of blood flow in the the clavicular region into the upper cervical area
lumen: blood on the outer margin flows faster and disease in these segments can be measured
than blood on the inside of the bend (Fig. 2.5); directly. However, changes in vessels remote from
in addition, turbulence is generated, even if those under examination may affect the findings
plaques are not present. Furthermore, assessing at any given point. As discussed previously,
the angle of flow for calculation of velocity can contralateral occlusion results in increased flow
be difficult, so that accurate, representative through the patent carotid with higher than
velocities are hard to define, especially if there normal velocities as a result of this. Aortic valve
is associated disease. Whenever possible, flow disease (Figs 3.12 and 3.24) or common carotid
characteristics should be assessed as far as possible artery origin stenosis on the left will not be
from the curved segment. visualised directly on carotid examination; the
presence of disease in these areas can only be
High-grade stenoses inferred from damping of the waveform with, or
In critical stenoses (>85% diameter reduction, without, turbulence. Similarly disease more distal
see Ch. 2) velocity decreases; in addition, there in the carotid siphon region will not be detected
is only a small volume of blood flowing through by cervical carotid examination.
the narrow residual lumen.The signal strength is
therefore poor and the Doppler shift is lower Carotid stents
than might be expected. It is important to ensure Insertion of stents into stenotic segments of the
that the machine is set up appropriately to common and internal carotid arteries is becoming
detect these weaker, lower Doppler shifts, if they increasingly recognised as an alternative to
are not to be missed and a false diagnosis of endarterectomy for treatment of significant
occlusion reached (Fig. 3.14). The higher sensi- extracranial carotid disease. Insertion of a stent 65
3
Fig. 3.24 Carotid waveform in a
patient with aortic valve stenosis
showing delayed systolic
acceleration.
affects the biomechanical properties of the arterial will result in different estimations of the degree
wall resulting primarily in increased stiffness; of stenosis for a bifurcation lesion depending on
this means that velocity measurements from a whether the diameter of the residual lumen is
non-stenotic stented segment will be higher than compared with the diameter of the common
those from an equivalent unstented artery. Lal carotid artery, the diameter of the internal carotid
et al53 have proposed that a normal velocity artery, the estimated diameter at the bulb as
through a stented segment can be up to 1.5 m s-1. calculated from calcification in the wall, or from
In patients with bilateral severe stenoses, stent- the alignment of the vessels. Indeed, NASCET
ing or operating on one side should result in a and ECST1,2 used different techniques for
change in blood flow in the non-operated artery measuring the degree of angiographic stenosis,
so, as discussed previously, this means that a so that the results of the two studies are not
careful review of the current situation should be directly comparable; an 80% ECST diameter
made before any decisions are made in relation stenosis for a given lesion corresponds to a
to the other side.30 stenosis of 50% in the NASCET study for the
same lesion. There is also a degree of variation
Accuracy in relation to other between different observers in the estimation of
techniques stenosis on arteriography, which may be up to
Arteriography 20%.57 In addition, angiography has some intrinsic
Many studies have shown that spectral Doppler procedural risks arising from catheterisation and
and colour Doppler have satisfactory accuracy exposure to contrast and radiation. The risk of
in the diagnosis and assessment of disease when minor stroke from carotid arteriography has been
compared with arteriography. Some care must be reported to range from 1.3 to 4.5% and in the
taken when considering studies comparing the ACAS the combined neurological morbidity/
two techniques, as different types of angiography mortality from arteriography in patients with
are used as the gold standard: direct carotid injec- haemodynamically significant stenoses was 1.2%.58
tion, arch injection, digital subtraction arterio- The overall performance of Doppler ultrasound
graphy (DSA) and venous DSA have all been compared with arteriography is good. Cardullo
used. Different methods of measuring the degree et al59 reviewed 16 spectral Doppler studies with
66 of stenosis are also used (Fig. 3.25).54–56 These 2146 Doppler–arteriogram comparisons; non-
3
Transcranial Doppler of the cerebral circulation

colour duplex Doppler had an overall sensitivity on disease at the origins of the carotids and also
of 96%, specificity of 86%, positive predictive in the intracranial circulation. However, they also
value of 89%, negative predictive value of 94% have significant drawbacks: they both ‘image’
and accuracy of 91% for the diagnosis of a the blood and infer the wall characteristics from
diameter stenosis greater than 50%. Subsequently, the shape of the blood flowing in the vessel,
further studies have confirmed the value of colour whereas colour Doppler provides information
Doppler with similar or better levels of accuracy on the wall of the vessel and any plaque, as well
and also its value in improving diagnostic confi- as the flowing blood. MRA is still limited to some
dence, clarifying difficult situations and reducing extent by signal voids at relatively high velocities,
examination times.60, 61 In particular the difficult as well as problems from signals arising in plaque
distinction between a critical stenosis and a haemorrhage and lipid deposits.63 Another dis-
complete occlusion can be achieved in nearly all advantage of CTA and MRA, apart from expo-
cases with the use of colour Doppler.33, 62 sure of the patient to contrast agents and radia-
tion (CTA), is the significant amount of post-
Magnetic resonance angiography and processing and image manipulation required for
computed tomography angiography an accurate assessment of the examination to be
Developments in CTA have made the use of CT achieved. Several authors have suggested that
for the assessment of larger vessels a realistic Doppler is satisfactory for initial screening of
proposition. MRA is also developing at a rapid patients for carotid disease and for identification
pace. These two techniques have an advantage of patients requiring surgery in most cases but
over ultrasound in that they provide information that CTA or MRA are useful additional tests
Fig. 3.25 Methods for estimating

1. ECST method percentage diameter stenosis in
arteriography.
D–d
D = % diameter stenosis
D
d
Example
Margin of bulb estimated 10 – 2
= 80% diameter stenosis
from arteriogram for D 10
2. NASCET method
D Example
4–2
Diameter of ICA taken for D 4
3. Common carotid diameter method
d D
Example
Diameter of upper CCA 8–2
is taken for D 8
67
3
which can be used instead of angiography if the technique to learn and to perform reliably as the
ultrasound is inadequate or indeterminate, or if vessels must be located without any imaging
clinical uncertainty persists.64–67 The need for information.
arteriography in the majority of patients being Modern ultrasound equipment can now be
considered for surgery is now very much reduced, configured to get some imaging detail and colour
although arguments for and against persist.68–70 Doppler information from within the adult skull
Compared with arteriography, CTA and in many cases; best results are obtained with
MRA, Doppler ultrasound is relatively cheap, dedicated transducers and software. This allows
rapid, non-invasive and accurate for the diag- localisation and positive identification of the
nosis of extracranial carotid disease. Whilst it major arteries and specific segments of these.
will not provide information on siphon disease, The main problem is the bone of the skull vault,
or aortic arch disease this is not usually a where it has been estimated that the attenuation
significant problem in most patients in relation can vary from 15–25 dB to 40–60 dB depending
to the decision to perform an endarterectomy. If on the type and thickness of the bone for a single
patients are being considered for stenting, then passage across the skull vault.74 As the sound
they will need an assessment of the aortic arch pulse has to pass through the skull on both the
and carotid origins, normally by MRA or CTA. inward and outward segments of its passage,
If a policy to operate on ultrasound alone is to there is therefore considerable loss of energy.
be implemented then the department/laboratory Power Doppler is of value in locating the vessels
must ensure that scanning protocols and results and the advent of intravascular ultrasound
are continuously reviewed, audited and validated contrast agents has improved the signal-to-noise
with care taken to identify patients who will ratio significantly; it also makes the location of
benefit from further imaging. intracranial vessels more straightforward.75
TRANSCRANIAL DOPPLER OF THE Technique

There are three potential sites of access for
CEREBRAL CIRCULATION
transcranial examinations in the adult: the trans-
The use of ultrasound to assess intracranial temporal window, the suboccipital approach and
structures is not a new phenomenon. One of the the transorbital approach. The transtemporal
first clinical applications of ultrasound as a window is used for assessment of the internal
diagnostic technique was in the assessment of carotid arteries, the middle, anterior and poste-
midline intracranial structures with A-scan rior cerebral arteries. The window is located by
equipment. More recently, high-quality imaging applying liberal amounts of acoustic coupling
and Doppler studies have been obtained in gel to the hair and skin of the temporal region in
neonates through the patent fontanelles, or front and above the external auditory meatus.
through the relatively thin bone of the neonatal Slowly moving the transducer around will allow
skull. Transcranial Doppler was first described the point of best transmission to be identified. In
by Aaslid in 198271 and the technique of pulsed some 10% of subjects it may not be possible to
transcranial Doppler has subsequently been get worthwhile images and Doppler signals;76
developed in many centres. This technique there tends to be greater attenuation in older
provides useful information on the direction and patients, females and Afro-Caribbean patients.77
velocity of blood flow and the changes which The pituitary fossa and suprasellar cistern are
may occur in these with various physiological, the most recognisable structures in the transverse
pharmacological or pathological conditions; it plane. Once these have been identified, colour
can also be used to monitor the intracranial Doppler can be used to locate the main arteries
circulation during carotid and other vascular and the direction of flow in these vessels (Fig.
68 surgical procedures.72, 73 However, it is a difficult 3.26). The ipsilateral middle cerebral artery is
3
Conclusions
The transorbital approach is used in pulsed
transcranial Doppler to assess the anterior cerebral
arteries, but it is not very convenient for colour
Doppler examinations as the transducers are rela-
tively larger. Care must be taken, if this approach
is used, to ensure that the transmit power is low
in order to reduce the risk to the retina.
The major cerebral veins and venous sinuses
are more difficult to demonstrate due to their
anatomical locations and slow flow within them,
but contrast agents have been reported to improve
this situation.75
Fig. 3.26 The anterior and middle cerebral arteries
on transcranial colour Doppler ultrasound through the Indications
transtemporal window. Transcranial colour Doppler has several advan-
tages when compared with conventional trans-
seen passing peripherally from the end of the
internal carotid artery. It cannot often be seen in
its entirety in a single scan plane and the trans- Table 3.9 Applications of transcranial Doppler
ducer position must be varied to follow it out to (TCD)
the Sylvian fissure, where it turns posteriorly.
TCD is able to provide useful information and
The origin of the contralateral middle cerebral clinical utility is established
artery can also usually be identified. The proxi- Screening children with sickle cell disease to
assess stroke risk
mal segments of the anterior cerebral arteries
Monitoring vasospasm after traumatic SAH
are also seen from the transtemporal approach.
TCD is able to provide information but clinical
The direction of flow in the ipsilateral anterior utility compared to other diagnostic tools
cerebral artery is normally away from the trans- remains to be determined
ducer and flow is towards the transducer in the Evaluation of occlusive lesions of intracranial
arteries in basal cisterns
contralateral vessel. This arrangement will be
Confirming brain death
altered if there is occlusion of the ipselateral
TCD is able to provide information but clinical
internal carotid artery and collateral flow through utility remains to be determined
the anterior communicating artery is present. In Monitoring thrombolysis of acute MCA occlusions
this situation, flow in the anterior cerebral artery Detection of microembolic signals
on the side of the occluded carotid will be reversed Monitoring during carotid endarterectomy and
towards the transducer. The posterior cerebral CABG
arteries can be seen arising from the vertebral Detection of impaired cerebral haemodynamics
with severe extracranial ICA disease
artery and passing around the cerebral peduncles.
Detection of vasospasm after traumatic SAH
The suboccipital window is located by scan-
Evaluation and monitoring of MCA territory
ning transversely in the midline under the occipital infarctions
bone. It is often better to position the transducer TCD is able to provide information but other
diagnostic tests are typically preferable
slightly to one side or the other of the midline as
Detection of L→R cardiac and extracardiac
the nuchal ligament can interfere with the clarity shunts
of the image. The vertebral arteries can be seen Evaluation of severe extracranial ICA disease
passing around the atlas and into the foramen
CABG, coronary artery bypass graft; ICA, internal carotid
magnum. The point where they join to form the artery; MCA, middle cerebral artery; SAH, subarachnoid
haemorrhage. From ‘Assessment: Transcranial Doppler
basilar artery may be seen if it lies low enough in ultrasonography’, the American Academy of Neurology.78
relation to the foramen magnum. 69
3
cranial Doppler. Visualisation of the intracranial cerebrovascular disorders, such as subarachnoid
anatomy allows rapid localisation and identifi- haemorrhage and stroke.
cation of specific vessels and specific segments
of particular vessels. The colour Doppler signal
CONCLUSIONS
allows rapid identification of flow patterns and
direction, so compression studies are less neces- Colour Doppler ultrasound provides a useful
sary in order to assess collateral flow. The ability technique for the assessment of the carotid and
to perform angle-corrected velocity measure- vertebral arteries in the neck. Careful attention
ments produces more accurate readings and the must be paid to the standard techniques which
ability to measure at specific sites allows better are used for the examinations and each centre
consistency for serial measurements. should use the Doppler criteria for stenosis that
Although still largely a research tool, the tech- provide the most accurate and reproducible
nique does have several potential applications. results in their experience. The availability of the
These were reviewed in 2004 by the American technique has reduced the necessity for carotid
Academy of Neurology (Table 3.9).78 In particular arteriography in most departments. MRA and
the ability to bring the ultrasound machine to CTA continue to improve but ultrasound
the patient allows the technique to be used to provides information on the nature of the plaque
monitor cerebral blood flow in a variety of which is not available from these techniques.
REFERENCES
1. European Carotid Surgery Trialists’ (ECST) bypass: a critical review of the literature. Eur J Vasc
Collaborative Group. MRC European Carotid Endovasc Surg 2002; 23:283–294.
Surgery Trial: interim results for symptomatic 9. Ivey TD, Strandness DE, Williams DB, et al.
patients with severe (70–99%) or with mild (0–29%) Management of patients with carotid bruit
carotid stenosis. Lancet 1991; 337:1235–1243. undergoing cardiopulmonary bypass. J Thorac
2. North American Symptomatic Carotid Endarterectomy Cardiovasc Surg 1984; 87:183–189.
Trial (NASCET) Collaborators. Beneficial effect of 10. Naylor AR, Merrick MV, Sandercock PAG, et al.
carotid endarterectomy in symptomatic patients Serial imaging of the carotid bifurcation and
with high-grade carotid stenosis. New Engl J Med cerebrovascular reserve after carotid endarterectomy.
1991; 325:445–453. Br J Surg 1993; 80:1278–1282.
3. Rothwell PM, Eliasziw M, Gutnikov SA, et al. 11. Bassi P, Lattuada P, Gomitoni A. Cervical cerebral
Carotid Endarterectomy Triallists’ Collaboration. artery dissection: a multicenter prospective study
Lancet 2003; 361:107–116. (preliminary report) Neurol Sci 2003; 24(suppl 1):
4. Cina CS, Clase CM, Haynes RB. Carotid S4–7.
endarterectomy for symptomatic carotid stenosis 12. Salonen JT, Salonen R. Ultrasound B-mode imaging
(Cochrane Review). In: The Cochrane Library, in observational studies of atherosclerotic
Issue 3. Chichester: John Wiley; 2004. progression. Circulation 1993; 87(suppl 3):II56–65.
5. Zwiebel WJ. Duplex sonography of the cerebral 13. Trigaux JP, Delchambre F, Van Beers B. Anatomical
arteries: efficacy, limitations and indications. variations of the carotid bifurcation: implications
AJR Am J Roentgenol 1992; 158:29–36. for digital subtraction angiography and ultrasound.
6. National Institute of Neurological Disorders and Br J Radiol 1990; 63:181–185.
Stroke. Clinical advisory: carotid endarterectomy for 14. Budorick NE, Rojratanakiat W, O’Boyle MK, et al.
patients with asymptomatic internal carotid artery Digital tapping of the temporal artery: significance
stenosis. Stroke 1994; 5:2523–2524. in carotid duplex sonography. J Ultrasound Med
7. Chambers BR, You RX, Donnan GA. Carotid 1996; 15:459–464.
endarterectomy for asymptomatic carotid stenosis 15. Salonen JT, Salonen R. Ultrasonically assessed
(Cochrane Review). In: The Cochrane Library, carotid morphology and the risk of coronary heart
Issue 3. Chichester: John Wiley; 2004. disease. Arterioscler Thromb 1991; 11:1245–1249.
8. Naylor AR, Mehta Z, Rothwell PM, et al. Carotid 16. Hallerstam S, Larsson PT, Zuber E, et al. Carotid
artery disease and stroke during coronary artery atherosclerosis is correlated with extent and
70
severity of coronary artery disease evaluated by external carotid arteries: diagnosis and surgical
3
References
myocardial perfusion scanning. Angiology 2004; management. J Vasc Surg 1993; 17:1019–1027.
55:281–288. 32. Krieghauser JS, Patel MD, Nelson KD. Carotid
17. Geroulakos G, O’Gorman DJ, Kalodiki E, et al. pseudostring sign from vasa vasorum collaterals.
The carotid intima-media thickness as a marker J Ultrasound Med 2003; 22:959–963.
of severe symptomatic coronary artery disease. 33. Mattos MA, Hodgson KJ, Ramsey DE, et al.
Eur Heart J 1994; 15:781–785. Identifying total carotid occlusion with colour flow
18. Kitamura A, Iso H, Ohira T, et al. Carotid intima- duplex scanning. Eur J Vasc Surg 1992; 6:204–210.
media thickness and plaque characteristics as 34. Androulakis AE, Labropoulos N, Allan R, et al.
a risk factor for stroke in Japanese elderly men. The role of common carotid artery end-diastolic
Stroke 2005; 36:32–37. velocity in near or total internal carotid artery
19. Buckenham TM, Wright IA. Ultrasound of the occlusion. Eur J Vasc Endovasc Surg 1996;
extracranial vertebral artery. Br J Radiol 2004; 11:140–147.
77:15–20. 35. Meves SH, Muhs A, Federlein J, et al. Recanalisation
20. Berguer R, Kieffer E. The aortic arch and its of acute symptomatic occlusions of the internal
branches: anatomy and blood flow. In: Kieffer E, carotid artery. J Neurol 2002; 249:188–192.
Berguer R, eds. Surgery of the arteries to the head. 36. Camporese G, Verlato F, Salmistraro G, et al.
New York: Springer Verlag; 1992:5–31. Spontaneous recanalization of internal carotid artery
21. Wain RA, Lyon RT, Veith FJ, et al. Accuracy of duplex occlusion evaluated with color flow imaging and
ultrasound in evaluating carotid anatomy before contrast arteriography. Int Angiol 2003; 22:64–71.
endarterectomy. J Vasc Surg 1998; 27:235–242. 37. Steffen CM, Gray-Weale AC, Byrne KE, et al.
22. Bluth EI, Stavros AT, Marich KW, et al. Carotid Carotid artery atheroma: ultrasound appearances in
duplex sonography: a multicentre recommendation symptomatic and asymptomatic vessels. Aust N Z
for standardized imaging and Doppler criteria. J Surg 1989; 59:529–534.
Radiographics 1988; 8:487–506. 38. Gronholdt ML, Nordestgaard BG, Schroeder TV,
23. Carpenter JP, Lexa FJ, Davies JT. Determination of et al. Ultrasonic lucent plaques predict future
60% or greater carotid artery stenosis by duplex strokes. Circulation 2001; 104:68–73.
Doppler ultrasonography. J Vasc Surg 1995; 39. Golledge J, Cuming R, Ellis M, et al. Carotid plaque
22:697–705. characteristics and presenting symptom. Br J Surg
24. Robinson ML, Sacks D, Perlmutter GS, et al. 1997; 84:1697–1701.
Diagnostic criteria for carotid duplex sonography. 40. O’Donnell TF, Erdoes L, Mackey WC, et al.
AJR Am J Roentgenol 1988; 151:1045–1049. Correlation of B-mode ultrasound imaging and
25. Grant EG, Benson CB, Moneta GL, et al. Carotid arteriography with pathologic findings at carotid
artery stenosis: gray-scale and Doppler US diagnosis endarterectomy. Arch Surg 1985; 120:443–449.
– Society of Radiologists in Ultrasound Consensus 41. Bluth EI, Kay D, Merritt CRB, et al. Sonographic
Conference. Radiology 2003; 229:340–346. characterization of carotid plaque: detection of
26. Hood DB, Mattos MA, Mansour A, et al. Prospective haemorrhage. AJR Am J Roentgenol 1986;
evaluation of new duplex criteria to identify 70% 146:1061–1065.
internal carotid artery stenosis. J Vasc Surg 1996; 42. Ratliff DA, Gallagher PJ, Hames TK, et al.
23:254–262. Characterisation of carotid artery disease:
27. Spencer MO, Reid JM. Quantitation of carotid comparison of duplex scanning with histology.
stenosis with continuous wave (CW) Doppler Ultrasound Med Biol 1985; 11:835–840.
ultrasound. Stroke 1979; 10:326–330. 43. O’Leary DH, Holen J, Ricotta JJ, et al. Carotid
28. Erickson SJ, Mewissen MW, Foley WD, et al. bifurcation disease: prediction of ulceration with
Stenosis of the internal carotid artery: assessment B-mode US. Radiology 1987; 162:523–525.
using colour Doppler imaging compared with 44. Hatsukami TS, Ferguson MS, Beach KW, et al.
angiography. AJR Am J Roentgenol 1989; Carotid plaque morphology and clinical events.
152:1299–1305. Stroke 1997; 28:95–100.
29. Steinke W, Meairs S, Ries S, et al. Sonographic 45. Thiele BL, Jones AM, Hobson RW, et al. Standards
assessment of carotid artery stenosis. Comparison in non-invasive cerebrovascular testing. J Vasc Surg
of power Doppler imaging and color Doppler flow 1992; 15:495–503.
imaging. Stoke 1996; 27:91–94. 46. Rosset E, Albertini J-N, Magnan PE, et al.
30. Sachar R, Yadav JS, Roffi M, et al. Severe bilateral Surgical treatment of extracranial internal carotid
carotid stenosis: the impact of ipsilateral stenting artery aneurysms. J Vasc Surg 2000; 31:713–723.
on Doppler-defined contralateral stenosis. J Am Coll 47. Mann WJ, Beck A, Schreiber J, et al.
Cardiol 2004; 43:1358–1362. Ultrasonography for evaluation of the carotid artery
31. Belkin M, Mackey WC, Pessin MS, et al. Common in head and neck cancers. Laryngoscope 1994;
carotid artery occlusion with patent internal and 104:885–888.
71
3
48. Barry R, Pienaar A, Pienaar C. Duplex Doppler 64. Cinat ME, Casalme C, Wilson SE, et al. Computed
evaluation of suspected lesions at the carotid tomography angiography validates duplex sonographic
bifurcation. Ann Vasc Surg 1993; 7:140–144. evaluation of carotid stenosis. Am Surg 2003;
49. de Bray JM, Lhoste P, Dubaz F, et al. Ultrasonic 69:842–847.
features of extracranial carotid dissections: 47 65. Back MR, Rogers GA, Wilson JS, et al. Magnetic
cases studied by angiography. J Ultrasound Med resonance angiography minimizes the need for
1994; 13:659–664. arteriography after inadequate carotid duplex
50. Logason K, Hardemark HG, Bärlin T, et al. Duplex ultrasound scanning. J Vasc Surg 2003;
scan findings in patients with spontaneous cervical 38:422–430.
artery dissections. Eur J Vasc Endovasc Surg 2002; 66. Herzig R, Burval S, Krupka B, et al. Comparison of
23:295–298. ultrasonography, CT angiography and digital
51. Steinke W, Rautenberg W, Schwartz A, et al. subtraction angiography in severe carotid stenoses.
Non-invasive monitoring of internal carotid artery Eur J Neurol 2004; 11:774–781.
dissection. Stroke 1994; 25:998–1005. 67. Buskens E, Nederkoorn PJ, Buijs-Van Der Woude T,
52. Sitzer M, Fürst G, Siebler M, et al. Usefulness of an et al. Imaging of carotid arteries in symptomatic
intravenous contrast medium in the characterization patients: cost-effectiveness of diagnostic strategies.
of high-grade internal carotid stenosis with colour Radiology 2004; 233:101–112.
Doppler-assisted duplex imaging. Stroke 1994; 68. Moore WS. For severe carotid stenosis found on
25:385–389. ultrasound, further arterial evaluation is unnecessary.
53. Lal BK, Hobson RW 2nd, Goldstein J, et al. Carotid Stroke 2003; 34:1816–1817.
stenting: is there a need to revise ultrasound 69. Rothwell PM. For severe carotid stenosis found on
velocity criteria? J Vasc Surg 2004; 39:58–66. ultrasound, further arterial evaluation is unnecessary:
54. Alexandrov AV, Bladin CF, Magisano, et al. the argument against. Stroke 2003; 34:1817–1819.
Measuring carotid stenosis. Time for a reappraisal. 70. Davis SM, Donnan GA. Is carotid angiography
Stroke 1993; 24:1292–1296. necessary? Editors disagree. Stroke 2003; 34:1819.
55. Rothwell PM, Gibson RJ, Slattery J, et al. 71. Aaslid R, Markwalder TM, Nornes H. Non-invasive
Prognostic value and reproducibility of transcranial Doppler ultrasound recording of flow
measurements of carotid stenosis: a comparison velocity in basal cerebral arteries. J Neurosurg
of three methods on 1001 angiograms. Stroke 1982; 57:769–774.
1994; 25:2440–2444. 72. Nedey C, Barjoud H, Chatelard P, et al. The role of
56. Phillips DJ. Recent advances in carotid artery intraoperative transcranial Doppler monitoring in
evaluation. In: Taylor KJW, Strandness DE, eds. carotid artery surgery. Ann Vasc Surg 1995;
Clinics in diagnostic ultrasound 26: duplex Doppler 9:247–251.
ultrasound. Edinburgh: Churchill Livingstone; 73. Doblar DD. Intraoperative transcranial ultrasonic
1990:25–44. monitoring for cardiac and vascular surgery.
57. Chikos PM, Fisher LD, Hirsch JH, et al. Observer Semin Cardiothorac Vasc Anesth 2004; 8:127–145.
variability in evaluating extracranial carotid stenoses. 74. White DN, Curry GR, Stevenson RJ. The acoustic
Stroke 1983; 14:885–892. characteristics of the skull. Ultrasound Med Biol
58. Executive Committee for the Asymptomatic Carotid 1978; 4:225–252.
Atherosclerosis (ACAS) Study. Endarterectomy for 75. Bauer A, Becker G, Krone A, et al. Transcranial
asymptomatic carotid artery stenosis. J Am Med duplex sonography using ultrasound contrast
Assoc 1995; 273:1421–1428. enhancers. Clinl Radiol 1996; 51(suppl 1):19–23.
59. Cardullo PA, Cutler BS, Brownell Wheeler H. 76. Ringelstein EB, Kahlscheuer E, Niggemeyer E,
Detection of carotid disease by duplex ultrasound. et al. Transcranial Doppler sonography: anatomical
J Diagn Med Sonogr 1986; 2:63–73. landmarks and normal velocity values. Ultrasound
60. Carroll BA. Carotid sonography. Radiology 1991; Med Biol 1990; 16:745–761.
178:303–313. 77. Halsey JH. Effect of emitted power on waveform
61. Sabeti S, Schillinger M, Mlekusch W, et al. intensity in transcranial Doppler. Stroke 1990;
Quantification of internal carotid artery stenosis 21:1573–1578.
with duplex US: comparative analysis of different 78. Sloan MA, Alexandrov AV, Tegeler CH, et al.
flow velocity criteria. Radiology 2004; 232:431–439. Assessment: Transcranial Doppler ultrasonography.
62. Lee DH, Gao FQ, Rankin RN, et al. Duplex and color Report of the Therapeutics and Technology
Doppler flow sonography of occlusion and near Assessment Subcommittee of the American
occlusion of the carotid artery. Am J Neuroradiol Academy of Neurology. Neurology 2004;
1996; 17:1267–1274. 62:1468–1481.
63. Sellar RJ. Imaging blood vessels of the head and
neck. J Neurol Neurosurg Psychiatry 1995;
59:225–237.
72
4
The peripheral arteries
Atheroma occurs to different degrees in different management. In subacute cases ultrasound allows
parts of an individual’s cardiovascular system and any subsequent arteriogram to be scheduled as
the lower limb arteries are particularly prone to either a straightforward mapping examination
the development of atherosclerosis. Approximately prior to bypass grafting, or as a more time-
2% of adults in late middle age in Western countries consuming angioplasty procedure.4,5 In many
have intermittent claudication1 and each year in cases ultrasound will provide sufficient informa-
England and Wales around 50 000 patients are tion for management decisions to be reached. In
admitted to hospital with a diagnosis of peripheral other cases, if further information is required,
arterial disease; 15 000 of these will require the subsequent magnetic resonance angiography
amputation.2 There are many factors which may (MRA)/computed tomography angiography
influence the development of disease and, in (CTA)/arteriogram can be tailored appropriately.
general terms, the prevalence of peripheral vascular Ultrasound provides an accurate assessment of
disease detected by non-invasive procedures is the major arteries which allows distinction between
about three times greater than the prevalence of patients with significant peripheral arterial disease
intermittent claudication.3 This chapter concen- and those without. At the other end of the spec-
trates on the use of ultrasound in the assessment trum, patients with atypical symptoms that might
of disease in the lower limb arteries, as this is the be due to ischaemia can be examined to exclude
area where most work is generated, but the value the presence of significant arterial disease.
of ultrasound in the investigation of a variety of
upper limb arterial disorders is also discussed. Bypass grafts and angioplasty
A variety of problems can occur with surgically
inserted bypass grafts, especially in the first year
INDICATIONS
after the operation. A graft surveillance programme
Peripheral vascular disease using ultrasound allows the identification of grafts
The main indications for performing Doppler at risk of failure and early remedial action to be
ultrasound of the arteries of the upper and lower taken. Similarly, patients in whom angioplasty
limbs are given in Table 4.1. The most common has been undertaken can be followed with ultra-
indication is the assessment of patients with sound to confirm residual patency, identify re-
ischaemic symptoms of the lower limb in order stenosis and assess improvements to flow following
to determine if they are likely to benefit from the procedure.
angioplasty or a bypass graft. The ultrasound
findings provide information on the extent and False aneurysms and other pulsatile
severity of disease; even in patients with limb- masses
threatening ischaemia ultrasound is a useful first- The assessment of pulsatile masses in relation to
line investigation that can provide the surgeon the arteries of the upper and lower limbs can be
with all the information that is required for patient performed rapidly and easily using ultrasound. 73
4
The common femoral artery runs from the
Table 4.1 Indications for Doppler ultrasound of
the peripheral arteries inguinal ligament to its division into superficial
and deep femoral arteries in the upper thigh; this
• Assessment of disease in patients with
ischaemic symptoms of the upper or lower limb
division is usually 3–6 cm distal to the inguinal
• Follow-up of bypass graft procedures
ligament. The deep femoral artery, or profunda
• Follow-up of angioplasty procedures femoris artery, passes posterolaterally to supply
• Diagnosis and follow-up of aneurysms of the the major thigh muscles. The importance of the
peripheral arteries profunda femoris lies in its role as a major
• Diagnosis and treatment of false aneurysms collateral pathway in patients with significant
• Diagnosis of pulsatile lumps superficial femoral artery disease. Several other
• Assessment of dialysis shunts branches arise from the external iliac, common
femoral and profunda femoris arteries and occa-
sionally one of these may be mistaken for the
Aneurysms can be distinguished from non- profunda femoris artery, especially if it is enlarged
vascular masses which lie adjacent to the artery. as a collateral supply.
The complications of catheterisation procedures, The superficial femoral artery passes downwards
including haematomas, arteriovenous fistulae along the anteromedial aspect of the thigh lying
and false or pseudoaneurysms, can be assessed anterior to the vein; in the lower third of the
and differentiated; in many cases, pseudoaneurysms thigh it passes into the adductor canal, deep to
can be treated under ultrasound control, thereby sartorius and the medial component of quadriceps
removing the need for a surgical procedure. femoris. Passing posteriorly behind the lower
femur it enters the popliteal fossa and becomes
Haemodialysis fistulae the popliteal artery, which lies anterior to the
Arteriovenous fistulae created for haemodialysis popliteal vein and gives off several branches, the
can be examined using ultrasound, allowing largest of which are the superior and inferior
identification of complications associated with geniculate arteries. Below the knee joint the
stenosis or occlusion, as well as estimation of popliteal artery divides into the anterior tibial
blood flow through the shunt, particularly if this artery and the tibioperoneal trunk, although the
is thought to be inadequate or excessive. exact level of the division may vary; after 2–4 cm
the latter divides into the posterior tibial artery
and the peroneal artery.
ANATOMY AND SCANNING
The anterior tibial artery passes forwards
TECHNIQUE
through the interosseous membrane between the
Anatomy – lower limb fibula and tibia. It then descends on the anterior
The arteries of the lower limb arise at the bifur- margin of the membrane, deep to the extensor
cation of the abdominal aorta (Fig. 4.1), the common muscles on the anterolateral aspect of the calf
iliac arteries run down the posterior wall of the (Fig. 4.2). At the ankle it passes across the front
pelvis and divide into the internal and external of the joint to become the dorsalis pedis artery
iliac arteries in front of the sacroiliac joint. The of the foot which runs from the front of the ankle
internal iliac artery continues down into the pelvis joint to the proximal end of the first intertarsal
and is difficult to demonstrate with transabdominal space where it gives off metatarsal branches and
ultrasound, although transvaginal or transrectal passes through the first intertarsal space to unite
scanning will show some of its branches. The with the lateral plantar artery and form the
external iliac artery continues around the side of plantar arterial arch.
the pelvis to the level of the inguinal ligament, it The posterior tibial artery passes down the
lies anteromedial to the psoas muscle and is deep medial aspect of the calf to pass behind the
74 normally superficial to the external iliac vein. medial malleolus, after which it divides into the
4
Anatomy and scanning technique

Abdominal
aorta
Right common iliac
Internal iliac
Common illiac
External iliac
Inferior gluteal Inguinal
ligament
Lateral
Common femoral circumflex
femoral artery
Medial
Lateral
circumflex
circumflex
femoral Profunda
femoral artery
Superficial femoris
Profunda femoris femoral
Popliteal
Anterior tibial Anterior tibial
Posterior tibial
Peroneal
Dorsalis pedis Lateral plantar

Arcuate
Plantar arch
Medial plantar
Fig. 4.1 The lower limb arteries.
medial and lateral plantar arteries of the foot; variable in calibre but the anterior tibial and
the lateral plantar artery joins with the dorsalis peroneal arteries may vary considerably in calibre
pedis artery in the plantar arch. and overall length in the calf.The arterial supply
The peroneal artery passes down the calf behind to the foot is not normally examined but if a
the tibia and interosseous membrane and divides bypass procedure to the pedal arteries is being
into several periarticular branches behind the considered then the dorsalis pedis and posterior
ankle joint. The size of the calf arteries can be tibial artery and its plantar branches should be
quite variable, the posterior tibial artery is the least assessed. 75
4
Fig. 4.2 Cross-section of calf,
Anterior tibial showing major relations of calf
artery and veins arteries and the three main access
Transducer points for demonstrating these
Tibia vessels.
Interosseous
membrane
Posterior
tibial artery
Fibula
and vein
Peroneal Long saphenous

artery vein
and veins
Transducer
Transducer Short saphenous vein
Scanning technique – lower limb Doppler trace should be recorded, as changes in

A full scan of the lower limb arteries can be time this may indicate the presence of significant
consuming. In some cases a full scan from aortic disease proximally, necessitating a careful direct
bifurcation to the ankle or foot is required but in examination of the iliac vessels. The bifurcation
other cases the examination can be tailored to of the common femoral artery into the profunda
specific levels depending on the diagnostic femoris and superficial femoral arteries is then
information required. It is therefore useful if the examined using colour and spectral Doppler.
diagnostic question can be clearly defined, so The profunda femoris artery should be examined
that the most appropriate examination can be over its proximal 5 cm, especially in patients with
performed. severe superficial femoral disease, in order to assess
the amount of collateral flow, or its potential
Common femoral, profunda femoris value as a graft origin or insertion point.
and superficial femoral arteries The superficial femoral artery is then followed
The examination begins with the patient lying along the length of the thigh using colour Doppler.
supine on the couch.The main steps in the exami- It is often better to move the transducer in
nation are given in Table 4.2. A linear array
transducer is used; usually 5–12 MHz depending
on the performance of the ultrasound system and Table 4.2 Basic steps in the examination of the
lower limb
the build of the patient, lower frequencies may
be required to examine the arteries in the adductor 1. Patient supine: scan common femoral, proximal
canal or in large patients. The distal external profunda and superficial femoral artery down to
adductor canal
iliac/common femoral artery is located using 2. Patient decubitus: scan adductor canal, popliteal
colour Doppler as it leaves the pelvis under the artery to bifurcation and tibioperoneal trunk, scan
posterior tibial and peroneal arteries
inguinal ligament lateral to the femoral vein.
3. Patient supine: scan anterior tibial arteries. Scan
Even if flow appears normal on colour Doppler iliac arteries and infrarenal aorta
76 and there is no evidence of local disease, a spectral
4

sequential steps, rather than sliding it down the a
thigh, as most machines require a few frames of
sampling at each position to provide a steady
image. In addition, the moving transducer
generates colour Doppler noise over the image,
obscuring vascular details. Doppler spectra are
obtained as necessary at points of possible disease.
Even in the absence of colour Doppler abnor-
malities, it is good practice to obtain routine
spectral assessments from the superficial femoral
artery in the upper, middle and lower thigh, in
order to confirm that there is no alteration in the
waveform that might suggest disease. Sometimes
the artery is difficult to see on colour or power
b
Doppler as the signals are weak or absent. In
these cases the artery may be visible by virtue of
calcified plaques in the wall of the vessel.
Alternatively, the superficial femoral vein, lying
behind the artery, can be used as a guide to the
position of the artery and spectral Doppler used
to demonstrate the presence or absence of arterial
flow. Echo-enhancing agents can be used if there
is any continuing uncertainty concerning the
patency of the artery.
There are three indirect signs of significant
disease which might be apparent during the
examination and which should prompt a careful
review if a cause for these changes has not been Fig. 4.3 (a) Collateral vessels in the muscles of the
identified. thigh. (b) A larger collateral vessel joining the lower
1. Colour Doppler may show the presence of superficial femoral artery.
collateral vessels in the muscles of the thigh
(Fig. 4.3a).
2. Collateral vessels may be seen leaving the of the adductor canal must be examined with
main artery (Fig. 4.3b). great care as it is a site where a short-segment
3. The character of the spectral waveform may stenosis or occlusion may be present, and this
show a change between two levels, indicating section of the vessel can be difficult to visualise
a segment of disease somewhere between the as it passes deep to the thigh muscles. In some
two points of measurement. cases the use of a lower scanning frequency may
help in visualisation.The superficial femoral artery
The adductor canal and popliteal fossa is examined as far down as it can be followed on
The patient is then turned into a lateral decu- the medial aspect of the thigh; the popliteal
bitus position so that the medial aspect of the leg artery is then located in the popliteal fossa and
being examined is uppermost (Fig. 4.4). This followed superiorly. In difficult cases a mark can
position is better than the prone position as it be put on the skin of the medial thigh to show the
allows access in continuity to the lower superficial lowest segment of vessel visualised in the supine
femoral artery, the adductor canal area, the position; the popliteal artery is then followed
popliteal region and the medial calf. The region superiorly in the decubitus position until the trans- 77
4
The posterior tibial artery is usually the easier
of the two branches of the tibioperoneal trunk to
locate. Often it can be located by placing the
transducer in a longitudinal position on the medial
aspect of the mid-calf area behind the tibia, using
colour or power Doppler to show the course of
the vessel, which can then be followed up and
down the calf. In obese or oedematous legs, or if
blood flow is impaired by disease, the posterior
tibial artery and the other calf arteries may be
difficult to locate. Scanning with colour Doppler
in the transverse plane using some angulation
towards the head or feet may show the relative
positions of the posterior tibial and peroneal arteries.
Alternatively, the associated veins can be used to
identify the region of the relevant artery:
squeezing the foot or lower calf will augment
flow in the deep veins, allowing these to be iden-
tified in either a longitudinal or transverse scan
plane. The posterior tibial artery can also be
located as it passes behind the medial malleolus,
where its position is constant, and then followed
Fig. 4.4 The position of the patient for examination
back up the leg.
of the adductor canal and popliteal fossa.
The peroneal artery runs more deeply down
the calf than the posterior tibial artery, lying
ducer reaches the level of the skin mark, ensuring closer to the posterior aspect of the tibia and the
that the vessel has been examined in continuity. interosseous membrane. It can be examined from
The popliteal artery is then examined and several approaches: firstly from a posteromedial
followed down to the point of division into the approach similar to that used for the posterior
tibioperoneal trunk and anterior tibial artery. tibial artery; secondly, it can often also be seen from
the anterolateral approach used for the anterior
Calf arteries tibial artery as it runs behind the interosseous
The complexity of the assessment of the calf membrane (Fig. 4.2); thirdly, a posterolateral
arteries depends on the clinical situation. If the approach may be of value in some cases.
examination is to exclude significant proximal The anterior tibial artery is examined from an
disease that would benefit from angioplasty or anterolateral approach through the extensor
bypass grafts, then it is usually sufficient to assess muscles lying between the tibia and fibula. The
the three calf arteries at the upper and mid-calf two bones can be identified on transverse scanning
level, recording whether they are patent or not, and the interosseous membrane located passing
in order to provide some assessment of the state between them. The anterior tibial artery lies on
of the distal run-off. In other cases a more detailed the membrane and can be located using colour
examination is required to clarify changes seen Doppler in either the longitudinal or transverse
on MRA/arteriography, or if a distal insertion plane. It usually lies nearer the fibula than the
point for a bypass graft is being sought. The tibia (Fig. 4.2).
increased sensitivity of power Doppler is useful The foot vessels are not usually examined but
in detecting weak signals from small or diseased the dorsalis pedis artery may be examined in front
78 but patent vessels. of the ankle joint before it passes deep to the
4

metatarsals; this is indicated if the artery is being visualisation of the internal iliac artery origin, on
considered for the insertion of a femorodistal tracking both the external iliac artery upwards
graft. and the common iliac artery downwards, means
The advent of power Doppler and echo- that the iliac vessels have been examined in their
enhancing agents has extended the role of entirety.
ultrasound in the assessment of vascular disease. The orientation of the iliac vessels as they pass
In the proximal lower limb and iliac vessels, the round the pelvis and the use of sector or curved-
location of the vessel and confirmation of occluded array transducers can lead to problems with
segments has been made easier and, in the distal beam–vessel geometry and obtaining satisfactory
part of the leg, they make assessment of the smaller angles of insonation. However, careful attention
vessels of the calf and foot easier. However, more to positioning the transducer will usually allow
work is required to evaluate further their role.6 an appropriate angle to be obtained.
Iliac arteries Anatomy – upper limb

Examination of the iliac vessels is carried out as The subclavian arteries arise from the brachio-
part of a general survey of the lower limb arteries, cephalic trunk on the right and directly from the
or if the clinical picture suggests a need to confirm arch of the aorta on the left (Fig. 4.5); however,
or exclude disease affecting these vessels, or if there is considerable normal variation in the
the Doppler findings at the groin suggest the patterns of their origination. The origin of the
likelihood of significant proximal disease. The
ease with which they can be visualised depends
on the build of the patient and the amount of
bowel gas present. It is usually necessary to use Subclavian artery
a 3–5 MHz transducer for satisfactory visuali- Vertebral artery Right common
carotid artery
sation. Some examiners will prepare patients for Thyrocervical trunk
iliac Doppler examinations with laxatives and Lateral thoracic artery Brachiocephalic
trunk
low-residue diets if it is considered likely that Thoraco-acromial artery
these vessels will be examined, although most Axillary artery
centres do not do this routinely. However, it is Subscapular
important not to make a false diagnosis of occlusion artery
because the vessel is obscured by bowel gas.
Internal thoracic
The external iliac artery can be followed up Brachial (mammary) artery
from the groin for a variable distance; the vein, artery
Profunda brachii artery
lying behind the artery, can be used to identify
Radial
the probable location of the artery if this is not Superior and inferior
collateral
apparent. Colour or power Doppler may also help ulnar collateral arteries
artery
locate the vessel, even if it is not visible on the
real-time scan image. Superiorly the common iliac Interosseous artery
artery can be identified arising from the aortic
bifurcation and then followed distally. Firm
pressure with the transducer will displace inter- Radial
Ulnar artery
vening amounts of bowel gas to a large extent, artery
although care must be taken not to compress the
artery and produce a false impression of a stenosis. Superficial palmar arch
Deep palmar arch
The internal iliac artery may be seen arising from
the common iliac artery and passing deeply into
the pelvis. This is a useful landmark as Fig. 4.5 The upper limb arteries on the right. 79
4
right subclavian artery can be examined behind aspect of the wrist, where it runs over the flexor
the right sternoclavicular joint, where the brachio- retinaculum and then divides into superficial
cephalic trunk divides into the right common and deep branches which anastomose with the
carotid artery and the subclavian artery.The origin equivalent branches of the radial artery to form
of the left subclavian artery from the aortic arch the superficial and deep palmar arches. It gives
cannot be demonstrated, although the more off recurrent ulnar arteries below the elbow and
distal segments can be seen as on the right side. the common interosseous artery, which forms
The subclavian artery on each side runs from its anterior and posterior divisions, running on either
point of origin to the outer border of the first rib side of the interosseous membrane.
where it becomes the axillary artery; the subcla-
vian vein lies in front of the artery. The main Scanning technique – upper limb
branches of the subclavian artery are the vertebral A 7–12 MHz transducer may be used in most
arteries, the thyrocervical trunk, the internal thoracic cases for examining the arteries in the upper limb,
(mammary) artery and the costocervical trunk. as there is less tissue to penetrate than in the leg.
The axillary artery runs from the lateral border The patient lies in a supine position with their
of the first rib to the outer, inferior margin of the head turned away from the side being examined,
pectoralis major muscle. It gives rise to several the arm is abducted with the elbow flexed and
branches which supply the muscles around the the back of the hand resting on the pillow by the
shoulder; the largest of these are the thoraco- patient’s head, so that the axilla and medial aspect
acromial trunk, the lateral thoracic artery and of the upper arm can be examined. Alternatively,
the subscapular artery. the arm can be abducted and supported on a
The brachial artery passes down the medial suitable shelf, or by asking the patient to hold
aspect of the upper arm to the cubital fossa below onto a suitable part of the ultrasound machine.
which it divides into the radial and ulnar arteries; The distal brachiocephalic artery on the right
the point of division sometimes lies higher in the and the mid-subclavian artery on the left can be
upper arm.Apart from muscular branches, the main visualised from a supraclavicular approach by
branches of the brachial artery are the profunda angling the transducer down towards the medi-
brachii artery, which is given off in the upper astinum from above the medial clavicle and
arm and passes behind the humerus; the superior sternoclavicular joint.The subclavian arteries on
and inferior ulnar collateral arteries arise from the both sides are seen behind the subclavian vein as
lower part of the brachial artery. It divides into the they run up to cross the first rib. In patients with
ulnar and radial arteries in the antecubital fossa. possible arterial compression syndromes, the artery
The radial artery runs down the radial (lateral) is examined with the arm in various positions so that
aspect of the forearm to the wrist, where it passes any narrowing or occlusion may be demonstrated.
over the radial styloid process, and over the The artery beyond the first rib is best examined
lateral aspect of the carpus. It then passes down from below the clavicle and followed distally. It
through the first interosseous space to form the can be followed in continuity as it runs deep to
lateral aspect of the deep palmar arch. It also has the pectoralis muscles through the axilla and
a superficial branch which anastomoses with the into the medial aspect of the upper arm; from
equivalent branch of the ulnar artery to form the here it can be tracked down to the cubital fossa.
superficial palmar arch. In the upper forearm The division of the artery into the two main forearm
the radial artery gives off the radial recurrent branches usually occurs below the cubital fossa
artery, which anastomoses with the profunda and each branch can be followed to the wrist; if
brachii artery, and several muscular branches in either the radial or the ulnar artery is difficult to
the forearm and at the wrist. trace down from the elbow, then the vessel should
The ulnar artery passes down the anterior ulnar be sought at the wrist and followed back up the
80 (medial) aspect of the forearm to the medial forearm.
4
Assessment of disease
the effects of these are additive, so that there is
ASSESSMENT OF DISEASE
still a significant drop in perfusion pressure distal
The assessment of lower limb atheroma is more to the affected segment.7
complex than for the carotids as the potential for In addition, the presence of serial stenoses
collateral supply around stenoses or occlusions can affect the estimation of the degree of a distal
is very much greater. The distinction must be stenosis, if this is not recognised. A significant
made between haemodynamically significant proximal stenosis or occlusion will result in a
disease and clinically significant disease. drop in perfusion pressure and velocity which
Two individuals may have the same degree of makes application of the peak systolic velocity
stenosis in their superficial femoral artery, but (PSV) and peak velocity ratios, the usual criteria
an 80% diameter reduction which develops acutely for quantifying a stenosis, problematical.8
will be significantly symptomatic, whereas the Power Doppler and echo-enhancing agents
same degree of stenosis developing over a period may allow an estimate of severity but if there
of time, allowing collateral channels to open, is clinical doubt, other imaging should be
may be much less disabling. The findings on considered.
Doppler must therefore not be considered in The same principles apply to the assessment
isolation but in the light of the full clinical of disease in the upper limb, but the type and
picture. distribution of disease in the arm is different
Colour Doppler allows the rapid identification from that seen in the leg. Ischaemic symptoms
of normal and abnormal segments of vessel. In in the arm may be the result of compression,
addition, some stenoses may show a colour embolic occlusion, or vasospasm and are less
Doppler tissue ‘bruit’ due to the tissue vibrations frequently due to localised atheroma.
set up by the blood passing through the stenosis. The main diagnostic criteria which are of value
It is valuable to relate the level of any diseased in the assessment of lower limb atheroma are
segments demonstrated on ultrasound to bony direct measurement of the stenosis, PSV ratios
landmarks which can be seen on angiography; and waveform changes.
this allows the appearances on the two exami-
nations to be compared and confirm that the Direct measurement
abnormality on the arteriogram is at the level of Direct measurement of a stenosis is often quite
the lesion seen on ultrasound or vice versa. The difficult in the lower limb arteries as these are
groin skin crease corresponds to the superior pubic relatively small, and it may be difficult to see the
ramus and can be used for lesions in the upper lumen clearly in the deeper parts of the thigh,
part of the thigh; the upper border of the patella particularly if there is disease present. However,
can be used for lesions in the lower part of the direct measurement of a stenosis may be possible
thigh; and the tibiofemoral joint space for popliteal in the lower external iliac, common femoral,
lesions. profunda femoris and upper superficial femoral
Some patients will show extensive diffuse disease arteries. Measurement of the diameter reduction
along much, or all, of the superficial femoral is performed after assessment of plaque distri-
artery but do not show any specific, localised bution in both longitudinal and transverse planes,
stenoses. It is important to note this appearance, so that the most appropriate diameter is selected.
as the overall haemodynamic effect may be severe When a segment of stenosis or occlusion is detected
enough to produce a significant pressure drop the length of the affected segment should be
along the vessel, thereby reducing limb perfusion, measured, as this will be relevant to the suitability
although this pattern of disease is not suitable of the lesion for angioplasty; segments of disease,
for angioplasty. Other patients may have several particularly occlusions, longer than 10 cm will
stenoses along the length of the vessel, each of not normally be considered for percutaneous
which is not haemodynamically significant but treatment.9 81
4
Peak velocity ratios
which occur in the lower limb arteries during the
Direct measurement of a stenosis is often not cardiac cycle. First there is the rise in pressure and
possible in the lower limb and the severity of the acceleration of blood flow at the onset of systole.
stenosis must then be estimated from the change There is then a short period of reversed flow as
in peak systolic velocity produced by the stenosis. the pressure wave is reflected from the constricted
Normal velocities in the lower limb arteries at distal arterioles.This is followed by a further period
rest are approximately 1.2 m s-1 in the iliac of forward flow produced by the elastic compli-
segments, 0.9 m s-1 in the superficial femoral ance of the main arteries in diastole (Fig. 4.7a).
segments and 0.7m s-1 in the popliteal segment.10 These changes are discussed in more detail in
Various criteria have been put forward for the Chapter 2.
quantification of lower limb arterial stenosis. Exercise modifies this pattern by reducing the
Those of Cossman et al11 have produced satis- peripheral resistance.This results in the reversed
factory results in the authors’ department and component being lost and increased diastolic
have the advantage of being easy to remember flow throughout the cardiac cycle (Fig. 4.7b). It
(Table 4.3). These criteria are based on the is for this reason that ultrasound examination of
PSV at the stenosis and the ratio of the PSV at the lower limb arteries should be performed in
the stenosis compared with the velocity 1–2 cm patients who have not had significant exercise of
upstream in a non-diseased segment. Colour the leg muscles for about 15 min. Conversely,
Doppler allows the position and direction of examination after exercise, or reactive hyperaemia,
peak velocity flow in the stenosis to be identified may be used in order to ‘stress’ the lower limb
and the sample volume can be placed appro- circulation and reveal stenoses which are not
priately, final adjustments of position being significant at rest, when blood flow is relatively
performed by listening to the pitch of the frequency low, but which become apparent with the higher
shift as the sample volume is moved through the volumes flowing when the distal circulation to
stenosis. A further velocity measurement is then the muscles opens up.12
made in a ‘normal’ segment of artery 1–2 cm Disease in the vessel at the point of measure-
upstream from the stenosis and the ratio calcu- ment, above it or below it, can affect the wave-
lated (Fig. 4.6). form, and if the vessel cannot be visualised in
continuity then a change in the waveform between
Waveform changes two points is indicative of disease. The two main
The normal waveform in the main arteries of the features which may be altered are the overall
resting lower limb has three components; four, shape of the waveform and the degree of spectral
or occasionally five, may be seen in fit young broadening as a result of flow disturbance;10 the
individuals.These represent the pressure changes major changes are shown in Table 4.4 and illus-
Table 4.3 Velocity criteria for the assessment of lower limb stenoses
Percentage stenosis Peak systolic velocity Velocity ratio

(m s-1)
Normal <1.5 <1.5:1
0–49 1.5–2.0 1.5–2:1
50–75 2.0–4.0 2–4:1
>75 >4.0 >4:1
Occlusion – –
From Cossman et al.11
82
a
a
4
b
b
Fig. 4.6 Measurement of the peak systolic velocity Fig. 4.7 The normal femoral artery waveform. (a) In a
ratio at a stenosis. (a) Velocity at the stenosis is limb at rest three components are visible. Aliasing in
3.4 m s-1. (b) Velocity above the stenosis is 0.66 m s-1, the profunda artery is due to the lower insonation
giving a ratio of approximately 5. angle giving a higher Doppler shift, not a focal
stenosis. (b) Following exercise there is increased
flow throughout diastole as a result of peripheral
dilation.
trated in Fig. 4.8. Proximal disease above the point
of measurement results in loss firstly of the third
and then of the second components of the wave-
form, as the normal passage of the pressure wave
along the artery is impaired.This interference with
Table 4.4 Waveform changes associated with
the passage of the pulse wave along the vessel is disease in the lower limb
also manifest by the slowing of the systolic accelera-
tion time.The width of the first, systolic complex • Loss of third and then second phase of the
waveform
is increased and the overall height is decreased.
• Increased acceleration time
These changes result in ‘damping’ of the • Widening of the systolic complex
waveform, which is most marked when there is a • Damping of the waveform
proximal occlusion. The turbulence generated • Spectral broadening
beyond a stenosis shows in the spectrum as spectral • Absent flow in occlusion
broadening and its presence therefore also indi- 83
4
a cates proximal disease, although not its severity,
as mild turbulence may be due to a minor stenosis
close by, or a more severe stenosis further away.
The spectral broadening may be seen throughout
the spectrum if the stenosis is close to the point
of measurement; as the distance from the stenosis
increases, the spectral broadening is seen in the
postsystolic deceleration phase only. The distur-
bance of flow created by a stenosis may take several
centimetres to resolve. The spectral broadening
is lost and the systolic forward flow component
is regained, but the reverse component and third
component are much less likely to reappear distal
to disease. In addition, if the distal limb is
ischaemic, this will result in dilatation of the
capillaries and increased flow throughout diastole.
b The presence of some, or all, of these changes
requires that the vessel be carefully examined
proximally to identify their source.
The presence of distal disease will also affect
the waveform, resulting in increased pulsatility,
with reduced diastolic flow, evident from the loss
of the third component; in addition, the peak
systolic velocity is reduced.This situation is most
often seen at the origin of a superficial femoral
artery with significant distal disease, although
the precise changes are variable, depending on
the degree of obstruction and the capacity of any
collateral channels.13
c
Assessment of aortoiliac disease
The clinical findings, or the appearances of the
waveform at the groin, may suggest the presence
of significant disease in the aortoiliac segments.14
The best method for assessment of these segments
is by direct visualisation with colour Doppler and
measurement of velocities as for the leg vessels.
Satisfactory examinations have been reported in
up to 90% of cases with careful scanning and
preparation.15 However, even if adequate direct
visualisation is not achieved, the likelihood of
Fig. 4.8 Abnormal lower limb artery waveforms. (a) significant proximal disease should be noted and,
Loss of reverse flow and diastolic flow due to distal depending on the clinical severity, it can be assessed
disease producing peripheral dilatation; a large plaque further using MRA, or by arteriography, if neces-
impinges on the lumen. (b) Broadening of the
sary. The main indirect indicators of significant
waveform and turbulence secondary to a proximal
iliac artery disease are spectral broadening, due
stenosis. (c) Damped waveform in the popliteal artery
84 to the turbulence set up by the stenosis, and
secondary to a proximal occluded segment.
4
Arterial bypass grafts

widening of the systolic complex as a result of
Table 4.5 Types of bypass graft
the stenosis slowing the systolic acceleration. As
noted earlier, power Doppler and echo-enhancing Types of graft
agents show promise in improving visualisation Femoropopliteal – may arise from external iliac
artery rather than common femoral artery
and assessment using ultrasound.6
Femorodistal – to calf vessels or dorsalis pedis
Femorofemoral crossover
Reporting the examination
Iliofemoral
The findings from a Doppler examination of the
Axillofemoral
peripheral arteries are best reported on a diagram
Materials used for graft
of the relevant arterial tree. The level and extent
Vein
of any stenoses can be indicated on this, together
In situ vein
with blood flow velocities at areas of stenosis and
Reversed vein
the standard sites of measurement. Non-visualised
Synthetic
areas should also be indicated. For patients with
Polytetrafluoroethylene (PTFE)
bypass grafts, the location of the graft can be
Gore-Tex
sketched in and relevant velocities indicated.
ARTERIAL BYPASS GRAFTS

The timing of graft failure is related to the
A variety of bypass grafts may be employed to alle- cause of the problem. Failures occurring within
viate ischaemic symptoms (Table 4.5). Autologous 6–8 weeks of surgery are usually due to technical
vein is the preferred material and usually the long problems arising from the surgery; 3–5% of grafts
saphenous vein is used, although other veins may fail at this stage, approximately 25% of all graft
occasionally be employed. Synthetic materials failures. Failures developing in the period begin-
[polytetrafluoroethylene (PTFE) or Dacron] may ning 3 months and extending to 2 years after
be used if the long saphenous veins are unsuit- surgery are usually due to neointimal hyperplasia;
able or unavailable. Infrainguinal femorodistal 12–37% of grafts fail during this period, approxi-
grafts are the commonest type of procedure; mately 70–80% of all graft failures; most of these
several problems may occur which result in graft will occur in the first 12 months. Beyond 2 years
failure (Table 4.6) and it has been shown that a after surgery the usual cause of failure is progres-
programme of graft surveillance in the postopera- sion of atherosclerosis, either in the native
tive period can reduce the number of failed grafts.16 vessels or in the graft itself.17 In addition to graft
Table 4.6 Causes of graft failure
Intrinsic Extrinsic
Stenosis Inflow disease progression
Proximal or distal anastomosis Outflow disease progression
Mid-graft Clamp injury
Diffuse myointimal hyperplasia Thromboembolism

Hypercoagulation states
Aneurysm
Sepsis
Anastomotic
Mid-graft
Haemodynamic failure occurs when the graft is patent but the limb
remains ischaemic
85
4
stenosis or occlusions other problems may occur, femoris artery. The native artery above the graft
including dehiscence at the origin or insertion and is assessed and the velocity of blood flow measured
false aneurysm formation, arteriovenous fistulae, with spectral Doppler. The origin of the graft is
infected collections and compression or kinking. then examined carefully and any increase in velocity,
The timing of the surveillance scans is based or disturbance of flow, noted (Fig. 4.9).
on this time scale for problems. An early scan The graft is followed down the length of the
4–6 weeks after the operation is performed, subse- thigh. Most grafts run along the medial aspect of
quently a scan is done at 3 months and then at the thigh; however some grafts, particularly repeat
3-monthly intervals until the end of the first year. grafts, may follow unusual courses, even crossing
It is not usually necessary to continue beyond this the thigh to run down the lateral aspect of the
if no cause for concern exists, as the majority of leg. Velocity measurements are obtained at any
failures will occur in the first year. However, if point of disturbed flow; if no disturbance is
there are particular reasons for concern, such as present, two to three velocity measurements are
mild or moderate stenosis (velocity ratio <2), then taken along the length of the graft to ensure that
surveillance can be extended as appropriate. Sur- there is satisfactory flow.
veillance programmes have been shown to be bene- At the graft insertion, the flow above, at and
ficial for in terms of enhanced graft patency and below the anastomosis is examined, any signifi-
limb survival rates.16,18 One study followed grafts cant changes in the velocities should be assessed
in 259 legs for up to 10 years and demonstrated as possible signs of stenosis. Some care must be
87.4% patency and 88.7% limb salvage rates taken in the interpretation of velocity increases at
5 years after the original operation; 10 years after both the origin and insertion of grafts, as moderate
the operation the equivalent figures were 80.4% changes may be the result of disparity in size
patency and 75.4% limb salvage.19 The situation between the graft and the native vessel and
is less clear cut for synthetic grafts. Whilst some therefore not pathological in origin, particularly
authors suggest that surveillance of synthetic grafts at the distal insertion, if this is into a relatively
is beneficial,20 other authors have found that syn- small calf artery or the dorsalis pedis artery.
thetic grafts are more likely to fail without warn- Synthetic grafts are relatively straightforward
ing signs being demonstrated on Doppler.21, 22 to assess, as any problems usually occur at the
Symptomatic grafts should always be examined, origin or insertion, rather than along the length
as a treatable lesion may be demonstrated prior of the graft. Vein grafts, however, can develop
to complete graft thrombosis and failure. problems at any point along their length, particu-
Technique of examination
It is of value if the request for a graft assessment
gives details of the surgery and the type of graft
inserted; ideally, a diagram of the course of the
graft should be provided.The examination should
begin at the groin and the graft located; trans-
verse scanning is helpful with identification of
the graft origin. Once located the graft should be
followed up to its point of origin from the native
artery. The majority of grafts are femoropopliteal
and run from the common femoral artery, or
lower external iliac artery, to the upper or lower
popliteal artery; occasionally the graft may origi-
nate from deeper in the pelvis, lower down the Fig. 4.9 Origin of a femoropopliteal bypass graft with
86 superficial femoral artery, or from the profunda a high velocity of 3.27 m s-1.
a
4
Arterial bypass grafts

larly at sites of avulsed valves or tied perforating
veins. Stenoses can occur secondary to surgery,
or as a result of intimal hyperplasia, which can
be stimulated by the turbulence at an irregularity
in the vein wall. In situ vein grafts may also have
persistent arteriovenous communications if a
perforating or superficial communicating vein
has been overlooked during the operation.These
may be quite small but their presence should be
suspected if there is a rapid, unexplained drop in
velocity along the graft, or if pulsatile venous
flow is seen in the common or superficial femoral
veins. Scanning transversely along the line of the
in situ graft makes it easier to identify these commu-
nicating vessels and demonstrate their course. b
A note should also be made of any collections

seen along the track of the graft. In the post-
operative period these are usually small collections
of serous fluid, haematomas, or small lymphoceles;
normally these resorb over a few weeks (Fig. 4.10a).
If infection is suspected (Fig. 4.10b) then a fine
needle (20–22G) can be used to perform a diag-
nostic aspiration, although care should be taken not
to introduce infection into a sterile collection.
If an occluded graft is demonstrated but the
leg is asymptomatic then it is worth checking if
the patient has had more than one bypass proce-
dure, as a second, patent graft may be present
Fig. 4.10 (a) A small trans-sonic lymphocoele around
elsewhere in the leg.
a femoropopliteal bypass graft; (b) an infected fluid
Other bypass procedures that may be performed collection around a synthetic graft showing low level
include aortoexternal iliac/femoral grafts, axillo- echoes within the fluid.
femoral grafts, femorofemoral (crossover) grafts.
These use synthetic grafts and do not usually enter clinical concerns. Distal lower limb grafts to the
into graft surveillance programmes, although an lower calf arteries, or dorsalis pedis artery are
ultrasound scan may be requested if there are also performed.
Table 4.7 Doppler criteria for graft stenosis and grafts at risk
Direct measurement of stenosis Moderate >50% diameter stenosis

Severe >70% diameter stenosis
Peak systolic velocity changes >1.5 m s-1 = 50–70% diameter
stenosis
>2.5 m s-1 >70% diameter stenosis
End diastolic velocity >1.0 m s-1 >70% diameter stenosis
Velocity ratio >2.5 >50% diameter stenosis
>3.5 >70% diameter stenosis
Peak systolic velocity <50 cm s-1 in narrowest segment
87
4
Features of a graft at risk
stenosis of more than 70%, or low mean graft
The main features which suggest that a graft is velocity, is a strong indicator of a graft at risk.16
at risk of failing are shown in Table 4.7. There is Gibson et al23 found that there was a significantly
good correlation between these criteria and the higher incidence of graft failure if there was a
incidence of subsequent graft failure.The finding graft related stenosis with a velocity ratio >3.5
of a fall in the ankle/brachial pressure index of and a mean graft PSV of <50 cm s-1, calculated
more than 0.15 in addition to the presence of a from velocities obtained at non-stenotic points
along the length of the graft. However, they found
a that a low mean graft velocity by itself without
an associated stenosis was not associated with
graft failure.
Other abnormalities which may be seen in rela-
tion to a bypass graft are a false aneurysm at the
origin or insertion due to dehiscence of the anasto-
mosis (Fig. 4.11), or an arteriovenous fistula.
DIALYSIS SHUNTS
Various types of arteriovenous communication
may be fashioned to allow for haemodialysis; these
are usually in the arm but if no suitable vessels are
available the leg may be used (Fig. 4.12). Colour
Doppler can be used to examine the supplying
artery, the anastomosis, or any interposed venous
or synthetic grafts, and the draining veins.24
Dialysis shunt function is normally monitored
by assessing performance during dialysis. The
main indications for examining a dialysis fistula
are when a shunt which has previously been
functioning satisfactorily starts to malfunction,
b
or when a recently created shunt does not seem
to be maturing adequately.
In grafts which are not functioning well it
should be remembered that problems can occur
with the artery anywhere along its length, at the
anastomosis, in the vein distal to the anastomosis
in the region where dialysis needles are inserted,
or in the veins proximally around the groin or
clavicle.25
Technique
The examination of upper limb arteriovenous
dialysis fistulae begins with the axillary or upper
brachial artery, which is then followed with colour
Fig. 4.11 (a) Dehiscence at the insertion of a graft Doppler distally until the anastomosis with the
into the lower popliteal artery; (b) aneurysmal vein or graft is identified.The region of the anas-
88 dilatation at the origin of a by pass graft. tomosis is then examined carefully for any evidence
Fig. 4.12 Diagrams of the different

4
Dialysis shunts
types of arteriovenous fistulae used
for dialysis. (a) Radiocephalic
(a) anastomosis; (b) brachiocephalic
anastomosis;
Cephalic vein
Radial artery
(b)
Cephalic vein
Radial artery
of stenosis. The amount of flow through such a Alternatively, gentle compression of the brachial
fistula is normally sufficiently high and turbulent artery with a pressure cuff or manual pressure
to produce tissue vibrations, which may obscure may reduce flow sufficiently for the tissue vibra-
the lumen but an increase in velocity of two to tions to be reduced and allow visualisation of the
three times the brachial arterial velocity does not vessel and anastomosis.
mean that there is a haemodynamically signifi- The venous drainage is then followed back up
cant stenosis. If the tissue bruit is a problem, the arm to the subclavian vein; normally there is
making sure that transducer pressure is minimal, a dominant draining vein but several veins may
and therefore not producing any venous compres- contribute to the drainage of the fistula and each
sion, may reduce this visible tissue vibration. must be followed proximally. If a stenosis is 89
4
Fig. 4.12 cont’d (c) two types of
fistulae using a synthetic PTFE
(c) graft.
Brachial
artery
Cephalic vein
Radial
artery
PTFE Cephalic
graft vein
suspected on the venous side, then the veins should strated on arteriography. Ultrasound depicted 92%
be examined carefully, making sure that trans- of significant stenoses, whereas graft volume flow
ducer pressure is as light as possible to ensure and the resistive index did not correlate with the
that inadvertent compression is not responsible presence of stenosis. Other workers have suggested
for any narrowing which may be demonstrated. a velocity >4 m s-1 as a marker of a significant
There may be a physiological increase in venous stenosis at an arteriovenous anastomosis.24 It
velocity as the vein passes through clavipectoral should also be recognised that a significant stenosis
fascia and careful assessment of this region should in the artery or at the anastomosis may result in
be carried out if venous problems are suspected. a fall in the velocity and volume of blood flowing
along the venous side of the fistula.
Stenosis Apart from stenosis affecting the inflow, outflow,
A stenotic lesion may affect a dialysis fistula on or anastomosis, other problems which may be
the arterial side, at the anastomosis, or on the identified in relation to dialysis fistulae are true
venous side. The criteria used for lower limb or false aneurysms at the sites of needle insertion,
arterial lesions do not apply in these circum- haematomas and abscesses; these result from the
stances, as flow velocities are generally high with repeated trauma of cannulation. Segmental throm-
inherent turbulence in the presence of a fistula bosis may occur on the venous side and complete
and visible tissue ‘bruits’ can occur in the absence thrombosis of the fistula may occur, usually
of a physiologically significant stenosis. Flow at affecting the venous rather than the arterial side.
the arteriovenous anastomosis may be very
turbulent and a doubling in velocity at this point Fistula steal syndrome
does not necessarily mean that there is a haemo- The low resistance of the fistula means that flow
dynamically significant stenosis present. However, to the distal parts of the limb may be impaired if
a sudden increase in velocity in the venous limb too much blood is diverted through the arte-
of the fistula should be assessed carefully, as this riovenous anastomosis. This is not normally a
may well represent a significant stenosis; Robbin problem but coexistent arterial disease in the limb
et al26 showed that a focal increase in PSV ratio arteries above or below the anastomosis can result
90 of 3 was associated with a 75% stenosis, as demon- in significantly impaired perfusion pressure
4
False aneurysms
distally. In one study27 10 of 212 patients (8.3%) venous channels, together with the turbulence of
had signs and symptoms of steal syndrome; this flow, can make assessment of fistula flow volume
was attributed to stenosis of the inflow arteries difficult to estimate on the venous side.24 On the
in five cases, excessive fistula flow in two cases, arterial side it is reasonable to assume that the
arterial disease distal to the anastomosis in two greatest proportion of blood in the supplying
cases, and no attributable cause was found in the brachial or radial artery will go through the
remaining patient. Steal syndromes should be fistula when the arm is in the resting state, and
assessed by examining the limb arteries above and therefore to estimate the blood flow volume in
below the anastomosis to identify any significant the fistula from the supplying artery. If necessary
stenoses. In patients with excessive flow in a radial the distal arterial supply can be occluded with a
artery fistula, antegrade flow may be seen in the pressure cuff.
ulnar artery but retrograde flow is seen in the Techniques for measuring volume flow have
distal radial artery as this is reversed with collateral been discussed in Chapter 1. The most straight-
flow through the palmar arches feeding the fistula.24 forward method for estimating volume flow is by
Compressing the venous side of the fistula will multiplying the time-averaged mean velocity in
reduce any excessive fistula flow and if this results the vessel by the cross-sectional area at the point
in a significant increase in distal flow then it is of measurement.Three aspects of technique should
likely that flow reduction surgery to the fistula be remembered: the time-averaged mean velocity
will be beneficial. Significant inflow stenoses may must be measured, not the time-averaged maxi-
benefit from angioplasty. mum velocity; the sample volume for the velocity
measurement should encompass the complete
Volume flow in the fistula cross-section of the vessel; and the cross-sectional
The assessment of volume blood flow using area should be measured at right angles to the
Doppler ultrasound is beset by various inherent long axis of the vessel in order to obtain the best
problems which result in a wide standard devia- estimate of volume flow. If the system calculates
tion for most flow measurements, as discussed the cross-sectional area from the diameter of the
in Chapter 1. However it is possible to measure artery, then this must also be measured at right
dialysis fistula flow sufficiently accurately to divide angles to the long axis of the artery (Fig. 4.13).
it into three broad groups: <200 mL min-1, which
is inadequate for dialysis; 200–800 mL min-1,
FALSE ANEURYSMS
which is satisfactory; and >800 mL min-1, which
is excessive and may be associated with high cardiac These are usually straightforward to diagnose on
output problems at rates over 1400–1600 mL min-1 colour Doppler ultrasound, although occasionally
depending on the cardiac reserve of the patient. there may be some difficulty if there is also a large
A fistula flow rate of at least 200 mL min-1 is haematoma present. The incidence of femoral
required for adequate dialysis, ideally rates of false aneurysms is increasing as more and more
300–400 mL min-1 are desirable. Some patients interventional procedures are carried out through
may tolerate flow rates in excess of 1000 mL min-1 femoral puncture sites; rates of 0.2–0.5% after
but others will start to show signs of cardiac diagnostic arteriography and 2–8% following
decompensation at and above this level of flow.28 coronary artery angioplasty and stent placement
In synthetic PTFE fistulae, flow rates less than have been reported.29 The characteristic appear-
500 mL min-1 are associated with a significantly ance is a hypoechoic area which shows swirling
higher risk of failure.24 blood flow on colour Doppler (Fig. 4.14a). It is
The venous side of the fistula may show important that the relationship of the aneurysm
branching into two or more venous channels; one to the femoral artery is identified to ensure that
of these is usually dominant and used for dialysis any therapeutic thrombin injection is made into
needle puncture. However, the multiplicity of the aneurysm and not the artery. This may be 91
4
a Fig. 4.13 Estimation of dialysis
fistula flow volume. (a) Flow in
the radial artery above the
fistula is estimated at 491 mL
min-1. (b) Another patient with a
large artery and excessive flow
of 8.8 L min-1.
difficult to define in the presence of a significant the presence of the fistula is the loss of the ‘to-
haematoma; in these cases identification of the and-fro’ flow in the track from the artery, with
artery above and below the haematoma and tracking only forward flow being shown which increases
back to the aneurysm area will allow some towards end-diastole. The need for intervention
assessment of the relationships to be reached. to treat a false aneurysm needs careful considera-
Spectral Doppler of the track between the artery tion: thrombin injection, surgery and possibly
and the aneurysm, or at the aneurysm neck, shows ultrasound guided compression are all recognised
a characteristic ‘to-and-fro’ flow signal as blood treatments but it should be remembered that a
flows in during systole and out during diastole significant proportion of pseudoaneurysms will
(Fig. 4.14b). Rarely, a false aneurysm may be thrombose spontaneously within 2–3 weeks; in
associated with an arteriovenous fistula passing one surgical series, 86% of 147 patients managed
92 from the cavity to an adjacent vein. The clue to conservatively showed spontaneous thrombosis
4
False aneurysms
b
Fig. 4.14 (a) A false aneurysm showing a patent lumen on colour Doppler; (b) the characteristic ‘to-and-fro’ flow
on spectral Doppler; (c) tip of the needle for thrombin injection seen within the false aneurysm; (d) thrombosed
false aneurysm following treatment with thrombin.
of their false aneurysms, or arteriovenous fistulae, aneurysmal sac under ultrasound control. In the
within a mean period of 23 days.30 great majority of cases, this will produce imme-
diate and complete thrombosis of the false aneurysm
Ultrasound-guided treatment within a few seconds. When the needle tip is
of false aneurysms positively identified within the aneurysm (Fig.
Thrombin injection 4.14c.), the thrombin solution is injected into the
Bovine thrombin and, more recently, human lumen of the aneurysm through a 22–25G needle
thrombin are now available in many centres and attached to an insulin type of syringe (1 mL) to
the treatment of choice for most false aneurysms allow careful titration of the amount injected;
is injection of human thrombin directly into the normally 400–1000 IU of thrombin will be 93
4
sufficient to produce thrombosis within 20–30 s continue down the native artery. This allows the
(Fig. 4.14d). Following the procedure the aneurysm and track to thrombose and therefore
patient should stay on bed rest for 1–3 h with remove the need for surgery.35 There are, how-
regular review of the distal circulation in the ever, some circumstances where it is recognised
affected limb and most centres undertake a that compression is unlikely to succeed and direct
check ultrasound examination at 24 h. In some referral for surgery should be considered; these
cases a small residual region of continuing flow are shown in Table 4.8.The most common contra-
may require a second injection. One study of indications for compression are the age of the
101 false aneurysms showed 95% primary success aneurysm and warfarin therapy; if it has been
rate after a single injection and 98% after a present for more than 7–10 days then the track
second injection in three cases; only two patients will have started to develop an endothelium and
required surgical repair.29 Sheiman et al31 looked the surrounding tissues are less compliant, so that
at the reasons for failure in a small series of false adequate compression becomes difficult.
aneurysms undergoing thrombin injection; they The procedure is quite time consuming for
found that arterial laceration of >8 mm was the the operator and uncomfortable for the patient,
cause of failure in four out of five cases and local so it is better to give some analgesia to the patient
infection the most likely cause in the remaining prior to the commencement of prolonged
patient. compression. The aneurysm and its track are
Bovine thrombin may produce allergic identified and compression is applied by pressing
responses in the patient and induce antibodies the transducer increasingly firmly down on these
to native clotting factors; therefore, it is better to until flow in them has ceased but flow is still
use human-derived thrombin if this is available. present in the native artery. This degree of pres-
Contraindications to thrombin injection include sure is then maintained for 10–15 min before
ischaemic skin changes over a large aneurysm, being released slowly. If flow is then seen in the
local infection, evidence of nerve compression, lumen, compression is reinstated for a further
or evidence of an arteriovenous fistula.Thrombin period of 10 min before again gently relaxing the
injection is reported to have worked satisfactorily pressure.These cycles of compression and relax-
after failed compression therapy in patients on ation are repeated until all flow in the aneurysm
full antiplatelet and anticoagulation therapy.32 lumen has stopped. Usually the lumen of the
There have been reports of the successful aneurysm thromboses in an irregular fashion from
treatment of false aneurysms in other sites by the outside inwards, until complete obliteration
thrombin injection; these include aneurysms in is achieved. In one large series,35 the average time
relation to haemodialysis access sites, an anterior for successful treatment of a simple unilocular
tibial artery pseudoaneurysm that developed after aneurysm was 43 min (SD ± 40 min) and 69 min
a tibial osteotomy and a common carotid artery (SD ± 54 min) for complex multiloculated lesions.
pseudoaneurysm that developed following an Another large series showed a 72% primary
unsuccessful central venous line insertion.33, 34 success rate using compression in 297 false
Ultrasound guided compression

If suitable thrombin preparations are not avail- Table 4.8 Circumstances when compression of
able, then treatment of false aneurysms using a false aneurysm is less likely to succeed
ultrasound-guided compression might be consid- • Aneurysms more than 7–10 days old
ered.The ability of colour Doppler to demonstrate • Associated infection
flowing blood in the aneurysm and track allows • Severe pain/discomfort
the operator to apply graded compression using • Large haematoma
pressure from the transducer, so that flow into • Aneurysms above the inguinal ligament
94 the aneurysm is stopped, whilst allowing flow to
4
Brachial compression syndromes

aneurysms, 7/12 patients who underwent a second
compression episode had successful occlusion
giving a secondary success rate of 74%.36
Following successful obliteration the patient is
scanned after 24 h to confirm that the aneurysm
remains occluded (Fig. 4.14d). Some operators
prefer to dispense with transducer compression
and use manual compression with their fingers
or a fist pressing over the site of the false aneurysm
and its neck; ultrasound is used intermittently to
assess the development of thrombosis. A variety
of mechanical compression devices have also been
proposed but results with these are variable.
Other pulsatile masses

Aneurysms
Aneurysms of the lower limb arteries may occur
in isolation, or as multiple lesions.They are more Fig. 4.15 Cystic adventitial disease of the popliteal
common in the popliteal arteries but may be found artery. The cystic component is seen above the
elsewhere in the legs or, more rarely in the upper patent lumen.
limb arteries.37 As with aortic aneurysms they can
enlarge over a period of time; their size and rate
of enlargement can be followed with ultrasound. be due to congenital fibrotic bands at the inser-
Their main clinical significance is that they can tions of the anterior and middle scalene muscles,
act as a source of emboli to the distal lower limb, or to compression associated with a cervical rib.
or may thrombose leading to acute ischaemia. The accompanying vein is also usually affected.
Some degree of compression may be seen in up
Cystic adventitial disease to 20% of normal subjects.41 The compression is
This is a rare condition in which fluid-filled cysts often positional, typically occurring when the
are found in the wall of the artery, usually the arm is elevated above the head, so various posi-
popliteal artery, or superficial femoral artery in tions of the arm and shoulder may need to be
the adductor canal. The aetiology of these cysts assessed. Sometimes the symptoms only occur
is unclear and various theories, including repeated with the patient in a particular position such as
trauma, ectopic synovial tissue, or a congenital standing or lying down. Usually the diagnosis is
abnormality linking the cyst to the adjacent knee straightforward on clinical grounds, with the
joint space or to adjacent tendons, have been pulse in the affected arm disappearing when the
proposed.38 Signs and symptoms of ischaemia arm is in the appropriate position. However, in
can be variable and confusing.39 Ultrasound is some cases the diagnosis, or the cause, is less clear
useful in diagnosis as it will show the narrowing cut and colour Doppler can be used to image
of the lumen, together with the associated cystic the subclavian artery as the arm is moved into
lesion producing the stenosis (Fig. 4.15).40 various positions with the patient supine or erect.
The transducer is usually placed in the supra-
clavicular position but scanning under the clavicle
BRACHIAL COMPRESSION
may be useful, particularly when the arm is fully
SYNDROMES
elevated. Careful examination of the artery as it
Thoracic outlet syndrome, or compression of the passes over the first rib may show changes in the
subclavian artery as it crosses the first rib, may waveform as the vessel is compressed (Fig. 4.16). 95
4
Sometimes access to the subclavian artery in the by the limited access to the popliteal fossa with
region of the first rib can be difficult. In these the knee flexed. However, careful examination
cases, scanning the axillary artery whilst the arm of the lower popliteal artery or posterior tibial
is put in different positions will show any changes artery with the knee in different positions of flexion
in velocity or waveform resulting from arterial will demonstrate changes in the arterial waveform
compression. A rebound increase in velocity resulting from compression. Athletes may get
may be noted in the distal arteries on release of compression of the popliteal artery between the
compression.42 lateral femoral condyle or upper tibia and the
hypertrophied gastrocnemius, or soleus and
Popliteal artery compression plantaris muscles. This can be demonstrated on
In cases of suspected entrapment of an artery or Doppler ultrasound by scanning with the foot in
graft in the adductor canal or popliteal fossa, a neutral position and then in a position of active
direct examination of the vessel is often restricted plantar flexion.43,44
Rarely, athletic patients may get iliac artery
compression from a bulky psoas muscle on
a strenuous exercise. Scanning following a period
of exercise shows turbulence, tissue bruits and
very high local velocities in the inguinal area.
ACCURACY IN RELATION TO
OTHER TECHNIQUES
The gold standard for the assessment of the accu-
racy of Doppler ultrasound is usually arterio-
graphy.The reservations on the accuracy of arte-
riography, which are discussed in Chapter 3 on
carotids, are also applicable to peripheral arterial
disease. Cossman et al11 compared colour Doppler
with arteriography in 84 limbs, from the iliac to
b
lower popliteal segments, using the criteria dis-
cussed earlier. For the detection of stenoses
greater than 50%, they found an overall sensi-
tivity of 87% (156/180 segments), specificity of
99% (444/449), accuracy of 95% (600/629),
positive predictive value of 96% and negative
predictive value of 95%. For the diagnosis of
arterial occlusion the overall sensitivity was 81%
(76/94), with specificity of 99% (463/466), accu-
racy of 96% (539/560), positive predictive value
of 95% and negative predictive value of 96%.
Polak45 reviewed five studies performed
between 1989 and 1992, including the study by
Fig. 4.16 Compression of the subclavian artery as
the arm is elevated/abducted. At rest flow in the Cossman et al.11 Colour Doppler was compared
axillary artery is approximately 1m min-1; (b) showing to arteriography and the overall sensitivity for
narrowing of the artery as it crosses the first rib as the detection of a stenosis greater than 50% was
the position of the arm is changed. The velocity 87.5% (316/361 segments), for an occluded segment
96 increases to 3 m min-1. the sensitivity was 92.6% (403/435), and the
4
Conclusions
overall specificity for the identification of normal tages of radiation and contrast injection. One
segments was 97% (1247/1282). These studies study50 looked at 44 patients using a four-slice
have assessed femoropopliteal disease, accurate multidetector CT system.Two vascular radiologists
assessment of infrapopliteal crural vessel disease compared the results to DSA images with sensi-
with duplex is a little more difficult, although a tivities of 79% and 72% for treatable lesions
sensitivity of 77% has been reported46 but if infor- (stenoses >50%) and specificities of 93% each.
mation on these vessels is clinically important then However, other workers51 have reported better
arteriography should be considered if there are results with overall sensitivity of 93% and 95%
any reservations about the duplex examination. specificity and an overall accuracy of 94%; although
Similar, positive results have been obtained results for the infracrural arteries showed only
for duplex in upper limb disease with 79% sensi- 85% sensitivity. CTA has also been used for the
tivity, 100% specificity and 99% accuracy for assessment of brachial artery compression and is
haemodynamically significant (50–70%) stenoses useful in demonstrating the relation of the artery
and 98% sensitivity, 99% specificity and 99% to adjacent bone and other musculoskeletal
accuracy for occlusion being reported by Tola et structures.
al47 who compared duplex ultrasound with
intraarterial digital subtraction angiography
CONCLUSIONS
(DSA) in 578 upper limb arterial segments.
MRA is now being applied to the peripheral Providing that the examination is performed
arteries with some promising results, particu- carefully by a skilled operator, Doppler ultrasound
larly with contrast enhancement, but availability provides a relatively cheap and an accurate tech-
and complexity, together with poor resolution in nique for the assessment of many patients with
the smaller calf vessels, mean that this is still disease or previous surgery to the peripheral arteries,
being developed, rather than a generally particularly in the lower limbs. In symptomatic
available technique. Hingorani et al48 compared patients it can be used as a first-line test to
arteriography, duplex ultrasound and MRA in identify those patients without significant disease,
33 patients undergoing lower limb revasculari- those patients who may benefit from angioplasty
sation procedures. There was a difference and patients who are likely to require surgical
between arteriography and duplex in three cases bypass. In many centres, ultrasound is now used
(10%) and this was considered to be clinically alone prior to surgery, with CTA or MRA used
significant in two cases. For MRA, there were to clarify the situation in problematic cases.
differences compared with arteriography in 12 Power Doppler and echo-enhancing agents will
cases (36%) and nine of these were considered increase the diagnostic sensitivity of Doppler
to be clinically significant. ultrasound and the need for arteriography in
CTA can also be used for the assessment of many cases should be substantially reduced.
the lower limb arteries49 but has the disadvan-
REFERENCES
1. Fowkes FGR. Epidemiology of atherosclerotic 4. Edwards JM, Coldwell DM, Goldman ML, et al.
disease in the lower limbs. Eur J Vasc Surg 1988; The role of duplex scanning in the selection of
2:283–291. patients for transluminal angioplasty. J Vasc Surg
2. Department of Health and Social Security, Office of 1991; 13:69–74.
Population Censuses and Surveys. Hospital 5. Ramaswami G, Al-Kutoubi A, Nicolaides AN, et al.
Inpatient Enquiry. London: HMSO; 1988. The role of duplex scanning in the diagnosis of
3. Leng GC, Evans CJ, Fowkes FGR. Epidemiology of lower limb arterial disease. Ann Vasc Surg 1999;
peripheral vascular diseases. Imaging 1995; 7:85–96. 13:494–500. 97
4
6. Langholz J, Schlief R, Schürmann R, et al. Contrast synthetic infrainguinal bypass grafts. Eur J Vasc
enhancement in leg vessels. Clin Radiol 1996; Endovasc Surg 1996; 11:441–445.
51(suppl 1):31–34. 22. Hoballah JJ, Nazzal RM, Ryan SM, et al. Is color
7. Flanigan DP, Tullis JP, Streeter VL, et al. Multiple duplex surveillance of infrainguinal
subcritical arterial stenoses: effect on poststenotic polytetrafluoroethylene grafts worthwhile? Am J
pressure and flow. Ann Surg 1977; 186:663–668. Surg 1997; 174:131–135.
8. Allard L, Cloutier G, Guo Z, et al. Review of the 23. Gibson KD, Caps MT, Gillen D, et al. Identification
assessment of single level and multilevel arterial of factors predictive of lower extremity vein graft
occlusive disease in lower limbs by duplex thrombosis. J Vasc Surg 2001; 33:24–31.
ultrasound. Ultrasound Med Biol 1999; 25:495–502. 24. Wiese P, Nonnast-Daniel B. Colour Doppler
9. Whyman MR, Allan PL, Gillespie IN, et al. ultrasound in dialysis access. Nephrol Dial
Screening patients with claudication from Transplant 2004; 19:1956–1963.
femoropopliteal disease before angioplasty using 25. Dumars MC, Thompson WE, Bluth EI, et al.
Doppler colour flow imaging. Br J Surg 1992; Management of suspected hemodialysis graft
79:907–909. dysfunction: usefulness of diagnostic US.
10. Jager KA, Ricketts HJ, Strandness DE Jr. Duplex Radiology 2002; 222:103–107.
scanning for evaluation of lower limb arterial 26. Robbin ML, Oser RF, Allon M, et al. Hemodialysis
disease. In: Bernstein EF, ed. Noninvasive access graft stenosis: US detection. Radiology
diagnostic techniques in vascular disease. St Louis: 1998; 208:655–661.
CV Mosby; 1985:619–631. 27. Malik J, Slavikova M, Maskova J. Dialysis access-
11. Cossman DV, Ellison JE, Wagner WH, et al. associated steal syndrome: the role of
Comparison of contrast arteriography to arterial ultrasonography. J Nephrol 2003; 16:903–907.
mapping with color-flow duplex imaging in the 28. Landwehr P. Haemodialysis shunts. In: Wolf K-J,
lower extremities. J Vasc Surg 1989; 10:522–529. Fobbe F, eds. Colour duplex sonography. New York:
12. van Asten WN, van Lier HJ, Beijnevald WJ, et al. Thieme; 1995:92–109.
Assessment of aortoiliac obstructive disease by 29. Maleux G, Hendrickx S, Vaninbroukx J, et al.
Doppler spectrum analysis of blood flow velocities Percutaneous injection of human thrombin to treat
in the common femoral artery at rest and during iatrogenic femoral pseudoaneurysms: short- and
reactive hyperaemia. Surgery 1991; 109:633–639. midterm ultrasound follow-up. Eur Radiol 2003;
13. Zierler RE. Duplex and color-flow imaging of the 13:209–212.
lower extremity arterial circulation. Semin 30. Toursarkissian B, Allen BT, Petrinec D, et al.
Ultrasound CT MR 1990; 11:168–179. Spontaneous closure of selected pseudoaneurysms
14. Sensier Y, Bell PR, London NJ. The ability of and arteriovenous fistulae. J Vasc Surg 1997;
qualitative assessment of the common femoral 25:803–808.
Doppler waveform to screen for significant aorto- 31. Sheiman RG, Mastromatteo M. Iatrogenic femoral
iliac disease. Eur J Vasc Endovasc Surg 1998; pseudoaneurysms that are unresponsive to
15:357–364. percutaneous thrombin injection: potential causes.
15. Rosfors S, Eriksson M, Hoglund N, et al. Duplex Am J Roentgenol 2003; 181:1301–1304.
ultrasound in patients with suspected aortoiliac 32. Gorge G, Kunz T. Thrombin injection for treatment of
occlusive disease. Eur J Vasc Surg 1993; 7:513–517. false aneurysms after failed compression therapy in
16. Bandyk DF. Essentials of graft surveillance. Semin patients on full dose antiplatelet and heparin
Vasc Surg 1993; 6:92–102. therapy. Catheter Cardiovasc Interv 2003;
17. Mills JL. Mechanisms of graft failure: the location, 58:505–509.
distribution and characteristics of lesions that 33. Ghershin E, Karram T, Gaitini D, et al. Percutaneous
predispose to graft failure. Semin Vasc Surg 1993; ultrasonographically guided thrombin injection of
6:78–91. iatrogenic pseudoaneurysms in unusual sites.
18. Fasih T, Rudol G, Ashour H, et al. Surveillance J Ultrasound Med 2003; 22:809–816.
versus nonsurveillance for femoro-popliteal bypass 34. Holder R, Hilton D, Martin J, et al. Percutaneous
grafts. Angiology 2004; 55:251–256. thrombin injection of carotid artery
19. Landry GJ, Moneta GL, Taylor LM Jr, et al. Long- pseudoaneurysm. J Endovasc Ther 2002; 9:25–28.
term outcome of revised lower-extremity grafts. 35. Coley BD, Roberts AC, Fellmeth BD, et al.
J Vasc Surg 2002; 35:56–62. Postangiographic femoral artery pseudoaneurysms:
20. Sanchez LA, Suggs WD, Veith FJ, et al. Is further experience with US-guided compression
surveillance to detect failing polytetrafluoroethylene repair. Radiology 1995; 194:307–311.
bypasses worthwhile: 12-year experience with 91 36. Eisenberg L, Paulson EK, Kliewer MA, et al.
grafts. J Vasc Surg 1993; 18:981–990. Sonographically guided compression repair of
21. Dunlop P, Sayers RD, Naylor AR, et al. The effect of pseudoaneurysms: further experience from a single
a surveillance programme on the patency of institution. Am J Roentgenol 1999; 173:1567–1573.
98
37. Morrissey NJ. Endovascular treatment of peripheral
46. Karacagil S, Lofberg AM, Granbo A, et al. Value of

4
References
arterial aneurysms. Mt Sinai J Med 2004; 71:1–3. duplex scanning in evaluation of crural and foot
38. Flanigan DP, Burnham SJ, Goodreau JJ, et al. arteries in limbs with severe lower limb ischaemia –
Summary of cases of adventitial disease of the a prospective comparison with angiography. Eur J
popliteal artery. Ann Surg 1979; 189:165–175. Vasc Endovasc Surg 1996; 12:300–303.
39. Cassar K, Engeset J. Cystic adventitial disease: 47. Tola M, Yurdakul M, Okten S, et al. Diagnosis of
a trap for the unwary. Eur J Endovasc Surg 2005; arterial occlusive disease of the upper extremities:
29:93–96. comparison of color duplex sonography and
40. Tsolakis IA, Walvatne CS, Caldwell MD. Cystic angiography. J Clin Ultrasound 2003; 31:407–411.
adventitial disease of the popliteal artery: diagnosis 48. Hingorani A, Ascher E, Markevich N, et al.
band treatment. Eur J Vasc Endovasc Surg 1998; A comparison of magnetic resonance angiography,
15:188–194. contrast arteriography, and duplex arteriography for
41. Longley DG, Schwabacher S, Yedlicka JW, et al. patients undergoing lower extremity
Thoracic outlet syndrome: evaluation of the revascularization. Ann Vasc Surg 2004; 18:294–301.
subclavian vessels by colour duplex sonography. 49. Becker CR, Wintersperger B, Jakobs TF.
Am J Radiol 1992; 158:623–630. Multidetector-row CT angiography of peripheral
42. Wadhwani R, Chaubal N, Sukthankar R, et al. arteries. Semin Ultrasound CT MR 2004;
Color Doppler and duplex sonography in 5 patients 24:268–279.
with thoracic outlet syndrome. J Ultrasound Med 51. Edwards AJ, Wells IP, Roobottom CA. Multidetector
2001; 20:795–801. row CT angiography of the lower limb arteries: a
43. Turnipseed WD, Pozniak M. Popliteal entrapment as prospective comparison of volume-rendered
a result of neurovascular compression by the techniques and intra-arterial digital subtraction
soleus and plantaris muscles. J Vasc Surg 1992; angiography. Clin Radiol 2005; 60:85–95.
15:285–294. 52. Romano M, Mainenti PP, Imbriaco M, et al.
44. Baltopoulos P, Filippou DK, Sigala F. Popliteal artery Multidetector row CT angiography of the abdominal
entrapment syndrome: anatomic or functional aorta and lower extremities in patients with
syndrome? Clin J Sports Med 2004; 14:8–12. peripheral arterial occlusive disease: diagnostic
45. Polak JF. Peripheral arterial disease. Evaluation with accuracy and interobserver agreement. Eur J Radiol
color flow and duplex sonography. Radiol Clin North 2004; 50:303–308.
Am 1995; 33:71–90.
99
5
The peripheral veins
The peripheral veins may be affected by a variety respect to the diagnosis, or exclusion, of dangerous
of disorders, which can be assessed by ultrasound. proximal thrombus in symptomatic patients.3 The
Deep vein thrombosis (DVT) and thrombo- results for asymptomatic thrombus in the lower
embolic disease are the most common indica- limbs are less encouraging and this should be
tions for investigation of the peripheral veins but recognised when using ultrasound to screen for
venous insufficiency and vein mapping are also DVT in asymptomatic patients.4
reasons for examining the veins. Anderson et al1 Recurrence of varicose veins following surgery
found an average annual incidence of 48 initial can pose many problems for the clinician trying
cases, 36 recurrent cases of DVT and 23 cases to clarify the venous anatomy. Colour Doppler
of pulmonary embolus per 100 000 population can be used instead of venography and varico-
in the Worcester DVT study.1 The prevalence of graphy in most cases and may be the only
varicose veins and chronic venous insufficiency examination required to define the anatomy
is more difficult to quantify, but it has been and function in patients with recurrent varicose
estimated that 10–15% of males and 20–25% of veins.5
females in an unselected Western population over The impact of postphlebitis syndromes and
15 years of age have visible tortuous varicose chronic venous insufficiency is a rather larger
veins; 2–5% of adult males and 3–7% of females problem than is apparent from its relatively low
have evidence of moderate or severe chronic clinical profile. In one large epidemiological study
venous insufficiency, with a point prevalence for of 4376 subjects, 62% had some evidence of
active ulceration of 0.1–0.2%.2 varicose veins; signs of chronic insufficiency were
present in 22%.6 Varicography shows perforator
veins which are obviously incompetent and some
INDICATIONS
The indications for ultrasound of the venous
system are shown in Table 5.1. The most
Table 5.1 Indications for venous ultrasound
frequent indication for ultrasound of the veins is
for the investigation of possible DVT in the • Diagnosis or exclusion of deep vein thrombosis in
lower limb and, occasionally, in the upper limb – the upper or lower limb, spontaneous or related
to indwelling catheters
especially if there have been central venous
• Assessment of secondary/recurrent varicose
catheters inserted for intensive care monitoring, veins
chemotherapy, dialysis or parenteral feeding. • Investigation of chronic venous insufficiency and
postphlebitis syndrome
Similarly, indwelling femoral catheters are prone
• Vein mapping prior to bypass grafts
to induce thrombosis and patients should be
• Assessment of primary varicose veins if there is
examined early if this is suspected. Ultrasound uncertainty following clinical examination
provides a non-invasive, reliable method for • Localisation of veins for cannulation
examining the venous system, particularly with 101
5
Anatomy – lower limb
incompetent superficial and deep venous segments,
but ultrasound has the advantage that the The veins of the lower limb are divided into
segments of the deep and superficial systems can deep and superficial systems. These are linked
be examined and the direction of blood flow by a variable number of perforator veins which
within each segment can be demonstrated. In carry blood from the superficial to the deep
addition, it is less unpleasant for the patient and systems (Figs 5.1 and 5.2).
allows multiple assessments to be performed
without discomfort. The main disadvantage is The deep veins
that it is fairly time-consuming, particularly in The anatomy of the lower limb veins is rather
complex cases, and requires a significant degree variable. Generally the veins accompany the
of expertise in order to perform examinations arteries but their number may vary and the
efficiently. communications with other veins along the way
The superficial veins of the legs and, occa- can show a variety of patterns; however, a general
sionally, the arms may be used for bypass grafts arrangement is usually apparent. In the calf
for the coronary or lower limb arteries. If there there are veins running with the main arteries:
is any doubt about their suitability as a conduit the posterior tibial, peroneal and anterior tibial veins;
following previous varicose vein surgery, or in there are usually two, occasionally three veins
terms of their calibre, ultrasound can be used to with each artery (Fig. 5.3). In addition there are
assess the diameter and length of vein available. venous channels, or sinuses, which drain the
In addition, the sonographer can map out the major muscle groups in the posterior calf. These
course of the vein to allow easier harvesting. are seen in the upper calf as they pass upwards
It may be difficult occasionally to locate a to join the other deep veins in the lower popliteal
suitable vein for central venous cannulation, region; the gastrocnemius and soleal veins are the
particularly in patients who have had multiple largest of these. The gastrocnemius vein is the
previous central venous lines, such as intensive more superficial and may be mistaken for the
care or chemotherapy patients. Ultrasound can short saphenous vein; clues to its true identity
be used to clarify the location and patency of are that it is usually accompanied by the artery
potentially suitable veins and, in difficult cases, to the muscle and it can be followed distally
the puncture may be made under direct ultra- down into the muscle rather than outwards to lie
sound visualisation. subcutaneously on the fascia around the calf,
which is the position of the short saphenous vein.
The calf veins join to form the popliteal vein,
ANATOMY AND SCANNING
or veins – there may be two, or sometimes three
TECHNIQUE
channels, especially if there is a dual superficial
The anatomy of the venous system in the limbs femoral vein. The popliteal vein runs up through
is more complex and variable than that of the the popliteal fossa, lying more posterior and
arteries. The meanings of the terms ‘proximal’ usually medial to the artery. As well as the veins
and ‘distal’ may cause confusion as the veins from the calf and calf muscles, it is joined by
start at the periphery and blood flows centrally the short saphenous vein at the saphenopopliteal
towards the heart so that ‘upstream’ is peripheral junction.
and ‘downstream’ is central, which is the oppo- The popliteal vein becomes the superficial
site from the situation in the arteries. The con- femoral vein at the upper border of the popliteal
vention is that proximal describes locations nearer fossa; rarely, the popliteal vein runs more deeply
the heart and distal refers to points further from to join with the profunda femoris vein. The
the heart; these terms are used in this way in this superficial femoral vein runs up the medial
chapter. aspect of the thigh, posterior to the superficial
102
5

Inferior
vena cava
Common iliac
External iliac
Internal iliac
Profunda Profunda
femoris Long saphenous femoris
Femoral
Popliteal
Popliteal
Short
saphenous
Peroneal
Anterior
tibial
Anterior tibial Posterior

tibial
Lateral/marginal
Lateral
marginal
Dorsal
Dorsal
Deep plantar
venous arch
venous arch
Medial Posterior Medial/lateral
marginal tibial plantar
Fig. 5.1 The veins of the lower limb, showing the superficial and deep systems.
femoral artery to join with the profunda femoris have significant segments of duplication (Fig. 5.4)
vein in the femoral triangle below the groin; the along its length in up to 25–30% of subjects,7, 8
profunda femoris vein drains the thigh muscles. these dual segments may have a variable relation
The confluence of the superficial femoral and to the artery, so that they may be overlooked
profunda femoris veins to form the common unless care is taken in the examination of the
femoral vein is normally a little more caudal thigh veins with both transverse and longitudinal
than the bifurcation of the common femoral views being obtained.
artery into the superficial femoral and profunda In the pelvis and groin, the anatomy is generally
femoris arteries.The superficial femoral vein may consistent. The superficial femoral vein and pro- 103
5
Posteromedial vein
of thigh (to profunda
femoris vein)
Posteromedial vein
(to profunda femoris
and superficial
femoral vein)
Long Communication with

saphenous short saphenous vein Posterolateral
vein vein (to profunda
DODD Long femoris vein)
(to superficial saphenous
femoral vein) vein
Communication Short
between saphenous
saphenous veins vein
BOYD
(to gastrocnemius vein)
Anterior
Gastrocnemius
tibial vein
perforators
Posterior Posterior arch vein
tibial vein (to gastrocnemius Soleus
and soleus veins) perforators
I COCKETT
II (to posterior tibial BASSI
III and muscle veins)
MAY
(to posterior tibial
or
and plantar veins)
KUSTER
Fig. 5.2 The major perforating veins in the lower limb.
funda femoris vein join to form the common muscles around the hip. These veins are variable
femoral vein, which lies medial to the common in size and number, and occasionally one of these
femoral artery. The common femoral vein is is large enough to be confused with the long
joined by the long saphenous vein at the sapheno- saphenous vein or profunda femoris vein but
femoral junction; the appearance of the common careful attention to the anatomy should clarify
femoral vein, long saphenous vein and artery in the situation.The common femoral vein becomes
transverse section is sometimes referred to as the the external iliac vein after it has passed under the
‘Mickey Mouse’ view (Fig. 5.5). The common inguinal ligament, and then it passes posteriorly
104 femoral vein is also joined by veins from the along the posterior pelvis, running alongside the
5

Fig. 5.3 Power Doppler showing paired deep calf
Fig. 5.5 The saphenofemoral junction showing the
veins running with the artery.
‘Mickey Mouse’ view.
external iliac artery.The internal iliac vein, which then join at the level of the aortic bifurcation to
drains the pelvic structures, joins with the external form the inferior vena cava, which normally passes
iliac vein deep in the pelvis to form the common cranially on the right side of the aorta. The left
iliac vein (Fig. 5.6). The two common iliac veins common iliac vein passes behind the right com-
Fig. 5.4 (a) Transverse view showing dual superficial femoral vein segments; (b) another example of multiple
superficial femoral vein segments showing a central artery with four venous channels adjacent to it. 105
5
the thigh, the superficial femoral vein usually has
one just below the confluence with the profunda
femoris vein and at several levels below this. The
iliac veins, in contrast, have relatively few valves;9
rarely a valve may be seen in the inferior vena
cava.
The superficial veins

The two main superficial venous channels in the
lower limb are the long and short saphenous
veins. The long saphenous vein arises from the
medial aspect of the dorsal venous arch of the
foot and passes in front of the medial malleolus
to run up the medial aspect of the calf and knee
Fig. 5.6 The confluence of the internal and external into the thigh. In the upper thigh, the long
iliac veins forming the common iliac vein. saphenous vein curves laterally and deeply to
join the common femoral vein just below the
inguinal ligament. The long saphenous vein has
mon iliac artery just distal to this confluence. In two components in the calf: the posterior divi-
a small number of individuals this confluence sion passes up from the medial malleolus and
does not occur and the two common iliac veins communicates with the perforator veins; the
continue cranially as dual inferior venae cavae; anterior division usually joins the posterior
this reflects the arrangement of paired cardinal division just below the level of the knee joint.
veins in the embryo. Duplication of the long saphenous vein can be
The deep veins have a series of valves along seen in the thigh in up to 50% of people,10 this
their course (Fig. 5.7).These are somewhat vari- usually takes the form of parallel channels. The
able in their number and location. They are long saphenous vein receives many superficial
most numerous in the veins below the knee; in tributaries and is connected to the deep veins by
b
a
106 Fig. 5.7 Normal valves in the profunda femoris vein; (a) valves shut, (b) valves open.
5

perforating veins; some of these tributaries in the will be dealt with in subsequent sections (Table
thigh can be quite prominent and may be mistaken 5.2). A 4–7 MHz linear transducer is the most
for the main vein if their true nature is not recog- suitable frequency as it provides sufficient
nised. The standard sites for the perforators are penetration, particularly in large or oedematous
at the level of the junction between the middle thighs. A higher-frequency transducer may be
and lower thirds of the thigh and in the calf, used for superficial veins, or in thinner legs. It is
although many other communicating veins have important to ensure that the system is set up for
been recorded and described (Fig. 5.2).11 In the the slower velocities found in veins, rather than
region of the saphenofemoral junction the long the significantly higher arterial velocities.
saphenous vein receives several tributaries drain- It is advantageous if there is a tilting couch
ing the groin, lower abdominal wall and perineum. available so that the patient can be moved from
These veins are of significance in the recurrence the horizontal to various degrees of head-up
of varicose veins following high ligation, as they elevation as necessary. In the absence of a tilting
provide a network of collateral channels which couch it is better if the patient can be examined
may bypass the resected segment. with the thorax higher than the legs, as this pro-
The short saphenous vein arises from the lateral duces some distension of the lower limb veins,
aspect of the dorsal venous arch of the foot, which makes them easier to identify and the
passing below and behind the lateral malleolus assessment of compression more straightforward.
to run up the posterolateral aspect of the calf to There are three components to the ultrasound
the popliteal fossa, where it passes through the examination of the veins for DVT: imaging,
deep fascia to join the popliteal vein. Classically, Doppler and compression. Thrombus may be
it enters the lateral aspect of the popliteal vein at seen in the vein, Doppler may show abnormal,
the level of the popliteal skin crease or within a or absent, flow signals and compression refers to
few centimetres above this but the level of the the fact that a normal vein is easily compressible
confluence can be quite variable. Occasionally – light pressure with the transducer will obliterate
it passes upwards to join the profunda femoris the lumen of the vein.Two points should be noted
vein in the lower thigh – a Giacomini vein.12 in relation to compression: first, compression
Burihan and Baptista-Silva13 dissected 200 adult should be performed in the transverse plane
cadaver legs and reported 20 different patterns (Fig. 5.8) for the reason that if it is done in the
of termination of the short saphenous vein. In longitudinal plane a thrombosed vein may dis-
27.5% of legs the short saphenous vein terminated appear as it is no longer in the scan plane, rather
in the principal deep vein of the leg (popliteal or
lower superficial femoral vein), in 25% of legs
the short saphenous vein, or a branch arising
from it, communicated with the long saphenous Table 5.2 Basic steps in the examination for
thrombosis
vein. In the remaining legs, there was a wide
variety and combination of communications with 1. Patient sitting on couch/trolley: compression of
other veins, including the deep femoral vein, the common femoral vein, superficial femoral vein
from groin to adductor canal
mid-thigh perforator vein, muscular veins and
2. Colour Doppler with augmentation, examine
even the inferior gluteal vein in three legs. common and superficial femoral veins
3. Patient decubitus, or with leg elevated:
compression and Doppler examination of
Scanning technique – lower limb popliteal vein(s)
The technique varies depending on the clinical 4. Patient sitting, if possible, with legs dependent:
indication. The most common indication is the examine the calf veins with compression and
colour Doppler
diagnosis or exclusion of DVT in the lower limb.
5. Patient supine: examine iliac veins if thrombus
This section therefore concentrates on this aspect suspected in these
and variations in technique for other indications 107
5
to confirm or exclude a diagnosis of DVT (Fig.
5.9). If there is any doubt then a spectral assess-
ment will allow a better appreciation of damped
flow, absent respiratory variation and impaired
augmentation.
Once the thigh veins have been examined the
patient is turned into a lateral position, with the
medial aspect of the leg being examined upper-
most, so that the popliteal veins can be exam-
ined. Again, compression and colour Doppler
are used to assess the veins. Some patients,
Fig. 5.8 Normal compression: the lumen of the vein
particularly postoperative hip patients, may not
(arrows) is completely obliterated by pressure from
the transducer.
be able to move into a decubitus position. In
these cases the popliteal veins are examined with
the knee partially flexed up off the couch, with
than because it has been compressed. Second, external rotation, if possible, so that the trans-
fresh thrombus is soft and gelatinous, so that firm ducer can be positioned in the popliteal fossa; a
pressure can produce a degree of compression, curved array can be of benefit in gaining access
which may give a false impression of patency. in this situation. Alternatively, the leg can be
The use of colour Doppler should clarify this elevated and supported off the couch by an
situation but care must be taken if this is not assistant. In addition to the popliteal vein, the
available. A further reason for scanning in the main muscular veins draining soleus and
transverse plane is that dual segments of the gastrocnemius should be assessed, especially if
superficial femoral vein will be identified more there is pain and tenderness associated with the
reliably. posterior calf muscles.
The examination begins at the groin, where The calf veins can be examined after the
the common femoral vein is located on a trans- popliteal vein with the patient in the decubitus
verse scan and compressed. Compression is then position on a tilted couch, or in the supine posi-
repeated at intervals of 3–5 cm down the length tion with the knee flexed up off the mattress, if
of the thigh to the adductor canal. At this point the patient is relatively immobile. Alternatively,
the superficial femoral vein is difficult to com- the patient can sit on the couch with their legs
press from an anterior approach as it is well
supported by the bulk of the anterior thigh
muscles. Compression is better achieved in this
region by placing a hand behind the medial
thigh and pushing up with the fingers against the
transducer. The scan plane is then changed to
longitudinal and the vein examined with colour
Doppler, or power Doppler, as the transducer is
moved up the thigh. Squeezing the calf gently
will augment flow and allow easier detection of
areas of flow or thrombosis; alternatively, the
patient can be asked to plantar-flex their toes,
which results in calf muscle contraction and
emptying of the calf veins.
Colour Doppler signals are often sufficient, in Fig. 5.9 The normal saphenofemoral junction
108 conjunction with the findings on compression, showing complete colour fill-in across the vein lumen.
5

over the side so that the dependent calf veins are to the inferior vena cava and aortic bifurcation;
well distended. The posterior tibial and deeper this can then be followed peripherally. In some
peroneal vessels are most easily located by scan- patients it is impossible to identify the deeper
ning in the transverse plane from the medial side pelvic portion of the iliac veins; however, if there
of the calf and identifying the arterial signals is a patent external iliac vein which shows respi-
on colour Doppler (Fig. 4.2). These veins may ratory variation with good augmentation and a
also be located on a longitudinal scan; again the patent upper common iliac vein, then it is highly
arterial signal provides a useful guide to the unlikely that there is significant thrombus in
position of the veins. If there are difficulties the invisible segment. Transvaginal scanning will
identifying the posterior tibial veins at the mid- show the deeper pelvic veins and may be con-
calf level then scanning the lower calf just above sidered if there is a need to visualise these vessels
the medial malleolus, where the vessels are super- directly. In thinner patients, or patients with
ficial and constant in location, may be of value; good pelvic access, the proximal internal iliac
the posterior tibial vessels can then be followed vein may be seen joining the external iliac vein in
back up the calf with augmentation of flow as the pelvis (Fig. 5.5). The inferior vena cava is
necessary in order to assess patency. In the mid- examined if thrombus is seen extending into this
and lower calf, squeezing the calf can produce vessel. It is important, whenever thrombus is
motion artefacts from movement of the calf diagnosed in a leg vein, that the proximal extent
muscles which obscure the flow signals from of the clot is identified, as this may have a signifi-
the veins; in these cases, squeezing the foot will cant impact on management decisions in rela-
produce adequate augmentation of flow. The tion to anticoagulation therapy, or the placing of
anterior tibial veins are examined from an antero- a filter.
lateral approach: scanning transversely, the tibia, In pregnant women in the later part of preg-
fibula and interosseous membrane are identi- nancy, the uterus will lie on the iliac veins in the
fied. The anterior tibial vessels are found on the supine position and compress them, thus reducing
superficial aspect of the interosseous membrane, flow and impairing augmentation in the lower
although it should be noted that these veins are limb veins. This can be alleviated by asking the
rarely involved in DVT in isolation from the patient to turn into a semi-decubitus position,
other calf veins. The peroneal veins may also be with the side being examined uppermost, so that
visualised deep to the interosseous membrane the uterus falls away, allowing better flow in the
in many patients from this anterolateral aspect, pelvic veins. An alternative is to examine the
allowing their examination if they have not been patient standing, as the uterus moves forward
identified from a posteromedial approach; a away from the iliac veins in this position.
posterolateral approach is also of value in iden-
tifying the deeply situated peroneal veins in Anatomy – upper limb
some patients. The veins of the upper limb are also divided into
The iliac veins are examined by following the deep and superficial groups (Fig. 5.10).The deep
external iliac vein upwards from the common veins are paired and accompany the arteries: the
femoral vein into the pelvis. A 3–5 MHz trans- radial, ulnar and brachial veins. There is a vari-
ducer is usually necessary for adequate penetra- able pattern of communicating veins between
tion. Firm pressure may be required to displace the deep venous channels and between the deep
bowel gas.This may produce narrowing or efface- venous channels and the superficial veins. The
ment of the more superficial segments of vein, superficial system is more variable than in the
resulting in an absence of signal and a possible leg but there are usually two main channels: the
false diagnosis of occlusion. If the pelvic veins cephalic vein on the radial aspect of the arm and
are difficult to trace superiorly then the common the basilic vein on the ulnar side. These commu-
iliac vein can usually be identified just distal nicate at the cubital fossa by way of the median 109
5
Scanning technique – upper limb
Internal jugular vein and neck

Brachiocephalic vein
Examination of the upper limb veins is normally
performed with the patient supine and the arm
Cephalic vein
abducted to about 90°; the patient may require
some support for the arm, or they can be asked
Axillary vein to hold onto some suitable part of the ultrasound
machine beside them. A transducer frequency of
Paired 5–12 MHz can be used.The examination begins
brachial veins
at the sternoclavicular joint, where the distal
Basilic vein brachiocephalic vein can be assessed and the
Median vein of confluence with the internal jugular vein exam-
the forearm ined, particularly if central lines have been
Cephalic vein
Paired inserted. The subclavian vein is examined from
ulnar veins above and then below the clavicle; it is seen lying
Paired
radial veins in front of the subclavian artery as it runs over
Basilic vein the first rib. The axillary vein is then followed
across the axilla into the upper arm, from where
Deep vein
the brachial veins can be examined down to the
Superficial vein
elbow. The veins below this are not usually
examined unless there is some specific reason,
such as the presence of a dialysis shunt. Augmen-
Fig. 5.10 The veins of the upper limb, showing the tation of flow is obtained by manual compres-
main superficial and deep veins. sion of the forearm, or upper arm; alternatively,
asking the patient to clench their fist will increase
venous flow. If there is a suspicion of a possible
cubital vein and they also communicate with the venous compression syndrome, the arm veins
deep brachial veins at this level. The basilic vein can be examined with the limb in different posi-
pierces the deep fascia on the medial aspect of tions of abduction; comparison of flow on the
the mid-upper arm to join the brachial veins and two sides can be of value.
this combined venous channel becomes the The internal jugular vein runs in the carotid
axillary vein when it enters the axilla.The cephalic sheath from the jugular foramen in the base
vein passes more cranially along the lateral aspect of the skull down to join with the subclavian
of the biceps. At the level of pectoralis major it vein; it lies superficial to the carotid artery.There
turns medially and deeply to pierce the clavipec- may be significant variation in size between
toral fascia below the clavicle and joins the upper the two sides. It is normally compressed easily
axillary vein.The axillary vein also receives other by pressure from the transducer, so a light touch
tributaries from the region of the shoulder joint is required. Flow in the vein is influenced
and the lateral chest wall. significantly by right heart activity, so colour
The axillary vein becomes the subclavian vein Doppler will show variable forward and reverse
as it crosses the first rib, where it lies in front of flow, with spectral Doppler showing the ‘a’, ‘c’
the artery; the main tributary of the subclavian and ‘v’ waves of the jugular pulse. Respiratory
vein is the external jugular vein. The subclavian variation will also be seen, with increased forward
vein on both sides joins with the internal jugular flow during inspiration when intrathoracic
vein behind the medial end of the clavicle to pressure is negative and slowing during expira-
form the brachiocephalic vein, which is also tion when the intrathoracic pressure is positive.
110 known as the innominate vein. Flow can therefore be modified by respiratory
5
Diagnosis of deep vein thrombosis

manoeuvres such as deep inspiration, or a occur to the spectral and colour Doppler find-
Valsalva. ings.The main changes associated with DVT are
shown in Table 5.4. The lower limb is examined
for possible thrombosis much more frequently
DIAGNOSIS OF DEEP VEIN
than the upper limb, although the features
THROMBOSIS
described are also applicable to the arm veins.
Clinical diagnosis of DVT is inaccurate and
clinical scoring systems, such as the Wells Score Compressibility
(Table 5.3),14 have been introduced to stratify As noted above, a normal vein is easily compressible
risk more accurately. In addition measurements with only mild to moderate pressure from the
of serum D-dimer can be used to further refine transducer, so that the lumen is completely
the selection of patients more likely to have a obliterated. A vein filled with thrombus will be
DVT who will benefit from an ultrasound scan.15 held open (Fig. 5.11), although it must be remem-
Patients with a low probability for DVT should bered that fresh thrombus has the consistency of
have a D-dimer estimation. If this is negative, jelly, so that it can be compressed to some extent
they are highly unlikely to have a DVT and do by strong pressure.
not require scanning; if the D-dimer is positive,
or the patient has an intermediate or high prob- Appearances of the vein and the
ability score for DVT, then a scan should be vein lumen
performed. D-dimer levels are less useful in The lumen of a normal vein is usually anechoic
patients who have recently undergone surgery as and, on colour Doppler, the whole lumen of the
false positives are more common. vein should be filled with colour, particularly on
The diagnosis of normal or thrombosed veins augmentation of flow. Although fresh thrombus
is based on the compressibility of the veins, the is anechoic, or hypoechoic, it becomes increasingly
appearance of the veins and the changes which echogenic as it matures. In addition fresh throm-
bus has a tendency to expand the vein and make
it look rounder and fuller than a normal vessel.16
Table 5.3 Pretest probability for deep vein This is accentuated at the upper end of the
thrombosis (DVT)
thrombus where the patent lumen above the clot
Active cancer (treatment ongoing, or within may be relatively poorly filled with blood due to
previous 6 months) +1 the distal obstruction by the thrombus (Fig. 5.12).
Paralysis, paresis, or recent immobilisation Fresh thrombus is not particularly adherent
of the lower extremity +1
to the vein wall, so that some blood may be seen
Recently bedridden for more than 3 days,
or major surgery within 4 weeks +1 around the periphery of the clot in the vein on
Localised tenderness along distribution
of deep venous system +1
Entire leg swollen +1
Calf swelling by more than 3 cm when Table 5.4 Signs of deep vein thrombosis
compared with the asymptomatic leg +1
(measured 10 cm below tibial tuberosity) • Absent or reduced compressibility
Pitting oedema (greater in symptomatic leg) +1 • Thrombus in the vein: static echoes, incomplete
Collateral superficial veins (non-varicose) +1 colour fill-in, expansion of the vein
Alternative diagnosis as likely or greater than • Static valve leaflets

that of DVT –2 • Absent flow on spectral or colour Doppler
From Wells et al.14 • Impaired or absent augmentation of flow
In patients with symptoms in both legs, the more • Loss of spontaneous flow and respiratory
symptomatic leg is used. variation
High: > 3; Intermediate: 1–2; Low: ≤ 0. • Increased flow in collateral channels
111
5
Fig. 5.11 Positive compression test, the

thrombosed vein (arrows) does not change calibre
on compression with the transducer.
colour Doppler. Another appearance which may

be seen in early thrombosis is that of a thin tail Fig. 5.12 A vein containing thrombus: low-level
of thrombus extending up the vein from its origin echoes are seen in the clot, the patent lumen above
and lying free in the lumen of the vein (Fig. 5.13). the thrombus is narrower than the thrombosed
Older thrombus becomes increasingly echogenic, segment.
adherent to the vein wall and contracts as it
becomes more organised and fibrotic. This may An exception to the rule that flowing blood is
result in the vein being reduced to a relatively not echogenic is seen in pregnancy, or any other
small echoic structure that may be difficult to situation where there is slow venous flow and a
locate. Alternatively, the thrombus may retract tendency to hyperviscosity. In these individuals
to one side of the vein, resulting in an asymmetric faint, mobile echoes are seen moving up the vein
lumen on colour Doppler. on real-time imaging; these accelerate on augmen-
a b
Fig. 5.13 (a) A small tail of thrombus extending up the vein (arrows) which is not sufficiently large to produce
any obstruction to flow and could be overlooked if visualisation of this area was poor; (b) power Doppler image
112 of the same thrombus showing flow around it.
5
Diagnosis of deep vein thrombosis

tation of flow. These echoes are produced by may show some slow antegrade flow, particularly
clumps, or aggregates of red cells, and do not if collateral channels are adequate, but this does
usually cause any significant difficulties in diag- not show any respiratory variation and the aug-
nosis with modern, real-time, colour Doppler mentation response is damped.
equipment.
Normal valves may be seen moving gently in Augmentation
the currents from blood passing them, particu- Normal venous flow is slow and can be improved
larly in the larger thigh veins (Fig. 5.7). One of by compression distal to the point of assess-
the earliest sites of deep vein thrombus forma- ment.There are various techniques for achieving
tion is in the sinus above a valve cusp, so that this which are discussed further in the section on
apparent rigidity or fixation of a cusp should chronic venous insufficiency, but for the assess-
raise the suspicion of possible early DVT and a ment of possible DVT manual compression of
careful examination of the area should be the calf is usually sufficient. The muscles of the
undertaken. calf are squeezed rapidly and firmly in order to
The walls of a normal vein are smooth and propel blood up the veins. In a normal venous
unobtrusive. Following recanalisation after a DVT, system there will be a rapid rise and fall in the
they become irregular, thickened and echogenic; frequency shift; whereas if there is a thrombosed
calcification may also occur in a small number segment in the veins, this will increase resistance
of cases. to flow with damping, or absence, of the aug-
mentation response (Fig. 5.14). It should be
Spectral Doppler findings remembered that increased resistance to flow
Spontaneous flow and respiratory anywhere in the vein above the point of com-
variation pression will result in impaired augmentation
Even at rest and with some head-up tilt there the thrombus may be above or below the point
should still be spontaneous flow along the vein of examination. Therefore, the demonstration of
which shows some respiratory variation or impaired augmentation should lead to a careful
phasicity, particularly in the proximal leg veins. search for thrombus in that limb; particularly in
This variation is produced by the intra- the calf or iliac segments.The squeeze of the calf
abdominal pressure changes on respiration and muscles should not be violent, or excessive, as
is the opposite of the changes found in the jugular patients will often have tender or painful calves;
vein and arm veins (see above). On inspiration in addition there is a small potential risk of
the diaphragm descends and the intra-abdominal dislodging a fresh friable thrombus, producing a
pressure rises; this results in decreased flow from pulmonary embolus.The risk of this is small and
the leg veins into the abdomen. On expiration reports of this type of event are few.17
the intra-abdominal pressure decreases and flow
from the legs increases. Similarly, if the patient Flow in collateral channels
holds their breath, flow in the leg veins slows When the normal venous channels are occluded,
and may cease until the patient relaxes, when blood may be seen in collateral veins. In the
there is relatively high flow from the legs. acute stage, intramuscular channels will not have
If there is thrombus occluding the vein there developed significantly but increased velocity
will not be any flow detected in the vein lumen and flow may be seen in the two saphenous veins,
at the level of the thrombus. Sometimes throm- or the profunda femoris vein, which provide
bosis is segmental, with a segment of iliac vein or ready-made collateral pathways. Over a period of
superficial femoral vein occluded but with patent several weeks the intramuscular venous channels
veins below this level; there is a higher incidence will develop and these may be apparent on colour
of this in pregnant patients and patients with pelvic Doppler; therefore their presence indicates a
tumours. Patent segments below the thrombus thrombus of some age, rather than fresh throm- 113
5
a
b
Fig. 5.14 (a) A normal augmentation response to squeezing the calf; there is a rapid rise and fall in the velocity
of blood past the transducer. (b) Abnormal augmentation, with damping of the response as a result of thrombus
impeding the flow of blood up the vein.
bus, unless there has been rethrombosis in a thrombosed veins show that some 64–75% of
segment of clearing clot. veins will recanalise completely, or in part, by
1 year after thrombosis,19 although valvular
Distinction of acute from chronic incompetence will be found at some level in the
thrombus majority of these.20 The remaining veins will
The features which suggest older, rather than show varying degrees of recanalisation, with a
fresh, thrombus are given in Table 5.5. How- thickened irregular wall around an uneven
ever, it is not always possible to define the age of lumen; or remain as fibrotic, permanently
a thrombus and, in these cases the management occluded structures. Abnormal collateral venous
of the patient must be based on the clinical channels will develop in the soft tissues around
picture. any segments which are significantly obstructed
Fresh thrombus is hypoechoic or anechoic. It for any length of time.
is not attached to the wall around the whole
circumference of the vein but if it fills the vein, Upper limb and jugular vein
the vein is a little expanded.14, 18 Increased flow thrombosis
may be detected in the profunda femoris vein, or The same principles apply to examination of
saphenous veins. As thrombus matures it becomes the upper limb and neck veins. Lack of com-
increasingly echogenic and starts to contract as pressibility of the deep veins of the arm and neck
it becomes organised. Longitudinal studies of and/or absence of flow on colour or power
Doppler are diagnostic of thrombosis.The larger,
more proximal veins, such as the axillary and
Table 5.5 Distinction between acute and subclavian, cannot be compressed due to their
chronic thrombus
location; diagnosis of thrombosis in these vessels
Acute Chronic will therefore depend on careful assessment using
colour or power Doppler. Indirect signs of throm-
Anechoic or hypoechoic Increasingly echogenic
Expansion of the vein Contraction of the vein
bosis include loss of respiratory phasicity or
Some compression Incompressible
cardiac variation, which indicates proximal occlu-
possible sion and are useful if central vein (innominate or
Thrombus ‘tail’ in lumen Clot adherent around superior vena cava) thrombosis is suspected.
the wall of the vein
Respiratory phasicity can be modified by asking
Absent or minimal Collateral channels in
collaterals the tissues the patient to breathe deeply, hold their breath
114 or perform a Valsalva manoeuvre. Comparison
5
Problems and pitfalls in the diagnosis of deep vein thrombosis

with the other side may be helpful, assuming that taken with the examination and machine settings,
this is normal. as well as with the selection of an appropriate
Baarslag et al21 compared colour Doppler transducer.
with venography and found 82% sensitivity and Dual superficial femoral veins may be over-
82% specificity for the diagnosis of upper limb looked unless they are actively sought with trans-
DVT; 63% of the patients who had thrombosis verse scanning. If they are not recognised, then
had an associated malignant disease and in 14% one component may be patent and seen on colour
of those with thrombosis this was associated Doppler, whereas the other component may
with an in-dwelling central venous catheter in contain thrombus and be overlooked (Fig. 5.15).
patients without malignant disease. There is a Similarly, non-occlusive thrombus may be
low risk of clinically significant pulmonary missed if the vein is not seen adequately. If there
embolus from upper limb DVT; in one series of is only a small amount of thrombus in the vein
65 patients with arm vein thrombosis, none of then good flow signals will be obtained on
the patients were found to have symptomatic spectral and colour Doppler and the presence of
pulmonary emboli.22 the thrombus may not be recognised (Fig. 5.13).
This is particularly important in obese or oede-
matous legs.
PROBLEMS AND PITFALLS
The calf veins are multitudinous in number
IN THE DIAGNOSIS OF DEEP
and variable in their anatomy. Even with a care-
VEIN THROMBOSIS
ful, patient, time-consuming examination it is
Some of these have been discussed already; difficult to exclude completely the presence of
however, the value of ultrasound as a technique a small segmental thrombus in a calf vein or
for the diagnosis of DVT depends on the operator muscular sinus. In a mobile patient with a little
performing a careful, complete examination, being calf tenderness or swelling this is not a problem,
aware of potential pitfalls and recognising when
a less than adequate examination has been per-
formed. The main problem areas which should
be remembered are shown in Table 5.6.
The essential requirement for a satisfactory
examination is good ultrasound access to the
veins of the limb. Many patients with a possible
diagnosis of DVT have swollen or oedematous
legs; this situation is aggravated if the patients
are also obese. If visualisation is poor then it is
possible to miss significant thrombus unless the
situation is recognised and appropriate care is
Table 5.6 Problems and pitfalls in the diagnosis

of deep vein thrombosis
• Swollen/oedematous/fat legs
• Dual thigh and popliteal veins
• Non-occlusive thrombus
• Segmental calf vein thrombus
• Segmental iliac vein thrombus Fig. 5.15 Dual superficial femoral vein segments;
• Pregnant patients the more posterior segment (arrows) is thrombosed
and could be overlooked. 115
5
as the body’s normal thrombolytic mechanisms that any subsequent extension can be appre-
will probably clear this. However, in a patient ciated. In addition, insertion of a caval filter
who is immobile following surgery or a stroke, a might be considered and it is important to know
small segmental calf thrombus indicates that the if access is possible from the groin through the
clotting cascade has been activated and there is iliac veins. Once a filter has been inserted, the
a possibility that this small thrombus may increase subsequent patency of the cava and iliac veins
in size, resulting in a significant, occlusive throm- can be assessed using ultrasound (Fig. 5.16).26
bus.Therefore a follow-up scan should be consid- During pregnancy several factors are present
ered in these patients in order to identify any which increase the risk of thrombosis. These
progression of thrombus from the calf. A study include changes to the coagulation system and
by Labropoulos et al23 reviewed 5250 patients; physiological changes to venous flow in the leg
isolated calf vein thrombus was found in 4.8% veins due to a combination of hormonal effects
(282 limbs in 251 patients). In these patients, and pressure from the enlarging uterus.27 Some
variable patterns of involvement of the calf veins of the issues relating to the ultrasound diagnosis
were demonstrated with the soleal veins involved of thrombosis associated with pregnancy have
in 20% of cases, gastrocnemius veins in 17%, already been discussed.There is also an increased
peroneal veins in 15% and the posterior tibial tendency to develop segmental proximal throm-
veins in 12%; in 64% of these positive cases, boses in the iliac and upper femoral veins. This
only a single vein group was involved. is more common on the left side,28, 29 perhaps
The accuracy of Doppler in the detection of reflecting the additional potential compression
asymptomatic thrombus is less impressive than from the right common iliac artery, which crosses
that for symptomatic thrombus,4, 24 and the tech- the left common iliac vein just beyond the aortic
nique is therefore inadequate as a screening tool bifurcation. If isolated iliac thrombosis is suspected
for the detection of asymptomatic thrombus. and the ultrasound examination is less than ade-
This is probably because asymptomatic thrombi quate, then consideration should be given to
are more likely to be small and non-occlusive; in
addition, there is a higher incidence of distal
thrombi in the calf veins, which may be more
difficult to demonstrate with ultrasound.3
The external and common iliac veins may not
be demonstrated in their entirety due to obesity
or overlying bowel gas. Care must be taken to
exclude segmental iliac vein thrombosis, espe-
cially if this is a possibility following pelvic surgery;
however, it is very rare for iliac thrombosis not
to include the common femoral vein.25 The
internal iliac veins are difficult to assess but any
thrombus arising in these, which extends into
the common iliac vein and significantly impedes
blood flow, may be suggested by an impaired
augmentation response in the femoral veins, or
loss of respiratory variation on deep breathing or
panting. However, non-occlusive thrombus
which is insufficient to produce this effect may
be overlooked; transvaginal scanning may be of Fig. 5.16 A caval filter in place. Note the change in
value in difficult cases. It is important that the colour due to the alteration in direction of flow in
116 proximal extent of any thrombus is defined so relation to the transducer.
5
Accuracy in relation to other techniques

further imaging with magnetic resonance (MR), a
or contrast venography.27 Patients who have
undergone caesarean section will have a higher
risk of developing a DVT.
Other causes of leg swelling,

pain or tenderness
Unlike venography, ultrasound allows examina-
tion of other structures in the pelvis and leg.
Other pathologies may be seen which account
for the patient’s symptoms of a swollen, or pain-
ful, tender leg; these are given in Table 5.7. It is
important to remember that, even if a ruptured
popliteal cyst is seen (Fig. 5.17), or a superficial
thrombophlebitis is demonstrated (Fig. 5.18),
the deep veins must still be examined carefully,
as a coexistent DVT may otherwise be overlooked.
Labropoulos et al30 demonstrated popliteal cysts b
in 3% of asymptomatic individuals, rising to
10% of patients with symptoms of possible DVT
and 20% of patients with painful knees.
Langsfeld et al31 found popliteal cysts in 3% of
patients being examined for possible DVT, 7%
of those with cysts had a coexisting DVT.
ACCURACY IN RELATION
TO OTHER TECHNIQUES
Despite these potential problems, ultrasound is
a good non-invasive method for the diagnosis of Fig. 5.17 (a) A popliteal (Baker’s) cyst behind the
knee joint; (b) another patient with a ruptured
symptomatic DVT, especially between the lower
popliteal cyst resulting in fluid tracking down through
popliteal region and the groin.3 The key to its
the calf.
value in any given department is that the sono-
graphers must not only be well trained in the
technique, but must also be able to recognise further measures, such as venography or a repeat
an inadequate examination so that appropriate scan, can be arranged. Should venography be
required to clarify areas of doubt, this can be
focused on the area of concern identified at the
Table 5.7 Other causes of leg swelling, pain or ultrasound examination and only a limited
tenderness examination may be required.
Many studies have shown that, in comparison
• Popliteal (Bakers) cysts
to venography, ultrasound is an accurate tech-
• Haematoma/muscle injury
nique for the diagnosis of symptomatic DVT in
• Superficial thrombophlebitis
• Iliac nodes/pelvic masses
the femoropopliteal segments, even in the absence
• Arteriovenous fistula
of colour Doppler.3 Used alone, compression is
• Lymphoedema
an accurate method for detecting DVT, with
sensitivities of 89% and specificity of 100% 117
5
markedly and reached a similar level to that
obtained in the upper part of the limb36 (Table
5.8). In another study,35 32% of studies of calf
veins were inadequate; if these were excluded
then ultrasound showed 93% sensitivity, 98%
specificity and 97% accuracy for the diagnosis of
lower limb DVT.
In a review of outcomes following negative
femoropopliteal ultrasound examinations, Gottlieb
and Widjaja37 showed that only 0.7% of cases
developed a subsequent pulmonary embolus, they
also reviewed 1797 similar patients reported in
the literature and noted that only four (0.2%) of
these had developed a pulmonary embolus
following a negative ultrasound examination of
Fig. 5.18 Thrombosis of the superficial veins is easily
the thigh area in patients symptomatic for DVT.
recognised but the deep veins should still be examined. It is important to draw a distinction between
the accuracy of ultrasound for the diagnosis
of symptomatic thrombus and asymptomatic
being reported for proximal thrombosis,32 and thrombus. The results for the latter are less good
sensitivities of 86–92% and specificities of as, almost by definition, asymptomatic thrombus
96–100% for careful examination of the calf will be non-occlusive in many cases and there-
veins.33 The additional use of colour Doppler fore easier to miss. Weinmann et al noted an
allows very accurate diagnosis of DVT, particu- overall sensitivity in six reported series of only
larly in the femoropopliteal segments. With the 59% for proximal thrombus, although the
development of colour Doppler techniques, specificity was 98%.3 In addition, asymptomatic
further studies have shown the value of ultra- thrombus may be small, or involve one or only a
sound and that the calf veins can be examined few calf vein segments. A further review by Wells
satisfactorily in most cases (Table 5.8).34, 35 The of 17 screening studies in orthopaedic patients
need for an adequate examination must be showed a sensitivity of 62%, specificity of 97%
emphasised. In one study, the initial results in and a positive predictive value of 66% in those
the calf were significantly less accurate than the studies which had been carried out with an
results for the femoropopliteal segment, but adequate scientific method.38
when the examinations were reviewed and only The continuing developments with MR
those which were technically adequate were imaging (MRI) and multislice computed tomo-
considered, the overall accuracy improved graphy (CT) mean that it is now feasible to
Table 5.8 Results of Doppler ultrasound in the diagnosis of symptomatic deep vein thrombosis
Author No. of patients Sensitivity (%) Specificity (%) Regions studied

Rose et al 36 75 96 100 Iliac and femoropopliteal
75 92 100 Femoropopliteal only
75 73 86 Calf only (all studies)
45 95 100 Adequate calf vein studies
Baxter et al34 40 100 100 Femoropopliteal
40 95 100 Calf only
Theodorou et al35 136 93 98 Femoropopliteal
118
5
Recurrent varicose veins and chronic venous insufficiency

consider using these for the diagnosis of DVT. responses in the tissues. These changes result
Several authors have suggested that performing in the development of varicose veins, varicose
a CT scan of the pelvis and upper legs in patients eczema and, ultimately, varicose ulceration.
undergoing CT pulmonary arteriography for Treatment options include standard varicose vein
pulmonary embolus is a satisfactory way to surgical techniques, pressure stockings, dressings
confirm or exclude the presence of significant and, more recently, venous reconstruction tech-
proximal thrombus in the leg and pelvic niques.The pattern of damaged and incompetent
veins.39,40 However, this technique would not veins can be defined using Doppler ultrasound
be practical for the assessment of all cases of to examine the deep and superficial veins in order
possible DVT and considerations relating to to identify thrombosed or partially recanalised
radiation dose and contrast injection would need veins. Incompetent venous segments, together
to be taken into account. Similarly, MR veno- with incompetent perforating veins, can be
graphy is also of some value41 as it shows not mapped out and appropriate surgical or medical
only the thrombus in the lumen of the vein as techniques applied. Approximately 1% of the
a filling defect but can also show thrombus population will have venous leg ulceration at
directly due to the methaemoglobin within it; in some point in their lives,2 and up to 22% will
addition it also shows the perivascular inflam- have evidence of chronic venous insufficiency.5
matory reaction to acute thrombosis.42 As with Diagnosis and assessment of primary varicose
CT, MR venography is not practical or suitable veins has traditionally been based on clinical
for initial assessment of all patients with possible assessment in conjunction with hand-held
DVT, although incidental findings of DVT in Doppler devices but it has been shown that a
abdominal and pelvic examinations can easily formal colour Doppler assessment prior to surgery
be recognised and research into its role is will alter the proposed operative procedure in
continuing. a number of cases. Mercer et al44 showed that
relying on clinical assessment and hand-held
Doppler would have resulted in inadequate, or
RECURRENT VARICOSE VEINS
inappropriate operations in 24% of patients
AND CHRONIC VENOUS
undergoing surgery for primary varicose veins.
INSUFFICIENCY
However, applying this principle to all cases of
The venous system of the lower limb is relatively primary varicose veins would result in a heavy
fragile and easily damaged by a variety of insults workload, so some consideration needs to be
including thrombosis, trauma and inflammation. given to patient selection and scanning only those
Previous thrombosis may not clear completely, in whom there are incomplete, or conflicting
resulting in chronic obstruction and damage to clinical findings.45
the valves. In limbs affected by DVT, 69% had Recurrence of varicose veins after surgery or
at least one segment of incompetent vein which sclerotherapy may occur.Three main patterns of
was more likely to occur in the previously recurrence have been described.46 A patent long
thrombosed segment.43 This damage results in saphenous vein may be present, suggesting that
loss of the protective action of the valves so that it has been missed at the time of the operation.
a continuous column of blood is present between Small collateral veins along the line of the long
the heart and the tissues of the calf, ankle and saphenous vein may enlarge to reconstitute the
foot. In the erect position this may extend over path of the vein (Fig. 5.19). Finally, drainage
1.25 m and the hydrostatic pressure exerted on can occur through venous collaterals which pass
the tissues interferes with the circulation of along a variety of courses remote from the normal
blood in the capillaries, the transfer of nutrients line of the vein. Colour Doppler is useful to assess
and waste matter between blood and the tissues the pattern of recurrence, so that appropriate
and may also promote local inflammatory surgical intervention may be planned.47 119
5
a
b
Fig. 5.19 Collateral channels at the saphenofemoral junction on colour Doppler (a) B-scan image; (b) colour
Doppler image.
Technique of examination flow. Incompetent valves will allow reverse flow

The patient is best examined standing, or with a back through them after forward flow has ceased
large degree of head-up tilt if the couch can be (Fig. 5.20), whereas competent valves will stop
elevated, otherwise inadequate pressure will be any reverse flow. Pressure cuffs that can be
exerted on the valves to test their competence inflated and deflated rapidly can be used to
and misleading measurements will be obtained. produce a similar effect and produce a more
As the examination may be time-consuming – standardised stimulus than manual compression.49
particularly if both legs are being examined – it They can also be used to compress a segment of
is useful if the patients have some means of leg in order to squeeze out the venous blood and
supporting themselves, such as a handle or rail then released suddenly so that any incompetent
on the wall; this enables them to stand in segments will show up by reversed filling from
reasonable comfort with their weight on the leg above. Alternatively, proximal compression may
that is not being examined and with slight flexion be applied to induce reverse flow. Getting the
of the leg under examination. Alternatively, they patient to perform a Valsalva manoeuvre will
may support themselves by holding the side of also show incompetent segments but there are
the ultrasound machine. It is useful if they are two disadvantages to this technique. First, the
asked to stand on a low plinth, as this makes effect will only demonstrate reverse flow as far as
examination of the popliteal and calf regions less the first competent valve, so that any incom-
uncomfortable for the examiner. petent segments below this will not be demon-
Various techniques can be used to assess strated. Second, it is quite difficult to explain to
competence or incompetence of a venous many patients the exact nature and method for
segment.48 The most convenient method for performing a Valsalva. Asking the patient to blow
general assessment is to squeeze firmly the into a high-resistance spirometer circuit can
120 patient’s calf, or lower thigh, to promote forward produce the desired sudden increase in intra-
a
5
Recurrent varicose veins and chronic venous insufficiency

b
Fig. 5.20 An incompetent segment of superficial

c femoral vein showing forward (blue) and reverse flow
(red); (c) the spectral tracing with reverse flow below
the baseline lasting for approximately 3 s.
abdominal pressure and is easier for many together and blood in the venous segment settles
patients to understand. In some patients reflux under the influence of gravity. Reflux should not
will be seen simply with inspiration. be confused with the reversed flow which occurs
Reflux can be defined as reverse flow occur- with turbulence, particularly in the common
ring after the cessation of forward flow. It is femoral vein and popliteal veins. The difference
generally held to be significant if it lasts for more is usually apparent on colour Doppler, and
than 0.5 s,50, 51 although the time taken for reflux turbulence is seen on spectral Doppler as reverse
to cease does not correlate particularly well with flow occurring at the same time as forward flow
the severity of reflux as measured by air (Fig. 5.21).
plethysmography.52 Shorter periods of reversed The examination begins in the groin, where
flow may be seen in normal veins and represent the common femoral vein, profunda femoris
the short period as the valve cusps come vein and saphenofemoral junction are identified 121
5
The patency and competence of the deep and
superficial veins of the thigh are then assessed
down to the level of the knee. Whilst examining
the long saphenous vein the presence of incom-
petent perforators should be sought (Fig. 5.22),
especially if the vein becomes incompetent at a
level below the saphenofemoral junction. These
can be identified most easily by scanning down
the vein transversely whilst applying recurrent
compression to the calf or lower thigh and look-
ing for outward flow with colour Doppler. The
commonest of these perforating veins is in the
Fig. 5.21 Turbulence in a vein showing both red lower thigh at the level of the junction of the
and blue signal in the lumen on colour Doppler and middle and lower thirds and is called the mid-
simultaneous forward and reversed flow on the thigh perforator vein (Fig. 5.2). The use of
spectral display. tourniquets may help clarify difficult cases but
this is not usually required with colour Doppler.
and assessed. If there is a history of previous The patient is then turned so that the popliteal
venous surgery the details are sometimes uncer- region can be examined with the knee partially
tain, or even wrong, and the region of the flexed.The veins in the popliteal fossa are assessed
saphenofemoral junction should be examined and the saphenopopliteal junction is examined.
carefully to assess the type of surgery, whether it The level of the saphenopopliteal junction should
was successful and whether there are any signifi- also be noted, especially if this is not in the
cant collaterals, or recanalised segments which expected location. As with the saphenofemoral
are incompetent. The loss of the normal smooth junction, recurrence after surgery can alter the
curve of the long saphenous vein as it passes anatomy and pattern of flow so that care is
laterally and deeply towards the common femoral needed in defining the situation.
vein is suggestive of previous surgery with subse- Examination of the calf veins may also be
quent recanalisation or collateral formation. performed, although the findings tend to be
a b
Fig. 5.22 (a) An incompetent calf perforating vein

with a diameter of approximately 5 mm passing
through the superficial fascia; (b) colour Doppler
shows flow passing from the deep to the superficial
veins.
122
5
Saphenous vein mapping

more variable and their significance more diffi- tion is required in order to identify perforating
cult to interpret. Incompetence may be seen in veins and superficial branches communicating
a similar fashion to that demonstrated more with the main vein, as these must be ligated
proximally. Sometimes the veins appear dilated during the operation to stop arteriovenous
and it is felt that they should be incompetent, fistulae developing.
but it is very difficult, or impossible, to induce
significant forward flow in the vessels, or any Technique
subsequent reflux. Any incompetent calf perfo- The examination is performed with the patient
rator veins should also be sought using colour standing, if possible, as this produces distension
Doppler, looking for outward flow from the deep of the vein, allowing easier location due to
to the superficial systems (Fig. 5.2).The anatomy dilatation and a better estimation of the calibre
and function of the calf veins and the calf perfo- of the vessel. If the patient is unable to stand
rator veins may have important implications for they can be examined sitting with their legs over
the development of varicose changes, and this the side of the couch; if this is not possible then
area is the subject of continuing research. they can be assessed lying supine with a low-
If necessary, varices can be traced proximally pressure tourniquet applied in order to produce
in order to identify the point of communication distension of the superficial veins.
with the deep or superficial segments. This is One of the problems associated with this
usually best done with the transducer at right examination is that ultrasound gel gums up
angles to the line of the vein being followed, fibre-tipped markers, making it impossible to
judicious compression of lower varices will show mark the course of the vein on the skin, or the
the course of the veins on colour Doppler and location of perforators. In order to avoid this
confirm the presence of reflux, where appro- problem the skin should not be covered with gel
priate. Care must be taken not to compress in the normal manner but the gel should be
superficial veins with excessive transducer pres- applied to the transducer and this then placed in
sure when following the veins. the region of the saphenofemoral junction. Once
the vein has been located the transducer is
aligned along its course, the skin is marked over
SAPHENOUS VEIN MAPPING
the vein at the lower end of the transducer. The
The long saphenous vein is the preferred con- transducer is then moved so that its upper end is
duit for arterial bypass grafting in the coronary on the skin mark, aligned along the vein, and a
arteries and lower limb. If there is any doubt further mark put at the location of the vein at the
concerning the suitability of the vein for the new position of the lower end of the transducer.
procedure, ultrasound can be used to assess the The course of the vein is followed down the
calibre and available length of the vein. Ideally limb, with skin markings being made at each
the vein should be more than 3–4 mm wide for transducer length. Care must be taken in the
much of its length and more than 2 mm at the calf, where the long saphenous vein has two main
ankle if a long, femorodistal graft is being components: the anterior branch usually passes
considered.53 The aim of the examination down to the front of the lateral malleolus and is
depends on the surgical procedure being the larger component, with the posterior branch
contemplated. If the vein is to be removed for a running posteriorly to the posteromedial aspect
coronary artery or reversed lower limb arterial of the calf. Some operators prefer to mark out
graft, then the examination can be limited to the vein with the patient supine as this approxi-
confirming the presence of the vein and mates better to the position during surgery.
assessing its calibre over the required length. If Once the main marks have been applied, the
an in situ lower limb arterial graft is to be location of the saphenofemoral junction, together
performed then a much more detailed examina- with any other perforator veins, dual segments 123
5
and tributaries, can be identified and marked. recognise that the technique has some limitations
This is usually easier to achieve by scanning and that several pitfalls exist.
transversely along the line of the long saphenous Ultrasound also provides a non-invasive tech-
vein with regular augmentation of flow from nique for the investigation of patients with chronic
squeezing the calf. The calibre of the vein is venous disease, or recurrence of varicose veins
measured in the transverse plane, taking care following surgery, enabling an accurate assess-
not to compress the vessel with pressure from ment of the pattern of incompetence or recur-
the transducer. rence to be established and allowing an appro-
priate surgical approach to be developed. It is
also of benefit in the assessment of patients with
CONCLUSIONS
primary varicose veins, especially if there is
Providing care and attention are paid to exami- uncertainty following clinical examination.
nation technique then colour Doppler ultrasound The long saphenous vein can be assessed for
is a reliable method for the diagnosis of DVT in its suitability as a bypass conduit for arterial or
symptomatic patients.The technique has become coronary artery bypass procedures. In addition,
the first-line investigation for DVT in many ultrasound provides a method for examining the
centres, allowing any subsequent venography to central veins prior to central venous line inser-
be restricted to the area of doubt or concern on tion if problems are anticipated in locating a
the ultrasound examination. It is important to suitable channel for line insertion.
REFERENCES
1. Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. relationship to varicose veins. Surg Gynecol Obstet
A population-based perspective of the hospital 1952; 85:537–542.
incidence and case-fatality rates of deep vein 10. Corrales NE, Irvine A, McGuinness CL, et al.
thrombosis and pulmonary embolism: the Worcester Incidence and pattern of long saphenous vein
DVT study. Arch Intern Med 1991; 151:933–938. duplication and its possible implications for
2. Callam MJ, Ruckley CV. The epidemiology of recurrence after varicose vein surgery. Br J Surg
chronic venous disease. A textbook of vascular 2002; 89:323–326.
medicine. London: Arnold; 1996:562–579. 11. Linton RR. The communicating veins of the lower
3. Baxter GM. The role of ultrasound in deep vein leg and the operative technique for their ligation.
thrombosis. Editorial. Clin Radiol 1997; 52:1–3. Ann Surg 1938; 107:582–593.
4. Weinmann EE, Salzman EW. Deep vein thrombosis: 12. Giacomini C. Osservazioni anatomische per service
a review. New Engl J Med 1994; 331:1630–1641. allo studio della circulazioni venosa delle extremitá
5. Phillips GWL, Paige J, Molan MP. A comparison of inferiori. Torino: Tip V Vercillino; 1873.
colour duplex ultrasound with venography and 13. Burihan E, Baptista-Silva JCC. Anatomical study of
varicography in the assessment of varicose veins. the small saphenous vein (saphena parva): types of
Clin Radiol 1995; 50:20–25. termination. Phlebology 1995; 10(suppl 1); 57–60.
6. Da Silva A, Widmer LK, Martin H, et al. Varicose 14. Wells PS, Anderson DR, Bormanis J, et al. Value of
veins and chronic insufficiency: prevalence and risk assessment of pretest probability of deep-vein
factors in 4376 subjects of the Basle Study II. Vasa thrombosis in clinical management. Lancet 1997;
1974; 3(2):118–125. 350:1795–1798.
7. Gordon AC, Wright I, Pugh ND. Duplication of the 15. Ilkhanipour K, Wolfson AB, Walker H, et al.
superficial femoral vein: recognition with duplex Combining clinical risk with D-dimer testing to rule
ultrasonography. Clin Radiol 1996; 51:622–624. out deep vein thrombosis. J Emerg Med 2004;
8. Quinlan DJ, Alikhan R, Gishen P, et al. Variations in 27:233–239.
lower limb venous anatomy: implications for US 16. Hertzberg BS, Kliewer MA, DeLong DM, et al.
diagnosis of deep vein thrombosis. Radiology 2003; Sonographic assessment of lower limb vein
228:443–448. diameters: implications for the diagnosis and
9. Basmajian JV. Distribution of valves in femoral, characterization of deep venous thrombosis.
124 external iliac and common iliac veins and their Am J Roentgenol 1997; 168:1253–1257.
17. Perlin SJ. Pulmonary embolism during compression
33. Atri M, Herba MJ, Reinhold C, et al. Accuracy of

5
References
US of the lower extremity. Radiology 1992; sonography in the evaluation of calf deep vein
184:165–166. thrombosis in both postoperative surveillance and
18. Zwiebel WJ, Priest DL. Colour duplex sonography symptomatic patients. Am J Radiol 1996;
of extremity veins. Semin Ultrasonogr CT MR 1990; 166:1361–1367.
11:136–137. 34. Baxter GM, Duffy P, Partridge E. Colour flow
19. Rosfors S, Eriksson M, Leijd B, et al. A prospective imaging of calf vein thrombosis. Clin Radiol 1992;
follow-up study of acute deep venous thrombosis 46:198–201.
using colour duplex ultrasound, phlebography and 35. Theodorou SJ, Theodorou DJ, Kakitsubata Y.
venous occlusion plethysmography. Internat Angiol Sonography and venography of the lower
1997; 16:39–44. extremities for diagnosing deep vein thrombosis
20. Franzeck UK, Schalch I, Jager KA, et al. Prospective in symptomatic patients. Clin Imaging 2003;
12-year follow-up study of clinical and haemodynamic 27:180–183.
sequelae after deep vein thrombosis in low-risk 36. Rose SC, Zwiebel WJ, Nelson BD, et al.
patients (Zurich study). Circulation 1996; 93:74–79. Symptomatic lower extremity deep venous
21. Baarslag HJ, van Beek EJ, Koopman MM, et al. thrombosis: accuracy, limitations and role of colour
Prospective study of color duplex ultrasonography duplex flow imaging in diagnosis. Radiology 1990;
compared with contrast venography in patients 175:639–644.
suspected of having deep venous thrombosis of 37. Gottlieb RH, Widjaja J. Clinical outcomes of
the upper extremities. Ann Intern Med 2002; untreated symptomatic patients with negative
136:865–872. findings on sonography of the thigh for deep vein
22. Mustafa S, Stein PD, Patel KC, et al. Upper thrombosis: our experience and a review of the
extremity deep venous thrombosis. Chest 2003; literature. Am J Roentgenol 1999; 172:1601–1604.
123:1953–1956. 38. Wells PS, Lensing AW, Davidson BL, et al.
23. Labropoulos N, Webb KM, Kang SS, et al. Accuracy of ultrasound for the diagnosis of deep
Patterns and distribution of isolated calf deep vein vein thrombosis in asymptomatic patients after
thrombosis. J Vasc Surg 1999; 30:787–791. orthopaedic surgery. A meta-analysis. Ann Intern
24. Davidson BL, Elliot CG, Lensing AWA. Low accuracy Med 1995; 122:47–53.
of colour Doppler ultrasound in the detection of 39. Loud PA, Katz DS, Bruce DA, et al. Deep venous
proximal leg vein thrombosis in asymptomatic high- thrombosis with suspected pulmonary embolism:
risk patients. Ann Intern Med 1992; 117:735–738. detection with combined CT venography and
25. Rose SC, Zwiebel WJ, Miller FJ. Distribution of pulmonary angiography. Radiology 2001;
acute lower extremity deep venous thrombosis 219:498–502.
in symptomatic and asymptomatic patients. 40. Lim KE, Hsu WC, Hsu YY, et al. Deep venous
J Ultrasound Med 1994; 13:243–250. thrombosis: comparison of indirect multidetector
26. Smart LM, Redhead DN, Allan PL, et al. Follow-up CT venography and sonography of lower
study of Gunther and LGM inferior vena cava filters. extremities in 26 patients. Clin Imaging 2004;
J Intern Radiol 1992; 7:115–118. 28:439–444.
27. Chan WS, Ginsberg JS. Diagnosis of deep vein 41. Fraser DG, Moody AR, Davidson IR, et al. Deep
thrombosis and pulmonary embolism in pregnancy. venous thrombosis: diagnosis by using venous
Thromb Res 2002; 107:85–91. enhanced subtracted peak arterial MR venography
28. Polak JF, Wilkinson DL. Ultrasonographic diagnosis versus conventional venography. Radiology 2003;
of symptomatic deep venous thrombosis in 226:812–820.
pregnancy. Am J Obstet Gynecol 1991; 165:625–629. 42. Froehlich JB, Prince MR, Greenfield LJ, et al.
29. Macklon NC, Greer IA, Bowman AW. An ultrasound ‘Bull’s-eye’ sign on gadolinium-enhanced magnetic
study of gestational and postural changes in the resonance venography determines thrombus
deep venous system of the leg in pregnancy. presence and age: a preliminary study. J Vasc Surg
Br J Obstet Gynaecol 1997; 104:191–197. 1997; 26:809–816.
30. Labropoulos N, Shifrin DA, et al. New insights into 43. Markel A, Manzo RA, Bergelin RO, et al. Valvular
the development of popliteal cysts. Br J Surg 2004; reflux after deep vein thrombosis: incidence and
91:1313–1318. time of occurrence. J Vasc Surg 1992; 15:377–384.
31. Langsfeld M, Matteson B, Johnson W, et al. 44. Mercer KG, Scott DJ, Berridge DC. Preoperative
Baker’s cysts mimicking the symptoms of deep duplex imaging is required before all operations
vein thrombosis: diagnosis with venous duplex for primary varicose veins. Br J Surg 1998;
scanning. J Vasc Surg 1997; 25:658–662. 85:1495–14977.
32. Cronan JJ, Dorfman GS, Scola FH, et al. Deep 45. Kent PJ, Weston MJ. Duplex scanning may be used
venous thrombosis: US assessment using vein selectively in patients with primary varicose veins.
compression. Radiology 1987; 162:191–194. Ann R Coll Surg Engl 1998; 80:388–393.
125
5
46. Stonebridge PA, Chalmers N, Beggs I, et al. Vascular Surgery/International Society for
Recurrent varicose veins: a varicographic analysis Cardiovascular Surgery clinical classification of
leading to a new, practical classification. Br J Surg chronic venous insufficiency. J Vasc Surg 1994;
1995; 82:60–62. 19:1001–1007.
47. Bradbury AW, Stonebridge PA, Callam MJ, et al. 51. Labropoulos N, Tiongson J, Pryor L, et al. Definition
Recurrent varicose veins: assessment of the of venous reflux in lower-extremity veins. J Vasc
saphenofemoral junction. Br J Surg 1994; Surg 2003; 38:793–798.
81:373–375. 52. Rodriguez AA, Whitehead CM, McLaughlin RL,
48. Allan PL. The role of ultrasound in the assessment et al. Duplex-derived valve closure times fail to
of chronic venous insufficiency. Ultrasound Q 2001; correlate with reflux flow volumes in patients with
17:3–10. chronic venous insufficiency. J Vasc Surg 1996;
49. Markel A, Meissner MH, Manzo RA, et al. 23:606–610.
A comparison of the cuff deflation method with 53. Leopold PW, Shandall A, Kupinski AM, et al. Role of
Valsalva’s maneuver and limb compression in B-mode venous mapping in infrainguinal in situ vein
detecting venous valvular reflux. Arch Surg 1994; arterial bypasses. Br J Surg 1989; 76:305–307.
129:701–705.
50. Iafrati MD, Welch H, O’Donnell TF, et al. Correlation
of venous non-invasive tests with the Society for
126
6
The aorta and inferior vena
cava
Paul L. Allan
Doppler examination in the abdomen is associated Second, longer-pulse repetition intervals are
with specific problems which are not encountered required to allow the sound to travel the greater
in peripheral vascular examinations, and these distances; this also limits the size of Doppler
are particularly relevant to examinations of the shift which can be measured as a result of the
aorta, inferior vena cava and their associated Nyquist limit (see Ch. 1). Operators should
vessels. therefore seek to minimise the scan depth and
Respiratory motion and cardiac pulsation use the highest-frequency transducer compatible
impair the examination, but getting the patient with adequate visualisation.
to suspend respiration for any length of time
results in relative hypoxia and subsequently
THE AORTA
increased respiratory movement. It is therefore
better to scan as much as possible during quiet Anatomy
respiration, asking the patient to hold their The aorta enters the abdomen at the level of T12
breath only for short periods in order to obtain and runs down the posterior abdominal wall to the
a spectral trace. In most cases only two or three left of the midline, with the inferior vena cava to
cardiac cycles are needed for assessment. its right side. It divides into the common iliac
Many vessels will always seem to be orien- arteries at the level of L4, which is about the level
tated at right angles to the scan plane, especially of the iliac crests. Para-aortic nodes are distributed
with sector or curved linear transducers. Different anteriorly and on both sides of the vessel.
angles of approach and repositioning of both the The abdominal aorta gives branches to the
transducer and the patient may be required in abdominal organs and to the abdominal wall.
an attempt to improve the Doppler angle. The parietal branches to the abdominal wall are
Bowel gas is also a problem as it can obscure not usually large enough to be seen regularly
a vessel, or produce distracting motion artefacts using colour Doppler and will not be considered
as it bubbles past; scanning after an overnight fast further. The visceral branches (Fig. 6.1) supply
may improve the situation, as may an injection the liver, kidneys, adrenal glands, gonads, spleen,
of hyoscine. It has been suggested that patients bowel and pancreas. The vessels to the adrenals
should receive bowel preparation as for an enema, and gonads are also usually too small to be seen
but this author feels that this is not usually justified reliably on ultrasound; the renal, hepatic and iliac
for the small advantage it may occasionally confer. arteries are covered elsewhere in greater detail.
Abdominal Doppler examinations are performed The splanchnic arteries supply the bowel and
on vessels which lie more deeply than the associated organs. The coeliac trunk (Fig. 6.2)
peripheral vessels and this has several conse- arises from the anterior aspect of the aorta just
quences. First, lower-frequency transducers are after it has entered the abdomen. The trunk is
used and this limits the size of Doppler shift only about 1 cm long and divides into three
which will be obtained for a given velocity. branches: the common hepatic artery, the splenic 127
6
The aorta and inferior vena cava
usually apparent on ultrasound.The artery ascends
Coeliac trunk in the lesser omentum as the proper hepatic artery,
in company with the portal vein and common
Hepatic bile duct, to the porta of the liver, where it divides
artery Splenic
artery into right and left hepatic arteries. The splenic
artery passes to the left and runs along the
Right renal Left renal superior margin of the body of the pancreas to
artery artery
the hilum of the spleen. It has a tortuous course
Superior and an arterial loop may be mistaken for a small
mesenteric Gonadal cyst in the pancreas if the situation is not recog-
artery arteries
nised; colour Doppler allows quick identification
of the true nature of the ‘cyst’. The right gastric
Inferior
mesenteric
artery arises from the splenic artery but is not
artery usually seen on ultrasound.
The superior mesenteric artery (Fig. 6.3) arises
1–2 cm below the coeliac trunk and supplies the
small bowel and colon to the distal transverse
colon. The superior mesenteric vein is seen on
the right side of the upper portion of the artery
and can be followed to its confluence with the
Fig. 6.1 The abdominal aorta and its major branches.
splenic vein, forming the portal vein. The indi-
vidual branches of the superior mesenteric artery
are not usually seen clearly on ultrasound. The
inferior mesenteric artery (Fig. 6.4) arises from the
anterior aorta about 3–4 cm above the bifur-
cation and runs inferiorly to the left side of the
Fig. 6.2 Transverse view of the coeliac axis showing

the splenic artery on the right and the hepatic artery
on the left.
artery and the left gastric artery. The common

hepatic artery passes to the right over the head of
the pancreas, where it gives off the gastro-
duodenal artery which can be seen passing Fig. 6.3 Longitudinal scan showing the origin of the
inferiorly between the head of the pancreas and coeliac axis superiorly and the superior mesenteric
the margin of the duodenum; the other branches artery just below this. Colour Doppler shows aliasing
128 from this segment of the hepatic artery are not and a tissue bruit around a stenotic coeliac axis origin.
6
The aorta
a arteries may arise from the superior mesenteric
artery or other arteries in the region.
Scanning technique
Aorta
The upper abdominal aorta can nearly always be
examined through the left lobe of the liver; the
coeliac trunk and superior mesenteric arteries
are also visible from this approach. A 3 or 5 MHz
transducer is used depending on the build of the
patient. The patient should fast for 8 h prior to
the examination, for two reasons: first, fasting
will improve visualisation of the aorta and its
branches; second, splanchnic blood flow will be
b in the basal fasting state, rather than the dynamic
postprandial state.
If the aorta is the main object of the inves-
tigation it is followed distally to its bifurcation.
The vessel should be scanned both longitudinally
and transversely, taking note of the overall diameter,
the presence of any aneurysmal dilatation and
any para-aortic masses or pathology. If visuali-
sation from an anterior approach is impaired,
then scanning in a coronal plane through the
right lobe of the liver will allow the upper aorta
to be visualised; scanning in a coronal oblique
plane from a left posterolateral approach can
provide a view of the mid- and lower aorta, together
with the bifurcation. The calibre of the vessel is
measured from the outer aspect of the vessel wall,
Fig. 6.4 (a) Transverse scan showing the inferior
ideally during systolic expansion. The systolic
mesenteric artery lying adjacent to the aorta (arrow);
anteroposterior diameter is the easiest and most
(b) colour Doppler image showing the inferior
repeatable measurement to make and this is
mesenteric artery and vein (arrows).
therefore used for follow-up of aneurysm patients.
It is important to ensure that the true antero-
posterior diameter is measured, particularly in
aorta. The inferior mesenteric vein may be seen ectatic, tortuous arteries, as oblique measure-
on the left of the artery but diverges as it passes ments will result in falsely high measurements.
up to join the splenic vein. Colour Doppler and spectral Doppler are used
Several variations in the anatomy of the to assess any potential disturbances of flow
splanchnic arteries are well recognised.The most which may result from atheroma or dissection.
important one in relation to ultrasound is the origin
of the right hepatic artery from the superior Splanchnic arteries
mesenteric artery. Occasionally the coeliac trunk The coeliac trunk and its main branches are
is absent with its branches arising separately examined using colour and spectral Doppler.
from the aorta; the left hepatic artery may arise The main trunk is short but it is directed towards
from the left gastric artery and accessory hepatic the transducer so that an excellent Doppler angle 129
6
is achieved. The proximal hepatic and splenic may occur, with a waveform similar to that seen
arteries, together with the superior mesenteric in the lower limb arteries (Fig. 6.5).2
artery, are often orientated almost at right angles The main abnormalities affecting the aorta
to the scan plane with an anterior approach are atheroma, aneurysm, dissection and para-
(Fig. 6.2), so that some experimentation with aortic masses. Atheroma can affect the aorta and
points of access is required to get acceptable produce stenosis (Fig. 6.6), or occlusion; aortic
Doppler angles.The hepatic artery is followed to disease, unless severe is usually overshadowed
the right and the gastroduodenal artery can be clinically by symptoms arising from the peripheral
identified beside the head of the pancreas. The or the coronary arteries. Sometimes there is
proper hepatic artery is traced towards the porta uncertainty as to whether aortic disease seen on
where it divides into the right and left hepatic arteriography is clinically significant. In these
arteries. The origin of the superior mesenteric cases velocity ratios taken from above and at the
artery is examined (Fig. 6.3) and the vessel traced
as far distally as it remains visible. Firm pressure
a
with the transducer may help in displacing bowel
gas from in front of the vessel, but care must be
taken not to compress the artery and produce a
spuriously high Doppler shift. Colour Doppler is
used to identify any abnormal areas of flow,
including ‘visible bruits’, or tissue vibrations, which
may be seen in cases of severe stenosis. Power
Doppler is of less value in the abdomen than in
peripheral vessels as arterial pulsation, respi-
ratory movement and bowel gas motion can all
cause marked motion artefacts which obscure
the signal from the vessel.
The inferior mesenteric artery is sometimes
difficult to locate. It can be found by scanning
transversely up from the bifurcation and it may b
be identified just to the left of the aorta, 2–4 cm
above the bifurcation (Fig. 6.4).
Normal and abnormal findings

Aorta
The calibre of the normal aorta varies with the
age, sex and build of the patient, being larger in
men, older patients and tall patients. The calibre
also varies with the level in the abdomen. Goldberg
et al found an average diameter of 22 mm above
the renal arteries, 18 mm just below the renal
arteries and 15 mm above the bifurcation.1 The
normal Doppler waveform in the aorta also varies
with location. In the upper aorta there is a narrow, Fig. 6.5 (a) The aortic waveform in the upper
well-defined systolic complex with forward flow abdomen showing diastolic flow. (b) The aortic
during diastole; below the renal arteries the waveform above the bifurcation with absent diastolic
diastolic flow is much reduced and above the flow and a waveform similar to that seen in the lower
130 bifurcation it is absent, or reversed diastolic flow limb arteries.
6
The aorta
scanning transversely with the ultrasound beam
at right angles to the long axis of the vessel, in
order to ensure a true anteroposterior measure-
ment. It is also important to locate the upper
and lower margins of the aneurysmal segment,
particularly in relation to the renal arteries. If
these cannot be identified with certainty it should
be remembered that the main renal arteries usually
arise from the aorta 1–2 cm below the superior
mesenteric artery, and this vessel can therefore
be used as an approximate marker for the renal
vessels.
Colour and spectral Doppler may show
Fig. 6.6 Colour Doppler image of a stenosis in the turbulent flow within the aneurysm, or indeed
middle aortic segment showing significant turbulence very slow flow with very little forward movement
and a peak systolic velocity of 3.7 m s-1. of the blood. However, examination of normal-
calibre vessels below the aneurysm will show
rapid reconstitution of the waveform as the
stenosis can be used to assess the degree of pressure wave is constrained by the narrower
haemodynamic compromise. However, accurate calibre. Doppler can also be used to confirm
criteria are not as fully developed for aortic renal blood flow, particularly after surgery,
stenosis (Fig. 6.6), as is the case for carotid and if there is any question that this has been cut
peripheral Doppler examinations, but in one off.
study3 a peak systolic velocity ratio of 2.8 corre- Ultrasound is used to monitor the rate of increase
lated (86% sensitivity, 84% specificity) with in size of the aneurysm over time. However, it
aortoiliac stenoses of >50% diameter reduction should be remembered that the calibre of the aorta
and a ratio of 5.0 showed some correlation (65% increases slowly with age and this should be
sensitivity, 91% specificity) with stenoses >75%. borne in mind when assessing the significance of
If the stenotic area is clearly seen, a direct any alteration in the size of smaller aneurysms.
measurement of diameter or area stenosis may Measuring any change in the calibre of the adja-
be obtained; colour or power Doppler is of value cent normal aorta may be helpful in assessing
in defining the margins of the residual lumen. the significance of any change in the diameter of
An aneurysm of the abdominal aorta can be an aneurysm.The main complication arising from
defined as an increase of the anteroposterior an aneurysm is leakage (Fig. 6.7) or rupture.
diameter over 3 cm, or a localised increase of 1.5 Ultrasound can be used to identify any retroperi-
times the diameter of the adjacent normal aorta. toneal haematoma which would indicate a leak
Aneurysms may extend into the abdomen from but computed tomography (CT), if available,
the thoracic aorta, or may arise within the provides a more accurate overall assessment of
abdominal aorta, usually affecting the infrarenal the situation, providing the patient’s condition is
segment. Aneurysms are nearly always true sufficiently stable to allow investigation. Rarely,
aneurysms secondary to degeneration of the an aortocaval or aortoduodenal fistula may
connective tissue in the vessel wall. Occasionally develop; high-volume pulsatile flow in the inferior
mycotic aneurysms secondary to infection, or vena cava is seen on Doppler in cases of caval
pseudoaneurysms secondary to trauma, may be fistula (Fig. 6.8).
seen. Ultrasound diagnosis is normally straight- Screening for aortic aneurysms in men over
forward; the cardinal measurement is the antero- 60–65 is beneficial in terms of reducing
posterior diameter, which is best obtained by mortality;4, 5 approximately 4% of men and 1% 131
6
thoracic aorta extending into the abdomen (Fig.
6.9). Rarely, it may originate in the abdominal
aorta, or result from trauma.The aorta is usually
dilated to some extent but dissection can occur
in the presence of a normal-calibre aorta.The flap
may be visible depending on its orientation in
relation to the ultrasound beam, and if a dissection
is suspected the aorta should be examined from
several different approaches in an effort to show
the flap. Spectral and colour Doppler will show
the presence and character of any flow in the
true and false lumens and, even if a flap is not
Fig. 6.7 Transverse image of a leaking aortic visible, the different flows in the two channels may
aneurysm with a haematoma visible (arrows). be apparent on Doppler; reversed flow may be
seen in the non-dominant channel due to
compression in systole; if one of the channels is
of women over 50 have been shown to have an thrombosed then the appearances can be a little
aorta more than 3 cm in diameter. However, confusing. Doppler can also be used to assess
surgery is not normally considered in incidental, blood flow in the major branches supplying the
asymptomatic aneurysms less than 5.5 cm in bowel, liver, kidneys and lower limbs.8 Clevert et
diameter. al9 reported on the role of ultrasound in the
Ultrasound can also be used in the follow up assessment of a series of 68 dissections, 25 of
of patients who have had endovascular repair of which involved the aortic and iliac segments. For
aortic aneurysms. Although contrast enhanced the 13 aortic dissections the sensitivity for colour
CT is considered the gold standard technique,6 Doppler was 85%, power Doppler 85% and B-
Doppler ultrasound with echo enhancing agents flow 98%; for the 12 iliac dissections, the sensi-
has been shown to be a useful method for assess- tivity for colour Doppler was 67%, power Doppler
ment and may show leaks that have not been 75% and 98% for B-flow when compared with
apparent on CT.7 the reference techniques (a mix of CT, magnetic
Dissection of the abdominal aorta nearly always resonance angiography and digital subtraction
results from the extension of a dissection of the angiography).
Fig 6.8 Colour Doppler image of an

aortocaval fistula in a patient with
an aneurysm. The spectral Doppler
gate is on the fistula and the
spectral display shows a turbulent
signal which is largely off the
spectral range at these settings.
132
a
6
The aorta
Fig 6.9 Longitudinal view of the abdominal aorta
b
showing a dissection flap.
Splanchnic arteries
Blood flow in the superior and inferior mesen-
teric arteries varies depending on whether the
patient is fasting, or has recently eaten. In the
fasting state the flow is consistent with a relatively
high-resistance vascular bed with low diastolic
flow. Following the ingestion of food there is a
reduction in the peripheral resistance of the
mesenteric vessels, resulting in increased diastolic
flow, together with an increase in peak systolic
velocity (Fig. 6.10).
The main indication for examination of blood
flow in the splanchnic arteries is the investigation Fig. 6.10 Flow in the superior mesenteric artery (a)
of possible intestinal ischaemia. One population before and (b) after food.
study10 showed a prevalence of 17.4% for mesen-
teric artery stenosis in a population with a mean makes the assessment of possible gut ischaemia
age of 77 years. Of the patients with mesenteric difficult.The demonstration of stenosis of two of
artery stenosis, 86% had isolated coeliac artery the three splanchnic arteries is strongly suggestive
stenosis (Fig. 6.3), 7% had combined coeliac of the diagnosis and, in the appropriate clinical
and superior mesenteric artery stenosis, 5% had situation, the demonstration of severe stenosis in
isolated superior mesenteric artery stenosis and one vessel with occlusion of another is also
2% had coeliac axis occlusion; however none of supportive of the diagnosis. Colour Doppler is
those affected had symptoms of intestinal of value in identifying the abnormal segment
ischaemia. The usual indication for ultrasound (Fig. 6.3), although care must be taken not to
examination is possible subacute or chronic mistake a high shift resulting from disease with
ischaemia, as significant acute ischaemia presents the high shift from normal velocity flow, which is
as an acute abdomen and is managed accord- seen due to the low Doppler angle resulting from
ingly. The splanchnic circulation is capable of the orientation of the coeliac trunk and proximal
developing multiple collateral channels and this superior mesenteric artery to the ultrasound beam. 133
6
In addition, colour Doppler may show a visible these conditions but may also arise following
tissue bruit if there is a significant stenosis. The liver transplantation.
proximal 2–3 cm of the vessels is the most
common site for disease and a peak velocity of Other applications
more than 2.8 m s-1 in the superior mesenteric The nature of para-aortic masses can be clarified
artery correlates well with a stenosis of more than using colour Doppler and masses can be distin-
70% diameter reduction with a sensitivity of 92% guished from aneurysms.
and a specificity of 96%; whereas the equivalent An aorta which is prominent but of normal
peak systolic velocity for the coeliac axis is 2 m s-1 calibre in a thin patient, or the aorta in a patient
(87% sensitivity, 80% specificity).11 Indirect signs with a marked lumbar lordosis, may be mistaken
of intestinal ischaemia include oedema of the clinically for a mass or an aneurysm. Ultrasound
mucosa and bowel wall and also reduced peri- can confirm the normal calibre of the vessel and
stalsis. In severe cases gas bubbles may be seen the lack of pathology in these patients.
in the portal vein flow; these produce a charac- Blood flow in the coeliac and mesenteric arteries
teristic popping sound on spectral Doppler (Fig. is also responsive to a variety of pharmacological
6.11). agents such as glucagon and somatostatin, or
The problems associated with the diagnosis pathological states such as cirrhosis and Crohn’s
of mesenteric ischaemia are illustrated by the disease.13 Doppler ultrasound can show flow
fact that 18% of patients over 60 years without changes associated with these situations and may
symptoms of mesenteric ischaemia have been hold some promise for the future in the assessment
shown to have significant disease on Doppler.10, 12 of disease activity, or response to treatment.
This emphasises the need to assess the findings Contrast enhanced colour Doppler has been
in the light of the clinical situation. used with some success in the postprocedural
Aneurysms of the hepatic and splenic arteries assessment of endovascular aneurysm repairs, in
may occur. Splenic artery aneurysms are some cases detecting leaks not seen on CT angio-
associated with acute pancreatitis and trauma; graphy.14, 15 However, ultrasound is less good for
hepatic artery aneurysms can be associated with detecting structural problems with the stent graft,
such as stent distortion or fracture.16 Unenhanced
colour Doppler is less reliable when compared
with CT angiography and digital subtraction arte-
riography. One suggested strategy is to alternate
CT and ultrasound in the follow-up of stent graft
patients16 as this could result in substantial cost
reduction, reduced radiation exposure and
reduction in the risk of contrast induced
nephrotoxicity.
THE INFERIOR VENA CAVA

Anatomy
The inferior vena cava is formed by the
confluence of the common iliac veins at the level
Fig. 6.11 The spectral display from the portal vein of of the 5th lumbar vertebra and runs cranially to
a patient with severe ischaemia and a transjugular the right of the midline. It passes through the
intrahepatic portosystemic shunt (TIPS) showing the diaphragm at the level of the 8th thoracic vertebra
characteristic high intensity echoes from bubbles of and enters the right atrium. In the embryo there
134 gas in the blood. is a complex arrangement of venous sinuses
6
The inferior vena cava

which form during embryogenesis, and several
of these contribute to the inferior vena cava; this
means that there are many variations of anatomy
which may be seen. The most common variation
is that the common iliac veins continue cranially
as paired ‘venae cavae’, with the left component
crossing to join the right side at the level of the
left renal vein. Many other variations have been
recorded; these are more easily assessed using
contrast-enhanced CT than ultrasound, but may
cause some confusion if they are seen during an
ultrasound examination and not recognised.
Technique
The inferior vena cava can be examined using Fig. 6.12 The inferior vena cava in the upper
the techniques described for the abdominal aorta. abdomen showing the variations in flow which occur
However, in the supine position the vessel may with respiration and cardiac activity; the various
be narrow in the anteroposterior plane and difficult components of the waveform in the inferior vena
to define. Scanning transversely with colour cava and hepatic veins are described in detail in
Doppler and using the aorta as a guide may allow Chapter 7 and Figure 7.12.
localisation of the vein in these circumstances; or
elevation of the leg(s) by an assistant will increase
flow and the calibre of the vein, thus making it whereas in fluid overload the cava is dilated and
more visible.The calibre of the cava will vary with there is cardiac periodicity detectable down into
the state of hydration of the patient. In a well- the iliac veins.
hydrated patient it will be distended, whereas in One of the most common indications for specific
a dehydrated patient it will be collapsed, narrow examination of the inferior vena cava is to assess
and more difficult to visualise. Excessive trans- whether thrombus from a pelvic or lower limb
ducer pressure applied in an effort to disperse deep vein thrombosis has extended up into the
bowel gas will also compress the cava, therefore cava. Thrombus may fill the lumen of the cava
a balance must be sought in order to visualise and even produce some expansion of the vessel;
segments of the vessel in some patients. alternatively, a tongue of thrombus may be seen
lying free in the lumen, extending up towards
Normal and abnormal findings the right atrium (Fig. 6.13).
Flow in the inferior vena cava is slow and varies Caval filters are inserted in some patients who
with both respiration and cardiac pulsation (Fig. are at risk of pulmonary embolus from more
6.12). On inspiration the diaphragm descends. distal thrombus. There are several types but all
This results in negative pressure in the chest and are inserted in the mid- or lower cava, below the
increased pressure in the abdomen, and blood renal veins. Identifying a metallic, echogenic
therefore flows from the abdomen to the chest; structure within the inferior vena cava above the
the reverse occurs on expiration. Superimposed level of the renal veins may indicate migration of
on this is the more rapid periodicity resulting the filter. There is a small risk that thrombus
from cardiac activity; this is seen particularly in may extend past the filter, or develop at the site
the upper abdomen.The prominence of the caval of the filter. Colour Doppler is a quick and easy
waveform also depends on the degree of hydration method for confirming patency of the cava around
of the patient. A dehydrated patient’s cava will and above the filter.17 The metal struts of the filter
be narrow and difficult to see below the renal veins, can be recognised within the lumen of the cava 135
6
Fig. 6.13 A tongue of thrombus
extending into the upper inferior
vena cava posterior to the left lobe
of the liver in a patient with deep
vein thrombosis.
and colour Doppler, or power Doppler, will show segment of native cava, which had been removed
blood flow past the level of the filter (Fig. 6.14). with the diseased liver. Many surgeons now perform
Renal tumours and hepatocellular carcinoma a ‘piggyback’ technique, where the native cava is
are two tumours which have a tendency to invade left in place, the inferior end of the donor caval
venous channels and, as a result, tumour thrombus segment is oversewn and the upper end anasto-
may extend into the cava (Fig. 6.15) from the renal mosed to the native cava. This results in a post-
or hepatic vein.18 Compromise of the venous operative appearance which can be confusing if
drainage of the liver or kidneys is shown by loss it is not recognised, as there will appear to be
of the normal cardiac and respiratory periodicity two cavae associated with the transplanted liver
of the venous waveform and the tumour thrombus (Fig. 6.16).
may be clearly seen as it extends into the caval Other postoperative problems which may occur
lumen. Some tumours in the retroperitoneum in relation to the cava following transplantation
may compress or directly invade the inferior vena include compression, if the new liver is relatively
cava causing obstruction to the venous return large; distortion of the cava may also occur if
from the lower abdomen and legs. Although the there is relative twisting of the caval channel as a
caudal segments of the inferior vena cava and result of fitting the donor liver into the native
the iliac veins usually remain patent, they will abdomen. In the longer term stenosis may develop
often be dilated, flow will be sluggish or reversed, at the sites of anastomosis. Liver transplantation
the flow profile will be flat, and there will be an is covered further in Chapter 7.
absence of the normal Valsalva response. Rarely, Retroperitoneal and other abdominal masses
intrinsic tumours, usually mesenchymal in origin, may compress or occlude the inferior vena cava.
such as fibrosarcomas or leiomyosarcomas may The situation is usually apparent, especially with
develop in the caval wall, lipomas have also been colour Doppler, which shows the cava entering
reported.19 the mass and becoming narrowed or occluded with
Following liver transplantation the cava should absent flow. Congenital webs can occur, particularly
be assessed to ensure satisfactory flow. The at the upper end of the cava.These may produce
appearances will depend on the type of anastomosis a variable degree of caval narrowing and in some
performed. In the past, the segment of donor cava cases may predispose to hepatic vein thrombosis
136 attached to the new liver replaced the equivalent and a Budd Chiari syndrome.
6
Conclusion
a
Fig. 6.15 Tumour thrombus in the inferior vena cava

from a renal carcinoma (arrows).
b
portocaval shunts are usually side-to-side shunts

in the upper abdomen at the level where the
proximal main portal vein passes close in front
of the cava (Fig. 6.17). A tissue bruit may be
apparent and the shunt is more easily identified
if the liver can be used as a window through to
the point of anastomosis; turning the patient up
onto the left side may facilitate visualisation.
However, these are rarely performed now, having
been replaced by transjugular intrahepatic
portosystemic shunts (TIPS) (see Ch. 7).
CONCLUSION
Fig. 6.14 (a) Real time image of an IVC containing a The aorta and inferior vena cava, together with
filter in the lower segment; (b) colour doppler image their major branches and tributaries, can be
of the same patient with a caval filter. The change in examined in most patients, provided that care
colour reflects the changing Doppler angle as the and attention is spent in finding the best scan
blood flows through the sector scan. plane and ensuring that the system is set appro-
priately for the specific examination concerned,
both in terms of imaging and Doppler settings.
Fistulae involving the cava may rarely occur Helical CT and three-dimensional recon-
spontaneously, often secondary to an aortic struction is rapidly becoming the technique of
aneurysm (Fig. 6.8), or they may be surgically choice for imaging the aorta, particularly if
created as is the case with portocaval shunts. In percutaneous stent/grafts are contemplated.
the case of aortocaval fistulae, colour Doppler may However, ultrasound continues to have a
show a visible tissue bruit with pulsatile flow in significant role in relation to both initial
the cava above the level of the fistula; sometimes diagnosis and follow-up of patients with aortic
the fistula itself is difficult to identify. Surgical or caval disease. 137
6
a
b Fig. 6.17 Colour Doppler image of a portocaval

shunt. The portal vein (arrows) has been surgically
anastomosed to the inferior vena cava.
Fig. 6.16 (a) A real time image of the upper caval

region in a liver transplant patient with a ‘piggyback’
caval anastomosis, showing the native cava
posteriorly and the donor cava just anterior to it;
(b) colour Doppler image of the same anastomosis.
REFERENCES
1. Goldberg BB, Ostrum BJ, Isard HJ. Ultrasonographic 4. US Preventive Services Task Force. Screening for
aortography. J Am Med Assoc 1996; 198:353–358. abdominal aortic aneurysm: recommendation
2. Taylor KJW, Burns PN, Woodcock JP, et al. statement. Ann Intern Med 2005; 142:198–202.
Blood flow in deep abdominal and pelvic vessels: 5. Frame PS. Screening for abdominal aortic aneurysm
ultrasonic pulsed Doppler analysis. Radiology 1985; (Editorial). Br Med J 2004; 329:E311–312.
154:487–493. 6. Raman KG, Missig-Carroll N, Richardson T, et al.
3. De Smet AA, Kitslaar PJ. A duplex criterion for Color-flow duplex ultrasound scan versus computed
aortoiliac stenosis. Eur J Vasc Surg 1990; tomographic scan in the surveillance of endovascular
138 4:275–278. aneurysm repair. J Vasc Surg 2003; 38:645–651.
7. Napoli V, Bargellini I, Sardella SG, et al. Abdominal
13. Perko MJ. Duplex ultrasound for assessment of

6
References
aortic aneurysm: contrast-enhanced US for missed superior mesenteric artery blood flow. Eur J Vasc
endoleaks after endoluminal repair. Radiology 2004; Endovasc Surg 2001; 21:106–117.
233:217–225. 14. Napoli V, Bargellini I, Sardella SG, et al. Abdominal
8. Thomas EA, Dubbins PA. Duplex ultrasound of the aortic aneurysm: contrast-enhanced US for missed
abdominal aorta – a neglected tool in aortic endoleaks after endoluminal repair. Radiology 2004;
dissection. Clin Radiol 1990; 42:330–334. 233:217–225.
9. Clevert DA, Rupp N, Reiser M, et al. Improved 15. Bendick PJ, Bove PG, Long GW, et al. Efficacy of
diagnosis of vascular dissection by ultrasound ultrasound contrast agents in the noninvasive
B-flow: a comparison with color-coded Doppler follow-up of aortic stent grafts. J Vasc Surg 2003;
and power Doppler sonography. Eur Radiol 2005; 37:381–385.
15:342–347. 16. Thurnher S, Cejna M. Imaging of aortic stent-grafts
10. Hansen KJ, Wilson DB, Craven TE, et al. and endoleaks. Radiol Clin North Am 2002;
Mesenteric artery disease in the elderly. J Vasc 40:799–833.
Surg 2004; 40:45–52. 17. Smart LM, Redhead DN, Allan PL, et al. Follow-up
11. Moneta GL. Screening for mesenteric vascular study of Gunther and LGM inferior vena caval
insufficiency and follow-up of mesenteric artery filters. J Intervent Radiol 1992; 7:115–118.
bypass procedures. Semin Vasc Surg 2001; 18. Bissada NK, Yakout HH, Babanouri A, et al.
14:186–192. Long-term experience with management of renal
12. Roobottom CA, Dubbins PA. Significant disease cell carcinoma involving the inferior vena cava.
of the coeliac and superior mesenteric arteries Urology 2003; 61:89–92.
in asymptomatic patients: predictive value of 19. Grassi R, Di Mizio R, Barberi A, et al. Case report.
Doppler sonography. Am J Radiol 1993; Ultrasound and CT findings in lipoma of the inferior
161:985–988. vena cava. Br J Radiol 2002; 75:69–71.
139
7
Doppler ultrasound
of the liver
Myron A. Pozniak
pathology.The Doppler survey may reveal a subtle

hypervascular lesion of which the examiner was
INTRODUCTION
otherwise unaware, or it may display hyper-
Recent ultrasound literature has many reports of vascularity of an observed lesion and awareness
extensive work and success with ultrasound contrast of a lesion’s vascularity often increases diagnostic
medium. Practitioners in Great Britain, Italy, Japan, certainty.The use of colour Doppler in abdominal
South Korea, Canada, and other countries have examinations also helps to differentiate vascular
reported very impressive results in lesion detec- from non-vascular structures. Care must be taken,
tion and characterisation with the application of however, to ensure that equipment settings are
microbubbles.1Unfortunately governing agencies appropriate: if gain, pulse repetition frequency,
across the world have not uniformly endorsed and filtration are not optimised, slow flow can be
these agents and there is limited ability to apply missed in vascular structures or artifactual colour
them in certain countries.This chapter is specifi- can be painted into non-vascular structures.
cally written for those ultrasound practitioners who
do not have access to intravenous ultrasound
GENERAL ASPECTS
contrast.
The standard abdominal ultrasound (US) Some sonologists place considerable emphasis
examination should include a brief, but precise on the measurement of flow velocity, but too great
survey with spectral and colour Doppler.2–5 This a dependence on this parameter may generate a
serves a twofold purpose: first, it adds valuable false sense of security, or lead to diagnostic errors.
haemodynamic information to the evaluation of Numerous systemic factors affect blood flow in
the abdominal organs, in most cases reinforcing abdominal vessels. These include the patient’s
normality, but occasionally revealing an unexpected state of hydration, cardiac output, blood pressure,
finding; second, by consistently integrating vascular compliance, the interval since previous
Doppler into the routine abdominal examination, food, and the haemodynamic effects of medica-
sonologists will continually refine their Doppler tions. These factors affect measured velocities in
skills so that significant haemodynamic abnor- a variety of ways and to varying degrees; so that
malities can be identified quickly and evaluated although the velocity may be above or below the
accurately. Although a cursory Doppler survey normal levels expected in any individual vessel,
of the major vessels may add 2–3 min to an it may not necessarily reflect focal disordered
abdominal examination, regular practice enables haemodynamics. Furthermore, defining flow
the examiner to become more familiar with the within a vessel as normal or abnormal by simply
equipment, more adept at perceiving abnormali- comparing the measured velocity to a predeter-
ties, and more expert in analysing the results. mined normal range is a poor method of estab-
Sometimes altered blood flow may be the lishing a diagnosis, as a few degrees difference in
only abnormal finding to suggest the presence of the angle of insonation or improper angle correc- 141
7
Doppler ultrasound of the liver
tion can significantly change the measured velocity. volume is appropriate for rapid interrogation of
Assigning the proper degree of angle correction a broad area, for example, when ruling out hepatic
may be difficult if the vessel is poorly visualised, artery thrombosis in a liver transplant recipient.
curved, or visualised only over a short segment. If, however, the examiner wants to precisely charac-
Varying the width of the sample volume can terise flow within a vessel and evaluate waveform
be advantageous when examining the abdomen. detail, then the sample volume must be small
If the examiner is screening for vascular patency and placed near the centre of the vessel, thereby
or trying to locate a specific vessel, a large sample interrogating the highest velocity lamina (Fig. 7.1).
a
Laminar flow
Relative velocity
c d
Fig. 7.1 (a) Schematic representation of normal laminar flow. The velocity along the wall of a vessel is slowed
because of drag, therefore the relative velocity is less than that measured at the centre of the vessel. (b) Colour
Doppler image of portal vein flow. A green tag was assigned to high-level velocities towards the transducer and
the scale was set low to allow aliasing. Colour-encoding thus permits an accurate display of numerous lamina.
Note the transition from the slowest velocities, red through orange, green and blue from the periphery to the
centre of the vessel. The actual velocity displayed on a spectral tracing will be critically dependent on sample
volume placement relative to these various velocity lamina. (c) Normal spectral Doppler tracing of portal vein
flow. A small sample volume is placed near the vessel wall. The velocity as interrogated near the wall only
measures approximately 0.11 ms-1. (d) Spectral Doppler tracing of the same portal vein with the sample volume
now placed more centrally interrogating the higher-velocity lamina. The measured velocity is now 0.27 ms-1.
142 Simply changing location of the sample volume is sufficient to cause a twofold change in measured velocity.
7
General aspects
A wide sample volume, by incorporating the the patient fast for 6–8 h prior to an abdominal
slower lamina along the wall together with the examination helps minimise the amount of
faster central lamina, will broaden the spectral abdominal gas, thereby increasing the likelihood
Doppler tracing and mimic turbulence (Fig. 7.2).6 that an appropriate sonographic window will be
There is no specific acoustic window that is ideal available for any particular vessel of interest. In
for all patients and the operator must determine addition, consistently scanning fasting patients
the best approach on an individual basis. This decreases the risk of misinterpreting flow dynamics
usually requires trying multiple windows at varying altered by a nutrient load.
degrees of inspiration. During respiration, the Patient obesity is a severe limiting factor for
upper abdominal organs move back and forth an adequate Doppler examination. Delineation
under the US transducer.When patients are able of anatomical detail is impaired when the exami-
to cooperate, the operator should ask them to nation is conducted at lower frequencies. During
intermittently hold their breath during the US imaging, the operator may need to press
Doppler examination and to breathe gently at firmly to displace some of the adipose tissue and
other times. This improves the colour Doppler position the transducer closer to the area of interest.
image and allows acquisition of longer spectral Such a manoeuvre, however, is not appropriate
Doppler tracings. Patients who are unable to during a Doppler examination as pressure from
hold their breath can pose a significant problem the transducer compresses the underlying organ
and the operator may have to carefully ‘ride’ the and its vasculature, thereby altering flow profiles
vessel in real time as it moves with respiration. and velocities. Compression of an organ or vessel
An experienced sonologist may be able to ‘rock’ with the transducer causes increased resistance to
the transducer back and forth in synchrony with blood flow (especially diastolic) thereby elevating
the patient’s respiration, thus maintaining the the perceived resistance to inflow (Fig. 7.3).
sample volume over the area of interest and
obtaining a longer tracing. If the patient is short Terminology
of breath or unable to cooperate, short segments In relation to Doppler ultrasound of the liver, it
of spectral tracings are all that may be possible. is important to be consistent in the use of terms
The presence of bowel gas is also an obvious relating to blood flow: the term pulsatility refers
impediment to a successful examination. Making to arterial flow, phasicity refers to changes in flow
b
a
Fig 7.2 (a) Spectral Doppler tracing of a portal vein with small, centrally placed sample volume. A thin lamina of
flow is interrogated and, therefore, the displayed tracing shows a narrow range of velocities with a ‘window’
below the tracing. (b) The sample volume has been opened wide to incorporate all velocity lamina across this
portal vein. Note the ‘filling in’ of the spectral tracing resulting in a perception of spectral broadening. 143
7
a
b
Fig. 7.3 (a) Spectral Doppler tracing of an interlobar renal transplant artery. This tracing was obtained with the
transducer gently contacting the lateral abdominal wall. The resistive measures approximately 65%. (b) The same
vessel is now interrogated with moderate pressure applied with the transducer. Note the increase in diastolic
velocity with result and elevation in resistive index to 100%. Pressure applied by the sonologist via the
transducer can increase the resistance to arterial inflow in any organ or vessel.
secondary to respiration, and the term periodicity applied to the portal vein, hepatic artery, and
is recommended when referring to velocity variation the hepatic veins of these patients and indeed in
in the hepatic and portal veins secondary to all RUQ examinations.7 This may reveal flow
cardiac activity. Normal flow in the portal vein alterations caused by inflammatory disease,
towards the liver is properly termed hepatopetal neoplasm, or other disorders which are too subtle
(as in centripetal force, not hepatopedal). Reversed or too small to cause imaging irregularities.
portal vein flow is referred to as hepatofugal (as Alterations in flow profiles and velocities in the
in centrifugal force). hepatic vessels may be the result of either
hepatic or cardiac disease, thereby helping to
differentiate patients needing cardiac evalua-
INDICATIONS
tion from those with liver disease who may
A significant percentage of patients referred for benefit from liver biopsy, or require further
right upper quadrant (RUQ) US typically have hepatic imaging using computed tomography
elevated liver enzymes of unknown aetiology, (CT), magnetic resonance imaging (MRI), or
incidentally detected on screening blood tests. angiography.
Although sonographic imaging of the liver may When portal hypertension is suspected, Doppler
reveal diffuse abnormality or focal disease, the US characterises changes in portal haemo-
majority of studies on these patients are often dynamics and identifies pathways of portosystemic
normal (Table 7.1). Doppler US should be collateralisation.8, 9 Doppler can confirm the
patency of surgical or percutaneous shunts which
have been performed in patients with bleeding
Table 7.1 Indications for Doppler ultrasound of oesophageal varices.10
the liver Identification and differentiation of bland
thrombus from tumour thrombus within the
• Part of the routine examination of the liver and
right upper quadrant hepatic or portal veins by Doppler has significant
• Assessment of portal hypertension implications for medical or surgical treatment
• Pre- and postprocedural assessment of planning.
transjugular portosystemic shunt (TIPS)
procedures
Doppler US plays a key role in the post-
• Postoperative follow-up of liver transplants
operative monitoring of liver transplant recipients,
• Assessment of focal liver disease confirming patency of the portal vein, hepatic
144 artery, and hepatic veins.
a
7
Scan technique
The role of Doppler in the characterisation of
parenchymal liver disease and screening for
hepatocellular carcinoma (HCC) is controversial. MHV
Marked alterations in flow profiles and velocities
can be seen and have been described in the
literature.11–20 It is rare, though, to be able to
precisely pinpoint a specific diagnosis based on
LHV
Doppler findings since there is considerable overlap
in velocity and waveform alterations among
various disease states.21 RHV
SCAN TECHNIQUE
The patient is usually scanned while in a supine
or left lateral decubitus position (Table 7.2).
Depending on vessel orientation and patient body
habitus, the portal vein and hepatic artery are
b
best interrogated by either a subcostal approach
pointing posterocephalad, or a right intercostal
approach pointing medially. Since the portal vein
and hepatic artery travel together in the portal
triad, along with the common duct, these approaches
should satisfactorily interrogate both vessels.
Scanning the left hepatic vein (and occasionally
the middle hepatic vein) is best accomplished
from a substernal approach.The transducer should
be oriented transversely, pointing posterocephalad,
and swept up and down across the liver. For the
right hepatic vein, a right lateral intercostal approach
is used with the transducer pointed cephalad. If
the patient’s liver extends below the costal margin
during inspiration, a subcostal transverse view,
angled cephalad, is useful for the confluence of Fig. 7.4 (a) Colour Doppler transverse view of the
the hepatic veins (Fig. 7.4). liver with a subcostal approach. This colour crow’s
foot appearance of the three hepatic veins confirms
their patency. The direction of flow is away from the
Table 7.2 Principles of the examination
transducer and towards the heart. Right, middle and
• General examination of liver parenchyma and left hepatic vein, RHV, MHV, LHV. (b) Power Doppler
abdomen image in this patient shows the convergence of five
• Colour and spectral Doppler assessment of the hepatic veins (HVs). The presence of accessory
portal vein, superior mesenteric and splenic
veins, together with main intrahepatic portal
branches is very common.
branches
• Colour and spectral Doppler assessment of the
hepatic artery from the coeliac axis to the porta,
Some patients, when asked to hold their
together with main intrahepatic branches breath perform a vigorous Valsalva manoeuvre.
• Colour and spectral Doppler assessment of the This results in increased intrathoracic pressure
main hepatic veins and the upper inferior vena
cava which may impede venous return, affecting flow
profiles and velocities, particularly in the hepatic 145
7
a
veins and inferior vena cava (IVC). This effect
may alter the hepatic vein profile, creating the
perception of hepatic venous outflow obstruction
(HVOO). An attempt should be made to scan
these patients in neutral breath hold to avoid
producing a misleading Doppler tracing.
VASCULAR ANATOMY
AND NORMAL FLOW PROFILES
(Table 7.3) HA
Portal vein PV
The portal vein normally supplies approximately
70% of incoming blood volume to the liver. This
relatively deoxygenated blood comes to the liver
after perfusing the intestine and spleen. It is rich
in nutrients after a meal, and arrives at the liver b
to be processed in the cells of the hepatic sinu-
soids. The portal vein is formed by the
confluence of the splenic and superior mesenteric
veins. It is accompanied by the hepatic artery
and common bile duct to form the portal triad
(Fig. 7.5); this has echogenic margins as it courses
into the liver, due to the paraportal extension of
Glisson’s capsule and the presence of some
perivascular fat. In the liver, these vessels progres-
sively branch to supply the liver segments;
anatomic variations of the portal vein are rare.
Fig. 7.5 Oblique colour Doppler images of the porta
The Couinaud system of segmental liver anatomy hepatis. (a) The hepatic artery (HA) accompanies the
divides the liver vertically along the planes of the portal vein (PV) and bile ducts. With the colour scale
hepatic veins, and horizontally along the planes appropriately set for the slow flow within the portal
of the left and right portal veins. The segmental vein, hepatic arterial flow often reveals colour aliasing
branches of the portal vein enter the centre of during systole. (b) With biliary dilatation colour
the Couinaud segments whose ultrasound Doppler provides clear discrimination between
appearance has been described by La Fortune et vessels and ducts (arrow).
al.22 The typical portal vein flow profile has a
relatively consistent velocity of approximately pulsation is damped by the capillaries of the
20 cm s-1 (± 5 cm s-1) towards the liver (Fig. 7.6a).23 intestine at one end of the portal system and by
The flow velocity is uniform because cardiac the liver sinusoids at the other end. Slight phasicity
may be seen on the portal vein spectral tracing
secondary to patient respiration and a mild degree
Table 7.3 Normal hepatic vessel velocities
of periodicity may be present, due either to
(fasting) retrograde pulsation transmitted from the right
heart via the hepatic vein (A-wave) or to the
Hepatic artery 30–40 cm s-1 systolic, 5–10 cm
s-1 diastolic inflow of blood during hepatic arterial systole.24
Portal vein 15–20 cm s-1 Because these brief pressure surges into the liver
146 transiently increase resistance to portal venous
7
Vascular anatomy and normal flow profiles

a b
Fig. 7.6 (a) Spectral Doppler tracing of normal portal vein flow. The flow velocity of 20 cm s-1 is relatively
uniform and in a hepatopetal direction. (b) Spectral Doppler tracing of a normal portal vein flow. Slight periodicity
is present in this patient’s portal vein tracing. The dip in antegrade velocity coincides with hepatic arterial
systole. Velocity variation may also occur due to pressure change with the hepatic vein A-wave.
inflow, they effect a momentary slowing of ante- individuals. The common hepatic artery originates
grade flow in the normal portal vein (Fig. 7.6b).25 from the celiac artery and after the origin of the
In studies of portal vein flow, Hosoki26 and gastroduodenal artery, it is called the proper
Wachsberg27 reported the presence of some degree hepatic artery; this then enters the liver alongside
of periodicity in 7% and 64% of their respective the portal vein (Fig. 7.5a) where it divides into
normal study populations. Although some perio- left and right hepatic arteries.There are numerous
dicity may be expected in portal vein flow, reversal variants of hepatic artery anatomy.These include
of flow, even briefly, should be considered an accessory vessels which exist in conjunction with
abnormal finding. normal branches of the hepatic artery and replaced
arteries which make up the sole supply of a
Hepatic artery segment or lobe. For example, a replaced right
The arterial blood supply of the liver arises solely hepatic artery arising from the superior mesenteric
from the celiac axis in approximately 76% of artery may be the sole blood supply to the entire
b
PV
SMA
IVC
Fig 7.7 (a) Transverse image of the mid-abdomen at the level of the superior mesenteric artery origin. A tubular
structure is seen coursing from the superior mesenteric artery (SMA) to the right lobe of the liver, between
portal vein (PV) and the inferior vena cava (IVC). (b) Colour Doppler identifies this tubular structure as a vessel.
An arterial signal on spectral Doppler, identification of its SMA origin and a course towards the right lobe of the
liver confirm this to be a replaced right hepatic artery. 147
7
while during diastole, it normally slows to approxi-
mately 10–20 cm s-1 which is normally less than
the velocity of the portal vein flow. The systolic
acceleration time is brisk, typically less than 0.07 s.
Technically, a good way to evaluate relative flow
velocities between the hepatic artery and the
portal vein is to increase the sample volume
size so that both vessels are incorporated into
the same tracing or to swing the sample volume
from one vessel to the other in the same tracing
(Fig. 7.9).
Fig. 7.8 Spectral Doppler tracing of a normal hepatic

Hepatic veins
artery. Systolic upstroke is brisk with acceleration
The hepatic veins are relatively straight, anechoic,
time <0.07 s. Resistive index measures 70%.
tubular structures that converge on the IVC
Velocity at end diastole approximates 19 cm s-1.
approximately 1 cm below its confluence with
the right atrium. The walls of the hepatic veins
right lobe of the liver, a situation occurring in are relatively hypoechoic which helps to differen-
approximately 11–25% of the population.28 In a tiate them from the portal veins in the more
slender patient, colour Doppler US may be able echogenic portal triads. There are no valves in
to identify the replaced right hepatic artery behind the hepatic veins.
the main portal vein as it courses towards the In most people, the right, middle, and left hepatic
right lobe from the superior mesenteric artery veins enter the IVC in a ‘crow’s foot’ configu-
(Fig. 7.7). The other variants occur less ration when viewed in the transverse plane (Fig.
frequently and are more difficult to identify by 7.4).The left and middle hepatic veins may enter
Doppler US. as a common trunk along the left anterolateral
The normal hepatic artery in a fasting patient aspect of the IVC. Approximately 30% of indi-
has a low resistance Doppler flow profile, [about viduals have additional hepatic veins that may be
60–70% resistive index (RI)] (Fig. 7.8). During identified by colour Doppler; a right superior
systole, the velocity is approximately 30–60 cm s-1; anterior segmental vein may be seen draining
a b
Fig. 7.9 Combined spectral Doppler tracing of the hepatic artery and portal vein. By opening the sample volume
wide (a), or rocking the transducer between the hepatic artery and portal vein in the same tracing (b), the
relative velocities between hepatic artery and portal vein can be directly compared. A normal velocity ratio
between hepatic artery and portal vein is present in these patients. The hepatic artery diastolic velocity in a
148 fasting patient is normally equal to or slightly less than portal vein velocity.
7
Vascular anatomy and normal flow profiles

The venous drainage from the central portion of
the liver parenchyma, including the caudate lobe,
empties directly into the IVC and cannot normally
HV IVC be perceived by colour Doppler since these veins
are small and central (Fig. 7.11). This separate
drainage pathway is responsible for the unique
behaviour of the caudate lobe in liver disease,
and for the distinctive enhancement pattern seen
on contrast enhanced CT scans of patients with
hepatic vein thrombosis.
Fig. 7.10 Longitudinal colour Doppler image of the
liver and inferior vena cava (IVC) through the right
The hepatic venous waveform
flank. There is a prominent accessory hepatic vein
The normal hepatic vein waveform is triphasic
(open arrow) from the inferior posterior segment of
the right lobe of the liver (Couinaud segment 6).
as a result of transmitted cardiac activity (Fig. 7.12)
This vein joins the inferior vena cava approximately and is similar to the jugular vein waveform; indeed,
3 cm inferior to the junction of the right middle and the labels on the hepatic vein components have
left hepatic veins (HV). been transposed from the jugular vein pressure
tracing. Most sonographic windows to the liver
demonstrate the hepatic veins so that flow
towards the heart is away from the transducer,
which registers as flow below the baseline but
during right atrial systole blood is forced back
into the liver and is therefore displayed above
the baseline. These directions are best described
as being antegrade (towards the heart) and retro-
IVC
grade (away from the heart).
This complicated tracing has been described
and normal velocity measurements determined
by Abu-Yousef.24 Figure 7.12b shows the hepatic
venous waveform in relationship with an electro-
cardiogram (ECG) tracing, tricuspid M-mode
scan, and atrial and ventricular status.The following
Fig. 7.11 CT scan at the level of the mid-liver in a stages can be identified:
patient with marked fatty infiltration. Note several 1. The most distinctive feature is the retrograde
small hepatic veins (arrow) draining directly into the A-wave, which is the result of right atrial
inferior vena cava (IVC). Because of this unique contraction and coincides with the P-wave on
drainage of the central liver, this segment behaves
the ECG. Since there is no valve between the
differently in patients with main hepatic vein
right atrium and the IVC, a burst of reversed
thrombosis or cirrhosis. The ultrasound appearance
flow travels down the IVC and into the
may be altered. Hypertrophy may be perceived.
hepatic veins, which has a mean velocity of
approximately 18 cm s-1.
into the middle hepatic vein, marginal hepatic 2. At the end of right atrial contraction, flow
veins may drain into the right and left hepatic returns to the antegrade direction as the atrium
veins, and a large accessory right hepatic vein may relaxes. However, on right ventricular systole
be seen entering the IVC several centimetres the tricuspid valve is slammed shut and actually
inferior to the junction of the three main hepatic bulges back into the right atrium, thus creating
veins in approximately 6–10% of people (Fig. 7.10). its own pressure wave, the C-wave, which may 149
7
(a)
(b) 1 2 3 4 5 1 2
A A
Hepatic vein
velocity tracing
C V C
S
R R
ECG P T P
Q S Q S
A E A
F
D
Tricuspid m-mode
Atrial Atrial Atrial

systole diastole systole
Ventricular Ventricular
systole diastole
Fig 7.12 (a) Normal hepatic vein spectral Doppler tracing. (b) Simultaneous tracings of an ECG, hepatic vein
spectral Doppler tracing, and mitral valve M-mode tracing with correlation to atrial and ventricular systole and
diastole. The divisions at the top of these tracing (1–5) correspond to the discussion in the text.
be perceived as a brief pause in the steadily 4. At the end of atrial filling, antegrade velocity
increasing antegrade flow. This C-wave decreases, or may even briefly reverse; this is
coincides with the beginning of ventricular known as the V-wave and has a mean velocity of
systole and occurs immediately after the QRS approximately -1.1 cm s-1. In relation to the
complex on the ECG. ECG, this occurs immediately following the
3. The right atrium continues to dilate and T-wave.
antegrade flow builds to a relatively high 5. As the right ventricle enters diastole, the
velocity of approximately 30 cm s-1. Even- tricuspid valve opens and flow in the hepatic
tually atrial filling approaches completion veins increases in the antegrade direction, as
and antegrade flow starts to slow. This both the right atrium and right ventricle fill.
transition, known as the S-wave, occurs Velocity builds to a mean of approximately
during ventricular systole within 0.15 s of the 22 cm s-1 and this phase is referred to as the
150 QRS complex. D-wave. Eventually, the right heart chambers
7
become filled passively and antegrade flow inspiration with increasing negative intrathoracic
begins to slow. We then return to the A-wave pressure, the S- and D-waves become more
as the atrium again contracts to begin another prominent, while the A- and V-waves are less
cardiac cycle. pronounced and may actually not manifest as
reversed flow (Fig. 7.14).
This waveform is seen in the hepatic veins
and upper IVC in the vast majority of patients.
ASSESSMENT OF DISEASE
However, not all individuals have a similar degree
of periodicity within the hepatic veins. The Portal vein
percentage of patients that manifest an identifi- Portal hypertension
able C-wave is relatively small (Fig. 7.13). The In hepatocellular disease, the sinusoids are
degree of flow reversal of the A-wave and V-wave damaged, destroyed or replaced. As the volume
may vary depending on the patient’s cardiac status, of normally functioning liver parenchyma decreases,
state of hydration, heart rate, and the distance of the resistance to portal venous flow increases,
Doppler interrogation from the heart. In a survey the portal vein dilates, and portal flow decreases
of a population of normal volunteers, a 9% and eventually reverses.8, 30–33 There is an elevation
incidence of a flattened hepatic vein flow profile in portal vein pressure, in excess of normal, by
has been reported.29 5–10 mmHg, resulting in portal hypertension.34
Because the heart is located within the thorax, Use of the ‘congestive index’ has been recom-
pressure changes caused by respiration affect the mended in helping to diagnose portal hypertension.
hepatic vein flow profile. When the patient This index is the ratio of the portal vein cross-
forcefully exhales or bears down against a closed sectional area (cm2) divided by the mean portal
glottis, the elevated intrathoracic pressure resists flow velocity (cm/ s-1), thereby taking into account
antegrade flow, causing the S- and D-waves to portal vein dilatation and decreased flow velocity,
be less prominent. The reversed component of the two physiological changes associated with
flow increases so the A- and V-waves become portal hypertension.35 In normal subjects, this
more pronounced. Conversely, during forced ratio is less than 0.7. Several other indices have
been suggested as useful in the prediction of liver
disease and its severity. These include hepatic
Fig. 7.13 Normal hepatic vein spectral Doppler

tracing. Even though the tracing is acquired within a Fig 7.14 Spectral Doppler tracing in a patient with
few centimetres of the heart, the C-wave cannot be severe alcohol-induced cirrhosis. Portal vein velocity
identified on this tracing. This is, indeed, the tracing has decreased although flow does remain
most often visualised in the hepatic vein and inferior hepatopetal through most of the cardiac cycle. Note
vena cava. Only a small component of retrograde the marked periodicity, however, which coincided
flow is present. with hepatic arterial systole. 151
7
and splenic arteries RIs, modified liver vascular veins, either a ‘cannon’ A-wave or a reversed
index (portal flow velocity/hepatic artery RI), S-wave39,40 (Fig. 7.17).
and portal hypertension index,36
Varices
(hepatic artery RI × 0.69) ×
As portal hypertension progresses and pressure
(splenic artery RI × 0.87)
rises to 15 or 20 mmHg, sufficient pressure
portal vein flow velocity
exists to cause the development of varices.These
As liver disease worsens, the periodicity in the collateral pathways shunt blood from the portal
portal vein may become more pronounced, usually to the systemic circulation. The more common
coinciding with hepatic arterial systole (Fig.
7.15).30,37 Finally, with end-stage liver disease,
continuous hepatofugal flow is observed, usually
with marked periodicity. Blood entering the liver in
the hepatic artery normally passes through the
hepatic sinusoids to the hepatic veins, but with
increasing hepatocellular disease, scarring and
fibrosis, the pathway of least resistance for the arterial
inflow becomes the portal vein with arterial blood
being shunted to the portal vein via vasa vasorum,
or via direct arteriovenous shunting at the level
of the sinusoids.Thus, the hepatofugal flow leaving
the liver in the portal vein is arterial blood
shunted from the hepatic artery38 (Fig. 7.16).
Fig 7.16 Portal vein spectral Doppler tracing in a
Pronounced periodicity may be seen in the
patient with severe hepatocellular damage secondary
portal vein, which does not coincide with hepatic to paracetamol (acetaminophen) overdose. Portal vein
arterial systole. This is usually due to cardiac flow is hepatofugal throughout the cardiac cycle.
disease, such as right ventricular dysfunction or
tricuspid regurgitation, and is caused by a promi-
nent reversed component of flow in the hepatic
Fig 7.17 Spectral Doppler tracing of portal vein flow

Fig. 7.15 Spectral Doppler tracing in a patient with in a patient with severe right ventricular dysfunction
severe alcohol induced cirrhosis. The portal vein flow and tricuspid regurgitation. There is marked periodicity
remains continually hepatopetal but the flow profile in the portal vein waveform. Hepatopetal flow
has developed marked periodicity to the point where decreases, and briefly reverses, coinciding with the
antegrade flow almost stops. This coincides with large regurgitant component of hepatic vein flow
152 arterial systole. during tricuspid regurgitation.
7
channels are the short gastric, left gastric and Short gastric varices coursing between the
coronary veins; recanalised paraumbilical veins; spleen and the greater curvature of the stomach
and splenorenal-mesenteric collaterals (Fig. 7.18). are best imaged via the left flank, using the
Other, less typical, pathways include perichole- enlarged spleen as a window (Fig. 7.19). Left
cystic, iliolumbar, gonadal, haemorrhoidal, and gastric or coronary vein varices running from the
ascending retrosternal veins. Indeed, almost any splenic or portal veins to the lesser curvature of
vein in the abdomen may serve as a potential the stomach are best imaged through the left
collateral to the systemic circulation and may be lobe of the liver (Figs 7.19b and 7.20a). Both
incorporated in a very convoluted shunt.41 sets of varices then converge on the gastro-
Gastroesophageal varix
Paraumbilical varix
Left branch of portal vein
Short gastric veins
Right branch of portal vein
Left kidney
Left gastric vein
Splenorenal
Portal vein varix
Right gastroepiploic vein

3
2
Left gastroepiploic vein
Splenic vein
Superior mesenteric vein
Inferior mesenteric vein
Left colic vein
Sigmoid and rectal veins
Portosystemic shunts
1. Oesophageal
2. Paraumbilical
3. Retroperitoneal 4
4. Rectal Hemorroidal varix
Fig 7.18 Major portosystemic varices encountered in portal hypertension. 153

7
oesophageal junction. From there, blood flow In utero, oxygenated blood flows from the
proceeds upwards through oesophageal varices placenta up the umbilical vein to the left portal
to eventually communicate with the azygous vein vein and through the ductus venosus into the IVC
and the systemic circulation (Fig. 7.20b). Because and right atrium. After birth, this pathway involutes
of the potential lethal risk from spontaneous, and the umbilical vein is represented by the
brisk haemorrhage from these varices, a variety ligamentum teres in the falciform ligament. In
of endoscopic, surgical, or percutaneous proce- portal hypertension, paraumbilical veins in this
dures may be employed to divert blood away ligament can dilate and carry blood from the left
from them.42 portal vein along the anterior abdominal wall to
a b
LGV
SGV
Fig. 7.19 (a) Longitudinal colour Doppler view of the left flank in a patient with portal hypertension. A tangled
web of vessels (arrows) is seen at the splenic hilum and extends cephalad towards the diaphragm (D). These
are short gastric varices coursing from splenic veins to the stomach and from there to the systemic circulation
via esophageal varices. (b) Contrast-enhanced CT of a patient with portal hypertension. Short gastric varices
(SGV) are best imaged from the left flank using the spleen as a sonographic window. The left gastric varix (LGV)
is best imaged through the left lobe of the liver.
b
a
Fig. 7.20 (a) Longitudinal colour Doppler image in the midline of a patient with portal hypertension. A large
tortuous left gastric varix is seen coursing from the region of the coeliac axis towards the gastro-oesophageal
junction. Whereas short gastric varices tend to be a plexus of small vessels, the left gastric varix is typically a
single large tortuous vessel. (b) Coloured shaded surface displays three-dimensional reconstruction of the portal
154 vasculature. The portal venous system is coloured red. The vessel represents a large left gastric varix. Note the
convoluted nature of this vessel as it courses cephalad, eventually branching into multiple oesophageal varices.
7
the umbilical region (Fig. 7.21).43 From the should not be confused with the Rokitansky Aschoff
umbilicus, the blood may pass to the superior or sinuses of hyperplastic cholecystosis. Colour
inferior epigastric veins, or through subcutaneous Doppler is useful to show flow within these vessels;
veins in the anterior abdominal wall, a ‘caput the spectral tracing is that of portal venous flow
medusae’, to reach the main systemic venous (Fig. 7.25). From the gallbladder, subhepatic
system. Because inferior epigastric varices run collaterals communicate with the abdominal wall
just deep to the rectus muscles, they are not and subcostal veins. Haemorrhoidal collaterals
apparent on clinical examination but are easily are not routinely studied by Doppler.
identified by colour Doppler (Fig. 7.22). Patients
with known portal hypertension, who present with Portal vein thrombosis
an umbilical hernia, should undergo imaging Portal vein thrombosis may be completely asymp-
evaluation prior to surgery as the hernia may contain tomatic in patients with cirrhosis; however, more
a dilated varix, rather than bowel (Fig. 7.23). than half of cases present with life-threatening
One advantage of this collateral pathway is that complications, such as gastrointestinal haemor-
it shunts blood away from those varices that can rhage or intestinal infarction.45 Portal vein throm-
cause life-threatening variceal bleeding.44 bosis must be considered when no Doppler signal
Splenorenal-mesenteric collaterals are typically quite is detected within the portal vein. It may be due
large and very tortuous. They are seen in the left to blood clot, or to tumour invasion. However,
flank coursing between the splenic hilum and the the examiner should first review the system set-
left renal vein (Fig. 7.24). Occasionally, this pathway up and re-evaluate scale, gain, and filtration
can continue via gonadal veins into the pelvis. settings (Fig. 7.26). If these are found to be set
Pericholecystic varices can occur in the gall- appropriately and there is still no perceptible
bladder wall and are associated with portal vein flow, the patient should be asked to perform a
thrombosis. US imaging may reveal cystic or Valsalva manoeuvre. This elevates intrathoracic
tubular structures in the gallbladder wall. These and right atrial pressure, transmitting higher
RM
PUV
Fig. 7. 21 (a) Longitudinal colour Doppler image of a patient with portal hypertension. A large vein carries flow
from the left portal vein towards the transducer. It courses along the falciform ligament and then turns caudad
along the interior abdominal wall coursing towards the umbilicus, deep to the rectus muscles (RM). (b) Shaded
surface display three-dimensional reformat of the upper abdominal solid organs and portal venous system.
Flow in the main portal vein courses towards the liver but immediately channels into a large left portal vein
(open arrow). Flow then continues anteriorly towards the anterior abdominal wall and down towards the
umbilicus in a large recanalised paraumbilical vein (pink). 155
7
a pressure to the IVC, hepatic veins and subse-
quently into the liver parenchyma.This increased
pressure causes even greater resistance to portal
PUV
venous inflow and may convert stagnant portal
IEV
flow to hepatofugal flow (Fig. 7.27). One may
also consider the use of an intravenous US
contrast agent to enhance perception of very
slow flow.
Early thrombosis of the portal vein may be
difficult to visualise with US since fresh clot can
be markedly hypoechoic (Fig. 7.28).46 As clot
b matures, it becomes more echogenic and retracts,
allowing partial recanalisation of the portal vein
(Fig. 7.29). Patients with long-standing portal
vein thrombosis may develop collateral flow into
the liver via a lace-like network of veins. This is
known as cavernous transformation of the portal
vein or a cavernoma.47, 48 Grey-scale imaging
alone can seldom visualise these vessels because
PUV of their small size, but colour Doppler reveals a
IVC
web of numerous serpiginous small veins which
typically involve a fairly wide area of the liver
hilum (Fig. 7.30). Spectral Doppler shows
portal flow in the branches of the cavernoma.
Neoplastic invasion
IEGV HCC has the propensity to invade the portal and
hepatic veins. Intravascular tumour is classified
as stage IV disease and is considered unresectable.
EIV
Involvement of the portal vein by tumour may
cause an increase in its cross-sectional area and
a decrease in portal vein flow. Tumour in the
portal vein will receive its blood supply from the
hepatic artery and spectral Doppler of the
‘thrombus’ will show an arterial waveform, which
usually projects in a hepatofugal direction,
Fig. 7.22 (a) Longitudinal colour Doppler image supplying the tumour as it grows out of the liver.
directly over the umbilicus. A recanalized paraumbilical A bland thrombus will not manifest such a tracing
vein (PUV) carries blood towards the umbilicus. on Doppler, so that invasive tumour can be
No caput medussa was present in this patient since
differentiated from bland thrombus and the diag-
flow continues from the umbilical region via the
nosis of stage IV HCC with vascular invasion
inferior epigastric vein (IEV). (b) 3-D CT angiogram in
confirmed (Fig. 7.31).49
a patient with portal hypertension. Note the large
recanalized paraumbilical vein (PUV) coursing towards
the umbilicus. From there flow continues back to the Portal vein aneurysm
systemic circulation via the left inferior epigastric vein Aneurysm of the portal vein has been reported,
(IEGV) to the left external iliac vein (EIV) and the but it is extremely rare. The vein may enlarge to
156 inferior vana cava (IVC). a diameter of 3 cm or larger. Spectral Doppler
a
b 7
PUV
Fig. 7.23 (a) (b) Three-dimensional rendering shaded

c surface display. The portal venous phase of the CT
angiogram in a patient with portal hypertension. A large
recanalised paraumbilical vein (PUV) courses down the
anterior abdominal wall (arrow). In the subumbilical region
this vessel is seen to dilate into an aneurysmal component
of this varix. (c) Photograph of the umbilicus in a patient
with alcoholic cirrhosis and end-stage liver disease.
This rust coloured mass is a herniated paraumbilical varix
that has extended to the rectus sheath and out through
the anterior abdominal wall. This varix is simply following
the anatomic pathway of the original umbilical vein.
S K
Fig. 7.24 (a) Longitudinal colour Doppler image of the left flank. A large varix occupies the space between the
splenic hilum (S) and the left kidney (K). This represents a convoluted splenorenal varix. (b) Three-dimensional
rendering of the portal phase of a CT angiogram. The portal venous system is illustrated in pink. The red vessel
represents a large convoluted splenorenal varix. The kidney and its renal vein are illustrated in yellow. (The spleen 157
has been removed for ease of visualisation of this varix.)
7
PV
Fig. 7.26 Colour Doppler image along the axis of the

Fig. 7.25 Colour Doppler image of the gall bladder porta hepatis in a patient with portal hypertension.
in a patient with portal hypertension and portal vein Colour flow is clearly evident in the hepatic artery.
thrombosis. Numerous vessels are identified in the The portal vein, however, shows no evident flow
gall bladder wall. Spectral Doppler revealed a portal suggesting thrombosis. Evaluation of the pulse repetition
vein waveform as would be expected in varices and frequency (arrow) shows that the value was set quite
not an arterial waveform as would be seen in the high. This effectively suppresses colour display of
cystic artery. slow flow giving the false impression of thrombosis.
should be applied to confirm a portal vein wave- Portal vein gas

form and rule out hepatic artery aneurysm since Gas may be seen in the portal vein and its
the latter carries a much higher incidence of branches in a variety of gastrointestinal disorders,
complications and rupture.50,51 such as sepsis, obstruction with distension, necro-
Fig. 7.27 Spectral Doppler tracings of a patient with end-stage liver disease being considered for liver
transplantation. (a) With neutral breath-hold, flow in this portal vein is barely perceived. It oscillates between
hepatofugal during arterial systole and hepatopetal during arterial diastole. (b) When instructed to forcefully
breathe in and out, this patient’s portal flow became more dynamic. During forced expiration with elevated
intrathoracic pressure there is increasing resistance to hepatic venous outflow forcing portal vein flow to
become hepatofugal. During inspiration with negative intrathoracic pressure, this decreases the resistance to
158 hepatic venous outflow thereby, causing flow to become almost stagnant.
7
Fig. 7.28 Oblique colour Doppler image of the porta
hepatis. The pulse repetition frequency is appropriately
set at a low level. Nevertheless, the trickle of flow is Fig. 7.30 Oblique colour Doppler image of the porta
seen in the extrahepatic portal vein but within the hepatis. A normal portal vein could not be visualised.
liver itself no colour flow is identified. Indeed there Instead there is a plexus of small vessels with
are some low level echoes within the portal vein hepatopetal flow (arrows). After portal vein thrombosis
indicating fresh clot (arrow). this web of small vessels reconstitutes portal flow to
the liver and is known as cavernous transformation
of the portal vein.
tising enterocolitis, infarction or ulceration. reveals sharp bidirectional spikes superimposed

Numerous tiny hyperechoic foci can be seen in on the Doppler tracing of the portal vein.52
the portal vein, flowing into the liver. Since these These spikes do not reflect a higher velocity of
bubbles are moving fairly rapidly, their percep- the air bubble but represent an artefact resulting
tion is improved by increasing scanner temporal from the system being set to display the Doppler
resolution by limiting the field of view to the area shift of red blood cells, so that the much more
of the portal vein and minimising, or turning off, intense echoes from the air bubbles register as a
frame averaging. The spectral Doppler tracing spikes of noise on the tracing (Fig. 7.32).53
Fig. 7.29 (a) Oblique image of the portal vein in a patient with hypercoagulability. Echogenic thrombus is seen
adherent to the portal vein wall (arrows). (b) Colour Doppler image transverse to the portahepatis. Flow is seen
coursing past the partially occluding thrombus in a hepatopetal direction. 159
7
Fig. 7.32 Spectral Doppler tracing of the portal vein in

Fig. 7.31 Spectral Doppler tracing over the portal vein
a patient with pneumatosis intestinalis. The spikes
in a patient with hepatocellular carcinoma (HCC).
present in this spectral Doppler tracing are caused by
An arterial waveform is present in a plug of tissue
air bubbles in the portal vein. These bubbles are
identified extending into the portal vein. This represents
traveling at the same velocity as the rest of the blood
HCC primarily perfused by arterial flow.
in the portal vein. With the spectral gain set for blood,
the intense sound reflection caused by the passing
air bubbles creates spikes of noise.
Hepatic artery
A comparison of hepatic arterial velocities with
those of the portal vein may be used as an indi- conditions (e.g. haemangioma, fatty infiltration),
cator of liver disease. The normal portal vein however, do not perceptibly affect main hepatic
velocity in a fasting patient is about 18 cm s-1, artery or portal vein flow (Fig. 7.34).
the normal systolic velocity in the hepatic artery Hepatic artery resistance has been studied in
is 25–40 cm s-1 and 10–15 cm s-1 in diastole. If several disease states. Alterations in resistance
the waveforms of the hepatic artery and portal may be observed but, to date, have not been shown
vein can be captured simultaneously on a Doppler to be sufficiently specific or sensitive in the diag-
tracing, the normal hepatic arterial diastolic nosis of any one particular condition.54,55 Rapid
velocity will therefore be seen to dip just below onset of oedema or inflammation of the liver may
that of the portal vein (Fig. 7.9). produce a substantial amount of congestion,
Almost all liver disease processes receive their leading to higher resistance to hepatic arterial
blood supply primarily from the hepatic artery. inflow and elevation of the RI. Elevated RI in the
As the process becomes more severe, or involves hepatic artery has been reported as a predictor
a larger area of liver, hepatic artery flow increases. of fulminant and severe acute liver failure.56,57
As liver disease worsens, the portal venous inflow Hypervascular disorders, especially those with
encounters progressively increasing resistance, arterial venous shunting, such as neoplasm, can
resulting in decreased hepatopetal portal velocity. lower the arterial resistance. A tardus parvus
Therefore, if the Doppler examination shows waveform and low resistance flow may also be
hepatic arterial diastolic velocities greater than perceived downstream from significant hepatic
that of the portal vein, the liver parenchyma artery stenosis, or a diaphragmatic crus defect of
should be carefully evaluated to rule out focal or the celiac axis (Fig. 7.35).
diffuse liver disease.8 This finding, however, is non-
specific and can be seen with neoplasm (both Hepatic artery aneurysm
primary and metastatic) and infection (viral, Aneurysm of the hepatic artery is usually
160 bacterial, parasitic or fungal) (Fig. 7.33). Benign extrahepatic and may be congenital or acquired.
Fig. 7.33 (a) Spectral Doppler tracing of the porta

7
a
hepatis in a patient with biopsy proven chemical
hepatitis. This tracing was obtained by a slight change
in orientation of the transducer during the scan,
traversing from portal vein to the hepatic artery.
Hepatic arterial velocities are markedly increased with
low resistive index and high diastolic flow. Portal vein
flow has decreased to the point where it is barely
receptable. (b) Combined hepatic artery and portal
vein tracing at the porta hepatis in a patient with HIV.
In contrast to (a) this tracing was obtained with a
sample volume opened wide, including the flow
profiles of both the hepatic artery and portal vein in
the same tracing. Similar alteration in velocities is
b
again perceived with bounding arterial flow and
markedly diminished portal vein inflow. (c) Similar
combined tracing in a patient with metastatic lung
cancer. The bounding arterial flow is supplying the
rapidly growing tumour metastases. Portal venous
flow is decreased as the liver sinusoids are being
replaced by tumour.
Pancreatitis, trauma, or liver biopsy are the most A clot may eventually develop within the
common aetiologies. Mycotic aneurysms can be aneurysm or pseudoaneurysm (Fig. 7.36). Com-
seen in immunocompromised patients, those with munication may develop from the aneurysm to the
bacterial endocarditis or those abusing intravenous portal vein or hepatic vein converting the aneurysm
drugs. Sonography demonstrates a rounded area to an arteriovenous fistula. Spectral Doppler flow
with swirling flow on colour. An arterial spectral profiles show bounding arterial inflow velocities,
tracing may be perceived but is usually quite swirling turbulent flow within the aneurysm and
distorted due to turbulence. arterialisation of the venous outflow.58,59 161
7
a
b
Fig. 7.34 (a) Transverse image of the right lobe of the liver. A globular but very echogenic mass like infiltrate is
perceived in the inferior aspect of the right lobe. The posteroinferior branch of the right hepatic artery and vein
are seen coursing into this area on colour Doppler. (b) Spectral Doppler tracing of the hepatic artery and portal
vein supplying this mass reveals the velocities are relatively normal. The ratio of flow is not altered between the
two vessels. Biopsy proved this mass to be focal fatty infiltration.
Fig. 7.35 (a) Spectral Doppler tracing of an hepatic artery in a young patient with upper abdominal pain.
The waveform shows a markedly tardus parvus configuration. There is a slow upstroke in systole and very low
resistive index. This indicates inflow compromise. (b) Combined spectral Doppler tracing of the celiac artery
and superior mesenteric artery (SMA) shows a normal flow velocity in the SMA. The celiac artery, however,
shows a very turbulent waveform with a velocity greater than 3 m s-1. MR angiography confirms celiac artery
compression by the arcuate ligament of the diaphragm.
Hereditary haemorrhagic telangiectasia frequent haemorrhages requiring transfusion.

(Osler–Rendu–Weber disease) Vascular lesions in the liver can evolve into arterial
This disease is characterised by multiple small venous fistulas and aneurysms.60,61 Ultrasound
aneurysmal telangiectases distributed over the skin, may reveal large hepatic arteries feeding large,
mucous membranes, alimentary tract, liver, brain ectatic, serpiginous arteriovenous malformations,
162 and spleen. These patients have a tendency for which in turn feed large draining veins.61,62
7
Hepatic veins
Hepatic venous outflow obstruction
The term Budd–Chiari syndrome is usually taken
to refer to hepatic vein thrombosis. Budd–Chiari,
however, refers to liver dysfunction due to any
cause of compromised hepatic vein outflow, both
thrombotic and non-thrombotic. Ludwig and
associates have recommended that the term
hepatic venous outflow obstruction (HVOO) should
be used instead of Budd–Chiari syndrome.63
This is appropriate since spectral and colour
Doppler are capable of identifying numerous
non-thrombotic causes of HVOO and differen-
tiating them from hepatic vein thrombosis.64
Etiology may be related to pregnancy, tumour,
Fig. 7.36 Transverse colour Doppler image of the
mid-liver. A rounded lesion is present with some hypercoagulable state or IVC membranes, but the
internal debris layering dependently. Flow is perceived majority of cases are idiopathic.65 The clinical
around this debris (arrows). An arterial spectral Doppler presentation of HVOO will vary, depending
waveform was identified in the feeding vessels. upon how rapidly it develops and the degree of
Angiography confirmed that this was a partially
thrombosed intrahepatic hepatic artery aneurysm.
Fig. 7.37 Transverse colour Doppler image of the

upper liver. Even with maximised Doppler sensitivity,
no flow was perceived in the middle hepatic vein of
this patient (arrow). The hepatic vein had become
thrombosed when a central line was inadvertently
advanced through the heart into the inferior vena cava Fig. 7.38 Transverse ultrasound image of the mid-
and up into the middle hepatic vein. This relatively liver. A thrombus is present within the middle hepatic
acute thrombus is hypoechoic compared to the vein (arrows). Fortunately there has been some
surrounding liver parenchyma. As thrombus matures, recanalisation of flow around this clot which is
however, the echogenicity tends to become isoechoic isoechoic to the surrounding liver. In the absence
with the adjacent liver and the thrombus becomes of this flow this clot would have been very difficult
163
harder to perceive. to see.
7
MHV
RHV
Fig. 7.39 Contrast enhanced axial CT of the mid-liver.

Note the uniform enhancement centrally in the
caudate lobe, but peripherally the enhancement is Fig. 7.40 Transverse colour Doppler view of the mid-
decreased and irregular. The three hepatic veins liver in a patient with catheter related inferior vena
(arrows) are unenhanced due to the fact that they cava thrombus. Flow in the right hepatic vein (RHV)
are thrombosed. is reversed and coursing towards the transducer.
A prominent collateral vein (arrow) then carries flow
to the middle hepatic vein (MHV) and back towards
obstruction. Approximately 50% of patients first the heart. Note the aliasing the more distal hepatic
present with RUQ pain and hepatomegaly. Almost vein due to the higher velocity caused by the
all develop ascites, while a few may develop mild increased volume of flow.
jaundice. Patients with chronic, partial obstruc-
tion may develop cirrhosis and portal hyperten-
sion. If the obstruction progresses to complete mine with absolute certainty that the cause is
occlusion, then shock, hepatic coma and death thrombosis.
may ensue. HVOO may be seen in three disease cate-
Sonographic imaging features of HVOO may gories: hepatic vein thrombosis’ non-thrombotic
include echogenic intraluminal material (either focal compromise of hepatic vein drainage (e.g.
thrombus or tumour), diffuse narrowing and stricture, web or neoplasm), or reduced compli-
compression of the veins from generalised liver ance of liver parenchyma (e.g. hepatitis, cirrhosis
swelling, or focal vascular compromise by a mass. or transplant rejection).
Doppler findings include complete absence of Hepatic vein thrombosis most commonly occurs
hepatic vein flow (Fig. 7.37) or localised flow in patients with a hypercoagulation disorder and
disturbances due to focal partial obstruction. In the use of oral contraceptives increases the risk
addition, the central portions of the hepatic veins of hepatic vein thrombosis two and a half times.
(distant from the IVC) that remain patent will Hepatic vein injury and phlebitis may also be
display low-velocity continuous flow, rather than associated with thrombosis, but in approximately
the normal flow pattern with periodicity. Finally, two-thirds of cases, the cause is idiopathic. The
liver congestion due to HVOO will also cause caudate lobe has a separate venous drainage to
flow abnormalities in portal vein flow, such as the IVC, which is usually spared from thrombosis,
diminished hepatopetal, bidirectional or hepatofugal resulting in caudate enlargement and normal
(reversed) flow. The diagnosis of complete enhancement on contrast enhanced CT (Fig.
thrombosis by US imaging is difficult since the 7.39). Occasionally, thrombosis may be limited
echogenicity of the clot is often similar to that of to one or two of the hepatic veins, resulting in
the adjacent liver parenchyma (Fig. 7.38). Because shunting of blood from the affected lobe to the
the identification of absent blood flow by Doppler unaffected side through hepatic venous collaterals
164 is an exclusionary diagnosis, it is difficult to deter- (Fig. 7.40). A strategically located mass (either
a
7
b
TT
Fig. 7.41 (a) Longitudinal view of the inferior vena

c cava in a patient with renal cell carcinoma. A large
plug of tumour thrombus is seen extending in the
inferior vena cava up to the level of the heart
(arrows). (b) Transverse view of the liver. The inferior
vena cava tumor thrombus (TT) obstructs the hepatic
vein ostia. Note the distention of the hepatic vein
(arrows). (c) Spectral Doppler tracing shows flattening
of the hepatic vein flow profile and complete absence
of periodicity. The triphasic waveform is not present,
since cardiac retropulsation cannot get past the
tumour thrombus.
benign or malignant) may expand and press upon of the hepatic vein ostia; this results in damping
the ostia of the hepatic veins, resulting in impaired of cardiac pressure changes into the IVC and the
venous drainage. Renal cell carcinoma can hepatic veins and flattening of the Doppler
extend from the renal vein into the IVC in 5% of waveform. The obstruction may eventually lead
cases. Rarely this may extend up to the right to hepatic vein thrombosis.
atrium and compromise the hepatic vein ostia, Diffuse hepatic parenchymal disease resulting in
which may result in elevation of the liver enzymes a reduction of liver compliance can easily
(Fig. 7.41). Renal carcinoma metastases to the compromise hepatic venous drainage as these
liver maintain this same propensity for vascular are a low pressure system (basically the same as
invasion (Fig. 7.42). the right atrium; i.e. -2/+7 mmHg). Both oedema
Membranous obstruction (fibrous web) of the IVC from acute inflammation and fibrosis from chronic
has been reported as one of the major causes of parenchymal disease (Fig. 7.43a) can compromise
HVOO in South Africa and Asia.66,67 The aetiology the hepatic veins, producing a relative HVOO.
of these is probably acquired since chronic hepatitis Some periodicity may be perceived in close
B infection is common in these patients, and up proximity to the junction of the hepatic veins with
to 50% may develop HCC.There is some support, the IVC, but it quickly fades as the sample volume
however, for a congenital hypothesis. The is moved further away from the heart (Fig.
obstruction is usually at, or just above, the level 7.43b).68,69 165
7
presents as jaundice, hepatomegaly, pain, ascites
and altered liver functions. The associated
coagulopathy usually precludes biopsy for diag-
nosis.70 Spectral Doppler of the hepatic veins
typically shows a normal flow profile. Portal vein
flow, however, is typically decreased and may
actually reverse.71 Hepatic artery flow typically
shows increased resistance.72
Altered or increased hepatic vein

periodicity
Near the heart, the normal spectral Doppler
tracing of the hepatic vein has a small retrograde
component of flow above the baseline, the A-
Fig. 7.42 Oblique view of the right hepatic vein in a
wave. It is relatively small when compared to the
patient with renal cell carcinoma metastatic to the
antegrade components below the baseline, the S-
liver. A small focus of tumour (arrow) is seen
and D- waves. The ratio of retrograde to ante-
extending from the large echogenic mass in the right
lobe into the right hepatic vein. Note distention of the grade flow decreases during inspiration (which
hepatic vein above the tumour thrombus and the lowers intrathoracic pressure) or when the Doppler
turbulent flow with eddy currents beyond the sample volume is moved further from the heart
invading tumour. (Fig. 7.14). A consistently large component of
retrograde flow is an abnormal finding, and if
identified, the examiner should consider the
Hepatic veno-occlusive disease presence of cardiac or pulmonary disease (Fig.
Non-thrombotic occlusion of small hepatic veins 7.44). Distortion of the triphasic hepatic vein
can occur in bone marrow transplant recipients, waveform can occur with many different cardiac
with alkaloid toxicity or secondary to chemotherapy disorders including cardiomyopathy, constrictive
or radiation therapy.The obstruction of terminal pericarditis, tamponade, tricuspid or pulmonary
hepatic venules by connective tissue and collagen valve disease, atrioventricular dissociation, atrial
a
b
Fig. 7.43 (a) Transverse image of the liver in a patient with severe cirrhosis. The liver is small, nodular and very
echogenic. (b) Spectral Doppler tracing of the middle hepatic within the liver. Patency of the vessel is confirmed
by colour Doppler but flow is markedly compromised by compression of this vessel. There is complete absence
166 of periodicity and a relatively slow velocity.
7
b
Fig. 7.44 (a) Transverse spectral and colour Doppler view of the hepatic veins. The hepatic veins are markedly
engorged as would be expected with chronic cardiac congestion. Spectral Doppler waveform shows appropriate
periodicity within these vessels. The overall amount of retrograde flow, however, is disproportionately large.
When summating the area of flow above and below the baseline it becomes apparent that there is relatively
little total antegrade flow as would be expected with cardiac congestion. (b) Spectral Doppler tracing in another
patient with cardiac congestion. Although periodicity is identified in the tracing it is more disordered than the
standard triphasic waveform. Flow away from the heart (above the baseline) is only slightly less than flow
towards the heart (below the baseline). This decreased total antegrade flow is a manifestation of cardiac
ingestion.
fibrillation and right ventricular dysfunction.

Pulmonary hypertension and massive pulmonary
embolism may also distort the hepatic vein
waveform.
Specific patterns of hepatic vein velocity profiles
have been described in restrictive cardiomyopathy,
pericardial constriction, and tamponade. A dramatic
increase in D-wave amplitude in patients with
restrictive cardiomyopathy reflects increased right
atrial pressure and rapid early diastolic filling,
Fig. 7.45 Spectral Doppler tracing of the hepatic vein characteristic of restriction. Reversal of the A-
obtained concurrently with an ECG tracing in a wave is seen in pericardial constriction and
patient with severe tricuspid regurgitation. Note that tamponade, and represents impaired late diastolic
the waveform is biphasic. There is a large reversed filling of the right ventricle. It is greatest during
component of flow but it occurs during the T-wave of
forced expiration due to the increased intratho-
the cardiac cycle (during ventricular systole). This is a
racic pressure transmitted to the pericardium.
reversed S-wave, fused with the V-wave. In the
Tricuspid regurgitation (TR) results in large
absence of an ECG tracing it may be difficult to
determine that this is indeed a reversed S-wave. volumes of blood leaking back through an
Identifying this large retrograde component of flow incompetent or diseased tricuspid valve. In mild
and disordered waveform, however, should be TR, the hepatic vein flow profile is characterised
enough to channel further workup to a cardiac by attenuation of the S-wave and a relative increase
evaluation. in V-wave amplitude. In severe TR, systolic 167
7
(a) (b)
a
v
Type I
d
s
(c)
a
Type II v
s d
s-v
Type III d
(d)
a d
Type IV
Fig. 7.46 (a) Hepatic vein flow profiles are characterised into four different types relative to right ventricular
dysfunction. With progressively worsening right ventricular dysfunction, the hepatic vein spectral Doppler
tracings become more and more distorted. Type 1- normal pattern – see Fig. 7.13. The retrograde component of
flow (A-wave) follows soon after the P-wave on the ECG tracing. (b) A type 2 pattern has accentuated A-wave.
Greater reversal of flow is secondary to reduced right ventricular compliance and/or increased atrial systolic
force. (c) The type 3 pattern shows an attenuated systolic forward flow (S-wave). This is due to a decrease in the
descent of the base of the right ventricle. Also, increased early diastolic flow is seen during the D-wave due to
increased early diastolic filling of the right ventricle. Again, note increased A-wave amplitude. (d) Type 4 pattern
is seen with severe right ventricular dysfunction, which is usually accompanied by severe tricuspid insufficiency.
Note that the dominant reversed component of flow now occurs well after the QRS. It is not related to the
P-wave.
reversal of the S-wave and fusion of the S- and through the tricuspid valve into the IVC and the
V-waves is seen. Instead of forward flow filling hepatic veins73 (Fig. 7.45).
the atrium during atrial diastole, the relatively In atrioventricular dissociation, the atrial and
168 high ventricular systolic pressure forces blood back ventricular electromechanical events occur inde-
7
b
a
Fig. 7.47 (a) Contrast enhanced CT scan in a patient with a biopsy proven hepatocellular carcinoma. A brightly
enhancing lesion on the early arterial phase is identified in segment 2 (left lobe lateral segment) (arrow).
(b) Spectral Doppler tracing of the left hepatic artery and portal vein show changes characteristic of liver disease
and neoplasm. Note the relatively high arterial diastolic flow and low resistance. Note, however, the attenuated
portal vein inflow (arrow).
pendently of each other.With complete heart block, ventricle and increased A-wave amplitude. Patients
atrial contraction against the closed tricuspid valve with severe RV dysfunction usually have asso-
can result in a markedly accentuated A-wave. ciated tricuspid insufficiency. This results in S-
Clinically, the increased jugular vein pulsations wave reversal. S-wave amplitude changes may also
are described as ‘cannon’ A-waves. The same reflect reduced right atrial compliance, reduced
phenomenon can occur in patients with prema- RV systolic function or tricuspid insufficiency.
ture ventricular ectopic beats which follow atrial A-wave amplitude changes indicate reduced RV
contraction. compliance or increased right atrial inotropy
With atrial arrhythmias, the A-wave may have (Fig. 7.46).
a varying relationship to the S-wave due to
premature atrial contraction or a variable PR Arterialised hepatic vein flow
interval as seen in Mobitz I heart block. In atrial An arterial waveform within the hepatic veins is
flutter, multiple small amplitude A-waves may an extremely rare finding. It may be seen with a
be present. The lack of organised atrial activity fistula from the hepatic artery to the hepatic vein
in atrial fibrillation leads to the loss of distinct A- following biopsy or surgery. Rarely, erosion of an
waves in the hepatic vein tracing. hepatic artery aneurysm into the hepatic vein
In patients with moderate to severe right will produce arterial pulsations in the veins.
ventricular (RV) dysfunction, hepatic vein Doppler
flow profiles fall into three basic patterns.The first Focal liver lesions – malignant
Doppler indication of RV dysfunction is accen- Hepatocellular carcinoma
tuation of the atrial A-wave, due to reduced RV HCC is the most common primary malignant
compliance which cannot accommodate all of neoplasm of the liver. It most frequently occurs
the right atrial output. Further deterioration of RV in patients with underlying chronic liver disease,
function results in attenuated systolic forward such as hepatitis B infection, and less commonly
flow (S-wave) due to a decrease in the descent of with alcoholic cirrhosis. US imaging is rela-
the base of the right ventricle. There is increased tively poor for the detection of HCC with a
early diastolic forward flow (D-wave), because sensitivity of only 50% for the sonographic iden-
of increased early diastolic filling of the right tification of malignancy in cirrhotic livers.74 The 169
7
distorted echo-texture of the liver parenchyma, hepatic neoplasms.Their sonographic appearances
together with the tumour’s variable sonographic are markedly variable, but these do not correlate
appearance, makes it difficult to distinguish the with cell type. One of the more frequently occurring
neoplasm. appearances is that of the target pattern or halo
HCC is typically a hypervascular neoplasm and sign. The hypoechoic rim surrounding the lesion
several authors have promoted the use of spectral is caused by compressed liver parenchyma, or by
and colour Doppler in helping to diagnose proliferating tumour at the edge of the lesion.
HCC.75,76 The tumour typically shows abundant Colour Doppler may reveal displacement of
neovascularity, resulting in decreased vascular normal liver vasculature by the expanding
impedence and resistance. Arterioportal venous metastatic lesion. Little if any flow is usually seen
shunting also occurs through the tumour.77 There- in the metastasis itself. Spectral Doppler can
fore the Doppler waveform in these tumours typi- reveal low resistance, high velocity flow in the
cally manifests increased diastolic flow with a low hepatic artery (Fig. 7.33c). However, this is not
resistance flow profile (Fig. 7.47). A basket pattern consistent enough to be of value in characterising
of flow within the lesion on colour Doppler has an unknown lesion. If metastasis is suspect, this
been described as characteristic for HCC;76 the change in the hepatic artery spectral Doppler
internal branching vessels within the tumour, waveform can be considered strongly suggestive,
combined with the network of surrounding vessels but it does not obviate the need for biopsy.
being responsible for this appearance. However, Power Doppler imaging can assess vascularity in
similar findings occur in other conditions. A low the majority of small liver nodules, but the
RI, high velocity flow with systolic frequency shifts pattern distribution of tumoral vascular signals
approximating 3 m s-1 may be seen in the main does not provide reliable differential diagnostic
hepatic artery but metastases to the liver, especially criteria.79
from hypervascular primaries, can manifest similar
changes in the hepatic artery flow profile. Focal liver lesions – benign
Peripheral portal vein flow alteration has been Hepatic steatosis (fatty liver)
suggested by some investigators to be valuable in In response to hepatocellular disease, the liver can
differentiating benign from malignant lesions. accumulate triglycerides within the hepatocytes.
Hepatofugal flow can be perceived on spectral and This reversible cellular response may be seen with
colour Doppler in proximity to HCC or metastatic obesity, alcoholic liver disease, diabetes, parenteral
disease. However, Miller et al78 showed that this nutrition, and numerous other disorders.80
kind of flow profile can also be identified in Ultrasound commonly reveals a bright echogenic
proximity to large haemangiomas, subcapsular liver with poor through transmission. The
haematomas, and other benign conditions. central vasculature is often poorly visualised due
Therefore, this finding cannot alter the standard to compression of these vessels by surrounding
differential diagnosis based on grey-scale findings. fat-laden parenchyma. It is well known that fatty
Low sensitivity and specificity unfortunately infiltration can be patchy and irregular.
limits the value of US imaging and Doppler in Occasionally the appearance can be nodular, but
diagnosing HCC. Biphasic CT is now considered most often it is non-spherical in shape with
the most accurate tool for the early detection of geometric margins or wedge shaped segmental
HCC, where a brightly enhancing hepatic lesion distribution. Doppler ultrasound typically shows
during the arterial phase of contrast injection is no change in the hepatic haemodynamics, or
considered HCC until proven otherwise. distribution of vessels. Hepatic artery and portal
vein velocities maintain a normal ratio. This is in
Metastases contrast to malignancy which usually shows an
Metastatic liver lesions occur with a frequency increased hepatic arterial velocity. Absence of
170 approximately 20 times greater than primary this velocity alteration may be helpful in further
7
reinforcing the impression of benign fatty infil- middle-aged women. It is believed to be hormone
tration in cases where the imaging appearance is dependent and not premalignant. Ultrasound often
confusing (Fig. 7.34). If significant doubt exists, reveals a solitary, small lesion (<3 cm), often at
CT, MR scintigraphy, or biopsy may be necessary the periphery of the liver. The echogenicity is
to clinch the diagnosis.81,82 varied with hypo-, hyper-, and isoechoic FNH
reported with equal frequency. A central scar has
Haemangiomas been reported as a dominant feature of FNH,
These are the most common solid benign however, it is seldom seen sonographically but,
neoplasms of the liver.They are hyperechoic when even if seen, it does not help clinch a benign
small (<3 cm), but the echogenicity of larger diagnosis since fibrolamellar HCC often also has
haemangiomas can be variable. Colour Doppler a central scar. FNH is typically hypervascular
adds little to the diagnosis of haemangioma as with a prominent central artery and radiating
the flow is typically too slow to register, even at branches in a stellate or spoke wheel configu-
the most sensitive settings. Flow may be ration with centrifugal flow; this is a unique
demonstrated occasionally with power Doppler characteristic of this lesion (Fig. 7.49). If doubt
(Fig. 7.48). Unfortunately, this may only serve still exists in the diagnosis then a technetium 99
to confuse the diagnosis, as the classic teaching sulfur colloid study should be performed as the
is that flow in haemangiomas is imperceptible. presence of Kupffer cells within the lesion can
In common with other benign masses, there is successfully differentiate FNH from other
rarely any alteration to the hepatic arterial inflow pathology.83–85
on spectral Doppler. Further evaluation with
scintigraphy, or MR, may be necessary to clinch Adenoma
the diagnosis. Hepatic adenoma, a rare benign liver tumour, is
being seen with increasing frequency. In females
Focal nodular hyperplasia it is related to oral contraceptive use, in males to
Focal nodular hyperplasia (FNH) is a relatively anabolic steroids.86 Adenomas appear as solid
rare hepatic lesion typically seen in young to masses with variable echogenicity due to differences
Fig. 7.48 (a) Transverse colour Doppler image of the right lobe of the liver. A large echogenic mass is easily
identified (arrow). No flow is perceived with the mass by colour Doppler despite settings at the most sensitive
level. (b) Power Doppler image of this lesion shows a leaf like pattern of signal within the mass. This is most
likely an artifact based on motion discrimination. With power Doppler artifact tends to occur within the pixels of
high intensity reflection. Biopsy did confirm this lesion to be a haemangioma. 171
7
a
b
Fig. 7.49 (a) Arterial phase of a contrast enhanced CT scan through the upper liver. A briskly enhancing round
mass with a central scar is clearly evident (arrows). (b) Colour Doppler through the lesion shows it should be
hypervascular. Biopsy confirmed this lesion to be focal nodular hyperplasia.
in fat content or haemorrhage. Doppler has not portocaval, and mesoatrial shunts.90 As with any
shown utility in their differentiation from other surgical anastomosis, stenosis and eventual
liver masses. thrombosis may develop. Imaging of shunt
integrity by Doppler ultrasound is sensitive, but
Infantile haemangioendothelioma because the shunt is usually retroperitoneal in
Infantile haemangioendothelioma is a rare liver location it can be extremely difficult to visualise.
tumour of infancy. It commonly presents as multiple It is mandatory that the sonographer knows the
lesions scattered throughout the liver with areas exact type and location of the shunt in order to
of infarction, haemorrhage and occasionally calci- focus the examination in the correct region. Brisk
fication.The tumour is composed of anastomosing hepatofugal flow in the main portal vein is a
vascular channels lined by one or more layers of secondary finding that indicates shunt patency,
endothelial cells. The incidence of congestive but the definitive diagnosis is made by direct
heart failure is high because of arterial venous visualisation of the shunt with colour Doppler91
(AV) shunting within the masses. It is considered (Fig. 7.51).
a benign tumour and most gradually regress after
presentation. Some children, however, succumb Transjugular intrahepatic
to the congestive heart failure associated with portosystemic shunts
the high degree of AV shunting. The ultrasound Treatment of portal hypertension and variceal
appearance of these lesions usually shows mixed bleeding by interventional diversion of portal
or variable echogenicity. Large feeding arteries flow as an alternative to surgery or sclerotherapy
and draining veins (Fig. 7.50) can be perceived was first proposed in 1969.92 Transjugular intra-
by colour Doppler with turbulent flow noted on hepatic portosystemic shunt (TIPS) placement
spectral Doppler.87–89 involves the percutaneous creation of a link between
the high-pressure portal system of a cirrhotic
patient and the low-pressure hepatic veins. A
MONITORING TREATMENT
transparenchymal track is created between the
OF LIVER DISEASE
hepatic vein and the portal vein, which is dilated
Surgical portosystemic shunts and a stent then inserted to maintain patency.
Numerous variants of surgical shunting have been Although the TIPS is most often placed between
devised over the years including mesocaval, distal the right hepatic vein and the right branch of the
172 splenorenal (Warren shunt), proximal splenorenal, portal vein, a left hepatic vein to left portal vein
7
Monitoring treatment of liver disease

b
HV
Fig. 7.50 (a) Oblique colour Doppler image of the right lobe of the liver. A large feeding artery is identified with
flow going into a tangled vascular malformation. (b) From this lesion flow is seen draining into a large hepatic
vein. Infantile haemangioendothelioma was confirmed at autopsy.
route may be selected for technical or anatomic bosis, measure vessel size and search for the
reasons. presence of portosystemic varices. This is best
accomplished with CT angiography. Preprocedural
Preprocedural assessment Doppler assessment allows a comparative appre-
Imaging evaluation should be performed before ciation of the results of the procedure; specifi-
TIPS placement to estimate liver size, evaluate cally changes in haemodynamics, spleen size and
for the presence of tumour, confirm patency of the amount of ascites.The varices (especially left
the portal vein and hepatic veins, rule out throm- gastric and recanalised paraumbilical) should be
identified prior to the procedure so that they
may be occluded by coil placement.
Postprocedural assessment
Unfortunately, TIPS complications are very
K common, the most significant being progressive
narrowing with eventual thrombosis and occlu-
sion. Clinical monitoring of TIPS function is
LT insensitive. By the time compromise becomes
RV
clinically evident, the TIPS is usually occluded.
SV The best way to monitor a TIPS is with Doppler
sonography and scanning at regular intervals
should be considered mandatory. Prompt iden-
tification of a stenosis with timely intervention
may prevent progression to thrombosis. A
Fig. 7.51 Oblique longitudinal view of the left flank
recently thrombosed TIPS may be recanalised
focused at the upper pole of the left kidney. The candy
successfully, but if the clot is allowed to mature
cane shaped large vessel is identified. The upper
then treatment usually requires placement of a
portion represents the splenic vein with flow coursing
towards the transducer. It has been hooked in
second TIPS.
surgically to the renal vein. The patient is status Routine Doppler monitoring after shunt
postsplenectomy anastomosis of the distal splenic placement should be considered mandatory as it
vein to the renal vein. The anastomosis is widely is an excellent non-invasive modality for sequential
patent with flow confirmed by Doppler. monitoring of TIPS patency. An evaluation of 173
7
the TIPS should be performed within 24 h after
its creation to confirm patency and establish base-
line flow directions and velocities. Subsequent
evaluation is conducted just prior to discharge of
the patient and then periodically thereafter; the
frequency varies among centres, but most re-
examine the shunt at approximately 3 month
intervals.
Optimum scanning parameters and normal
US findings following TIPS have been described
by a number of investigators.93–97 Transducers of
2.25–3.5 MHz are usually necessary because of
the frequently increased echogenicity and sound
attenuating features of the cirrhotic liver, along
with the fact that the shunt is usually fairly deep Fig. 7.52 Grey-scale longitudinal image of the right
lobe of the liver. The TIPS catheter is seen coursing
within the body. The Doppler gain should be set
from the porta hepatis region to the junction of the
as high as possible, without encountering noise.
right hepatic vein with the inferior vena cava (arrow).
The transducer is focused at the level of the
This long shunt is composed of two stent elements.
shunt or vessel of interest; the sample volume is Note the subtle narrowing at the mid-shunt where
placed in the centre with an angle of insonation the two stent elements overlap. The mesh like
less than 60° when feasible.This may be difficult appearance of the shunt is due to the echo reflection
to accomplish, especially in the middle segment off the individual wire elements of the stents.
of the shunt, where the direction of flow
frequently runs perpendicular to the insonating
beam. The pulse repetition frequency is set as
low as possible, but avoiding aliasing. Optimising
the scanning parameters will help minimise the
chances of making a false positive diagnosis of
shunt thrombosis.93,97
Normal findings
A complete TIPS evaluation includes a survey of
the abdomen to quantify ascites. The radiologist
should also look for intrahepatic, perihepatic or
subcapsular haematomas; intraperitoneal bleeding
(as indicated by increased volume or echogenicity
of the ascites); biliary obstruction; and echogenic
contents in the common bile duct or gallbladder
Fig. 7.53 Transverse view of the TIPS at its junction
that suggest haemobilia.95,98 The shunt is then
at the inferior vena cava. The opening of the stent is
located between the right portal vein and right
seen projecting into the right atrium (arrow).
hepatic vein. The stent is highly echogenic and
appears as two parallel, curvilinear lines, usually
uniform in diameter in the parenchyma but appear artificially narrowed.95 The stent should
slightly flared at the portal and hepatic venous extend from within the portal vein, across the
ends (Fig. 7.52). The shunt diameter is easily parenchyma, and into the hepatic vein. Imaging
measured96 but a curved TIPS that passes sometimes reveals malposition of the stent as a
174 obliquely through the plane of insonation may result of inappropriate deployment, or subsequent
7

migration down into the portal vein, or up into
the hepatic vein and right atrium (Fig. 7.53).
Flow within the stent is then evaluated by
Doppler ultrasound. The presence of blood flow
is easily confirmed, as the entire shunt lumen fills
with colour due to the relatively fast, turbulent
flow (Fig. 7.54). Assuming a right portal to right
hepatic vein TIPS, the velocity, flow direction
and waveform are then checked at the portal vein
end, mid-shunt and hepatic vein end. The main
portal vein and the left portal vein are assessed
and the right hepatic vein is checked both proximal
to and just beyond its junction with the stent
(Fig. 7.55). Spectral Doppler evaluation should Fig. 7.54 Oblique colour Doppler image of the TIPS.
verify that the direction of flow in the shunt is The entire lumen is saturated with colour indicating
from the portal vein to the hepatic vein. Flow patency. Note the non-uniformity of colour encoding
through the normal shunt is non-pulsatile, but flow – a function of narrowing at the area of stent overlap.
in a widely patent shunt may show periodicity
throughout the shunt due to right atrial pressure
changes being transmitted back through the at the hepatic venous end of the shunt, while the
shunt, against the direction of flow (Fig. 7.56). other half had high-velocity turbulent flow.96
Periodicity is most prominent near the hepatic Periodicity may be accentuated within the shunt
venous end.97 In one study, half of the patients in patients with tricuspid valve disease or
with patent TIPS demonstrated some periodicity congestive heart failure.93
Fig. 7.55 Illustration of the TIPS

and its related vasculature in the
standard right hepatic vein to right
portal vein configuration. Blue circles
indicate those points at which a
Doppler tracing should be obtained
in a complete TIPS ultrasound
evaluation. Figure courtesy of
Gerald Mulligan, MD.
Doppler locations
Mandatory
Optional
175
7
stent-bearing portion of the portal vein represent
flow in the portal vein, not in the shunt.95 Flow
in the left and right portal veins usually becomes
hepatofugal, flowing out of the diseased liver and
towards the inflow of the shunt93 (Fig. 7.57).
Depending on the diameter of the shunt and the
severity of the liver disease, however, flow may
continue to be hepatopetal into the parenchyma.
If the patient has patent paraumbilical vein
collaterals, these will continue to shunt blood
away from the liver. Flow in the left portal vein,
therefore, continues in a hepatopetal direction
despite normal TIPS function. If portal branch
Fig. 7.56 Doppler tracing obtained at the portal vein
flow changes direction over time from hepatofugal
end of the TIPS. Note the marked periodicity of flow
within the TIPS. The waveform is that commonly seen
to hepatopetal, a significant flow-limiting lesion
in hepatic veins and inferior vena cava. Identification is assumed to be present in the TIPS.
of this degree of periodicity at the portal vein and of The Doppler data are recorded and maintained
the stent is a confident indicator of a widely patent in a table format for follow-up (Fig. 7.58). Serial
shunt. documentation provides the best means of
identifying any variations in velocity and/or flow
direction over time and these changes are the
Flow velocities in the shunt vary widely, ranging best early indicator of shunt compromise.
from approximately 50 to 270 cm s-1.94,96,97
Velocities can also be quite variable through the
shunt itself, usually increasing from the portal
venous end to the hepatic venous end of the shunt.
The mean velocity of patent shunts has been
reported as 95 cm s-1 in the shunt near the portal
venous end93 and 120 cm s-1 in the middle
segment of the shunt.94 Flow across the shunt is
usually quite turbulent, especially when multiple
stent components are used, when overriding stents
can cause a relative narrowing of the shunt lumen.
Normal velocities in the main portal vein are
variable. Following TIPS insertion, the mean portal
vein velocity has been reported to increase from
7 cm s-1 to 24 cm s-1 in one study96 and from
20 cm s-1 to 38.4 cm s-1 in another.97 Hepatic
Fig. 7.57 Combined spectral Doppler tracing of the
arterial flow has also been shown to increase after
left branch of the portal vein and hepatic artery in a
TIPS, presumably because the shunt diverts the
patient with an appropriately functioning TIPS.
portal venous inflow away from the liver.95
The sample volume is opened wide in position so
In a properly functioning TIPS, flow direction that both vessels are interrogated in the same
in the portal system is towards the portal vein tracing. Note flow in the hepatic artery is towards the
end of the stent.Therefore, flow in the main portal transducer, therefore into the left lobe of the liver.
vein is hepatopetal and its velocity is typically The left portal vein flow, however, is away from the
quite brisk (between 20 and 50 cm s-1). It must transducer, therefore hepatofugal and towards the
176 be kept in mind that velocities measured in the inflow of the shunt.
7

Name: ID:
VELOCITIES DIRECTIONS
Date PV End Mid HV End Main PV Lt PV HV TIPS Comments
Fig. 7.58 TIPS data sheet. Each patient receiving a TIPS should have a data sheet maintained with velocities and
flow directions documented at each visit. Progressive compromise of the TIPS can then be more easily
diagnosed as progressive changes in velocities or a change in direction flow become manifest. Velocity
measurement in the mid-TIPS tends to be the most erratic. A persistent decrease in main portal vein velocity
over the sequence of studies is the most definitive indicator of progressive shunt compromise.
Shunt stenosis Several investigators have attempted to deter-

The two most common causes of TIPS compromise mine flow velocities which define the presence of
are neointimal hyperplasia throughout the shunt, TIPS stenosis93-95,100 but reported findings have
or a focal stenosis at the hepatic vein end. Most varied considerably. In one study, a velocity <50 cm
TIPS will have some degree of neointimal hyper- s-1 at the portal venous end was 100% sensitive
plasia, but this may progress to the point where and 93% specific.93 In another study, a velocity
it limits flow through the TIPS. At the point of <50 cm s-1 in the middle segment of the TIPS was
maximum stenosis within the TIPS, a high-velocity 78% sensitive and 99% specific, with a positive
jet may be perceived by Doppler (Fig. 7.59). predictive value of 96%, negative predictive
Other components of the TIPS and the portal value of 91% and accuracy of 92%.94 [When
system, however, will show decreased velocities. these investigators used a velocity <60 cm s-1 as
With sufficient compromise, flow in the branch the criterion, sensitivity increased to 84% but
portal veins becomes hepatopetal and flow in specificity dropped to 89% and accuracy to 87%.
the main portal vein may become hepatofugal. At <70 cm s-1, sensitivity was 89% but specificity
Focal hepatic vein stenosis can occur where the was 83% and accuracy was 85%. In another
proximal end of the TIPS abuts the hepatic vein. study, a velocity of 90 cm s-1 was applied, but the
Focal irritation of the vein wall by the stent wires sensitivity was only 87.5% with specificity of
may cause a bar of granulation tissue to build up. 95%.]101 These varied findings underscore the
This results in decreasing velocities throughout fact that flow velocities vary widely from patient
the shunt.95 A key Doppler finding of this focal to patient and that the best method for TIPS
stenosis is the presence of poststenotic flow evaluation is to use individual patient baseline
disturbances with a high-velocity jet and turbu- velocities obtained soon after TIPS placement.102
lence in the hepatic vein or IVC99 (Fig. 7.60). A change in velocity of ± 50 cm s-1 from baseline
The sonologist must therefore evaluate flow has been proposed as the threshold value for
beyond the end of the stent, sometimes even as predicting haemodynamically significant shunt
far as the right atrium. Flow in all three hepatic compromise.74 It becomes obvious with this
veins is normally towards the heart but a stenosis spectrum of reports that there is controversy
at the junction of the TIPS and the hepatic vein regarding the accuracy of Doppler ultrasound for
can cause flow compromise peripherally in the the detection of TIPS malfunction, nevertheless
vein with damping of periodicity, or segmental it is the best non-invasive means for following
flow reversal (Fig. 7.61). TIPS.103 177
7
a
b
Fig. 7.59 (a) Oblique colour Doppler image along the

c length of the TIPS shunt. Turbulent flow is seen
across this shunt. Note the absence of flow along the
shunt wall (arrows) indicating neointimal hyperplasia
along almost the entire length of the TIPS. (b) Flow is
seen along the length of the TIPS. At the portal vein
it is relatively uniform. At the mid-TIPS there is
narrowing and turbulence with the focus of neointimal
hyperplasia compromising the lumen. (c) Spectral
Doppler tracing at the mid-TIPS shows turbulent flow
at the high velocity of greater than 2.5 m s-1, well
above the accepted normal upper limit of 1.2 m s-1 for
the mid-TIPS.
If there is significant flow compromise through

the TIPS, main portal vein velocity decreases.
Flow in the left portal vein may change back to
a hepatopetal direction (Fig. 7.62), representing
a reversion to pre-TIPS haemodynamics (Table
7.4).95,97,99
Table 7.4 Criteria for compromised TIPS

function
• Shunt velocity of <50 cm s-1

Fig. 7.60 Spectral and colour Doppler image at the
• Increase or decrease in shunt velocity of >50 cm
hepatic vein end of the TIPS catheter. Note on the s-1compared with initial postprocedure value
colour image that the catheter ends well short of the • Focal region of increased velocity in the shunt or
inferior vena cava. The diameter of the hepatic vein hepatic vein
above the TIPS is relatively narrow (arrow). The • Hepatopetal flow in left and right portal vein
spectral Doppler tracing shows a velocity of 2.5 m s-1. branches
This high velocity jet is due to the focal stenosis due • Hepatofugal flow in main portal vein
178
to the hypertrophy bar of tissue.
7

a
Fig. 7.61 (a) Longitudinal colour Doppler image including the right hepatic vein and TIPS shunt. Note the poor
colour saturation of the TIPS indicating poor flow. Colour signal from the right hepatic vein shows flow coursing
through the transducer, which is away from the inferior vena cava. This is a result of focal stenosis at the hepatic
vein junction with the TIPS shunt. (b) Transverse image of the same patient as in Figure (a). Flow in the right
hepatic vein is indeed coursing away from the heart. It collateralises to the mid-hepatic vein which shows the
appropriate flow direction back towards the IVC.
Shunt occlusion sion must be considered. The absence of flow by

If no flow can be detected in the shunt and flow colour or power Doppler within the TIPS is a
velocity and direction in the portal vein is the same highly specific indicator of shunt thrombosis
as it was prior to TIPS placement, shunt occlu- (Fig. 7.63).92,93 However, before concluding
that the stent is occluded, meticulous scanning
for slow flow should be performed, since flow
velocity may be extremely low in a shunt that is
Fig. 7.62 Spectral Doppler tracing of the left hepatic

artery and portal vein including a portion of the TIPS
catheter. Absence of colour signal with the shunt
indicates TIPS thrombosis. This is further confirmed Fig. 7.63 Power Doppler image of the TIPS catheter.
by the direction of flow in the left portal vein which is Absence of colour signal within the mid-shunt
hepatopetal. This is in contrast to the normal left indicates complete TIPS thrombosis. Note the colour
portal vein hepatic artery flow direction as illustrated artefact from the stent wires, not to be confused
in Figure 7.57. with flow along the wall. 179
7
highly stenosed but still patent. Colour Doppler thrombosis for technical reasons is rarely a
settings (including pulse repetition frequency, problem. Identifying low flow, however, has the
gain and filtration) need to be optimised to same implication as occlusion, and that is the
differentiate true occlusion from very slow need for TIPS revision. Indeed, repeat inter-
flow. A properly functioning TIPS, however, is ventions are the key to long-term success of
not a low-flow system, so misinterpretation of TIPS.104
REFERENCES
1. Beissert M, Delorme S, Mutze S, et al. Doppler flowmetry in the diagnosis of portal
Comparison of B-mode and conventional hypertension. J Hepatol 1992; 16:298–303.
colour/power Doppler ultrasound, contrast- 13. Ohta M, Hashizume M, Kawanaka H, et al.
enhanced Doppler ultrasound and spiral CT in the Prognostic significance of hepatic vein waveforms
diagnosis of focal lesions of the liver: results of a by Doppler ultrasonography in cirrhotic patients
multicentre study. Ultraschall Med 2002; with portal hypertension. Am J Gastroenterol
23(4):245–250. 1995; 90:1853–1857.
2. Rao BK. Colour flow Doppler sonography of the 14. Pierce ME, Sewell R. Identification of hepatic
abdomen. Curr Opin Radiol 1991; 3:225–229. cirrhosis by duplex Doppler ultrasound value of
3. Czembirek H. Value of abdominal pulsed duplex the hepatic artery resistive index. Australas Radiol
sonography. Radiologe 1987; 27:98–105. 1990; 34:331–333.
4. Mostbeck G, Mallek R, Gebauer A, et al. 15. Iwao T, Toyonaga A, Shigemori H, et al. Hepatic
Duplex ultrasound and colour-coded Doppler artery haemodynamics responsiveness to altered
ultrasound of visceral blood vessels in abdominal portal blood flow in normal and cirrhotic livers.
diseases. Wien Klin Wochenschr 1992; Radiology 1996; 200:793–798.
4:227–233. 16. Vilgrain V, Lebrec D, Menu Y, et al. Comparison
5. Grant EG, Schiller VL, Millener P, et al. Colour between ultrasonographic signs and the degree
Doppler imaging of the hepatic vasculature. of portal hypertension in patients with cirrhosis.
AJR Am J Roentgenol 1992; 159:943–950. Gastrointest Radiol 1990; 15:218–222.
6. Kruskal JB, Newman PA, Sammons LG, et al. 17. Shikare SV, Bashir K, Abraham P, et al. Hepatic
Optimizing Doppler and colour flow US: perfusion index in portal hypertension of cirrhotic
application to hepatic sonography. Radiographics and non-cirrhotic aetiologies. Nucl Med Commun
2004; 24(3):657–675. 1996; 17:520–522.
7. Ralls PW. Colour Doppler sonography of the 18. Leen E, Goldberg JA, Anderson JR, et al.
hepatic artery and portal venous system. AJR Am Hepatic perfusion changes in patients with liver
J Roentgenol 1990; 155:517–525. metastases: comparison with those patients with
8. Iwao T, Toyonaga A, Oho K, et al. Value of Doppler cirrhosis. Gut 1993; 34:554–557.
ultrasound parameters of portal vein and hepatic 19. Moriyasu F. Doppler ultrasound in diagnosis of
artery in the diagnosis of cirrhosis and portal liver tumour and portal hypertension. Radiol Med
hypertension. Am J Gastroenterol 1997; (Torino) 1993; 85:44–55.
92:1012–1017. 20. Yasuhara K, Kimura K, Nakamura H, et al. Doppler
9. Harada A, Nonami T, Kasai Y, et al. Systemic velocity histogram analysis of hepatocellular
haemodynamics in non-cirrhotic portal carcinoma. J Clin Ultrasound 1995; 23: 225–231.
hypertension – a clinical study of 19 patients. 21. Gorka W, Kagalwalla A, McParland BJ, et al.
Jpn J Surg 1988; 18:620–625. Diagnostic value of Doppler ultrasound in the
10. Grant EG, Tessler FN, Gomes AS, et al. assessment liver cirrhosis in children:
Colour Doppler imaging of portosystemic shunts. histopathological correlation. J Clin Ultrasound
AJR Am J Roentgenol 1990; 154:393–397. 1996; 24:287–295.
11. Sacerdoti D, Merkel C, Bolognesi M, et al. 22. Lafortune M, Madore F, Patriquin H, et al.
Hepatic arterial resistance in cirrhosis with and Segmental anatomy of the liver: a sonographic
without portal vein thrombosis: relationships with approach to the Couinaud nomenclature.
portal haemodynamics. Gastroenterology 1995; Radiology 1991; 181:443–448.
108:1152–1158. 23. Zwiebel WJ. Sonographic diagnosis of hepatic
12. Zironi G, Gaiani S, Fenyves D, et al. Value of vascular disorders. Semin Ultrasound CT MR
180 measurement of mean portal flow velocity by 1995; 16(1):34–48.
24. Abu-Yousef MM. Normal and respiratory variations
on hepatic haemodynamics and the nature of the

7
References
of the hepatic and portal venous duplex Doppler arterioportal communications. Hepatology 1988;
waveforms with simultaneous electrocardiographic 8:16–20.
correlation. J Ultrasound Med 1992; 11:263–268. 39. Abu-Yousef MM, Milam SG, Farner RM.
25. Gallix BP, Taourel P, Dauzat M, et al. Flow Pulsatile portal vein flow: a sign of tricuspid
pulsatility in the portal venous system: a study of regurgitation on duplex Doppler sonography.
Doppler sonography in healthy adults. AJR Am J AJR Am J Roentgenol 1990; 155:785–788.
Roentgenol 1997; 169:141–144. 40. Loperfido F, Lombardo A, Amico CM, et al.
26. Hosoki T, Arisawa J, Marikawa T, et al. Portal blood Doppler analysis of portal vein flow in tricuspid
flow in congestive heart failure: pulsed duplex regurgitation. J Heart Value Dis 1993; 2:174–182.
sonographic findings. Radiology 1990; 41. Gorg C, Riera–Knorrenschild J, Dietrich J. Pictorial
174:733–736. review: colour Doppler ultrasound flow patterns in
27. Wachsberg RH, Needleman L, Wilson DJ. the portal venous system. Br J Radiol 2002;
Portal vein pulsatility in normal and cirrhotic 75(899):919–929.
adults without cardiac disease. J Clin Ultrasound 42. Wachsberg RH, Simmons MZ. Coronary vein
1995; 23:3–15. diameter and flow direction in patients with portal
28. Lippert H, Pabst R. Arterial variations in man. hypertension: evaluation with duplex sonography
Munchen: JF Bergmann; 1985:32–33. and correlation with variceal bleeding. AJR Am J
29. Shapiro RS, Winsberg F, Maldjian C, et al. Roentgenol 1994; 162:637–641.
Variability of hepatic vein Doppler tracings in 43. Saddekni S, Hutchinson DE, Cooperberg PL.
normal subjects. J Ultrasound Med 1993; The sonographically patent umbilical vein in portal
12:701–703. hypertension. Radiology 1982; 145:441–443.
30. Westra SJ, Zaninovic AC, Vargas J, et al. 44. Gupta D, Chawla YK, Dhiman RK, et al. Clinical
The value of portal vein pulsatility on duplex significance of patent paraumbilical vein in
sonograms as a sign of portal hypertension in patients with liver cirrhosis. Dig Dis Sci 2000;
children with liver disease. AJR Am J Roentgenol 45(9):1861–1864.
1995; 165:167–172. 45. Amitrano L, Guardascione MA, Brancaccio V, et
31. Gibson PR, Gibson RN, Donlan JD, et al. Duplex al. Risk factors and clinical presentation of portal
Doppler ultrasound of the ligamentum teres and vein thrombosis in patients with liver cirrhosis.
portal vein: a clinically useful adjunct in the J Hepatol 2004; 40(5):736–741.
evaluation of patients with known or suspected 46. Parvey HR, Raval B, Sandler CM. Portal vein
chronic liver disease or portal hypertension. thrombosis; imaging findings. AJR Am J
J Gastroenterol Hepatol 1991; 6:61–65. Roentgenol 1994; 162:77–81.
32. Kuo CH, Changchien CS, Tai DI, et al. Portal vein 47. Raby N, Meire HB. Duplex Doppler ultrasound in
velocity by duplex Doppler ultrasound as an the diagnosis of cavernous transformation of the
indication of the clinical severity of portal portal vein. Br J Radiol 1988; 61:586–588.
hypertension. Chan Keng I Hsueh 1995; 48. Weltin G, Taylor KJ, Carter AR, et al.
18: 217–223. Duplex Doppler: identification of cavernous
33. Kozaiwa K, Tajiri H, Yoshimura N, et al. Utility of transformation of the portal vein. AJR Am J
duplex Doppler ultrasound in evaluating portal Roentgenol 1985; 144:999–1001.
hypertension in children. J Pediatr Gastroenterol 49. Pozniak MA, Baus K. Hepatofugal arterial signal in
Nutr 1995; 21:215–219. the main portal vein – an indicator of intravascular
34. Gorg C, Schwerk WB, Gorg K, et al. Focal tumour spread. Radiology 1991; 180:663–666.
involvement of malignant lymphoma in the liver. 50. Vine HS, Sequeira JC, Widrich WC, et al.
Bildegebung 1991; 58:67–70. Portal vein aneurysm. AJR Am J Roentgenol
35. Moriyasu F, Nishida O, Ban N, et al. Congestion 1979; 132:557–560.
index of the portal vein. AJR Am J Roentgenol 51. Mhanna T, Bernard P, Pilleul F, et al. Portal vein
1986; 146:735–739. aneurysm: report of two cases.
36. Piscaglia F, Donati G, Serra C, et al. Value of Hepatogastroenterology 2004; 51(58):1162–1164.
splanchnic Doppler ultrasound in the diagnosis of 52. Merritt CRB, Goldsmith JP, Sharp MJ.
portal hypertension. Ultrasound Med Biol 2001; Sonographic detection of portal venous gas in
27(7):893–899. infants with necrotizing enterocolitis. AJR Am J
37. Barakat M. Portal vein pulsatility and spectral Roentgenol 1984; 143:1059–1062.
width changes in patients with portal 53. Maher MM, Tonra BM, Malone DE, et al.
hypertension: relation to the severity of liver Portal venous gas: detection by grey-scale
disease. Br J Radiol 2002; 75(893):417–421. and Doppler sonography in the absence of
38. Rector WG, Hoefs JC, Hossack KF, et al. correlative findings on computed tomography.
Hepatofugal portal flow in cirrhosis: observations Abdom Imaging 2001; 26(4):390–394.
181
7
54. Joynt LK, Platt JF, Rubin JM, et al. Hepatic artery 68. Barakat M. Non-pulsatile hepatic and portal vein
resistance before and after standard meal in waveforms in patients with liver cirrhosis:
subjects with diseased and healthy livers. concordant and discordant relationships. Br J
Radiology 1995; 96:489–492. Radiol 2004; 77(919):547–650.
55. Han SH, Rice S, Cohen SM, et al. Duplex Doppler 69. Janssen HL, Tan AC, Tilanus HW, et al. Pseudo-
ultrasound of the hepatic artery in patients with Budd-Chiari Syndrome: decompensated alcoholic
acute alcoholic hepatitis. J Clin Gastroenterol liver disease mimicking hepatic venous outflow
2002; 34(5):573–577. obstruction. Hepatogastroenterology 2002;
56. Tanaka K, Numata K, Morimoto M, et al. 49(45):810–812.
Elevated resistive index in the hepatic artery as a 70. Boyer TD. Portal hypertension and bleeding
predictor of fulminant hepatic failure in patients esophageal varices: portal hypertension.
with acute viral hepatitis: a prospective study In: Zakim D, Boyer TD, eds. Hepatology: a
using Doppler ultrasound. Dig Dis Sci 2004; textbook of liver disease. Philadelphia: WB
49(5):833–842. Saunders; 1990:592.
57. Deasy NP, Wendon J, Meire HB, et al. The value 71. Brown BP, Abu-Yousef M, Farner R, et al. Doppler
of serial Doppler ultrasound as a predictor of sonography: a non-invasive method for evaluation
clinical outcome and the need for transplantation of hepatic veno-occlusive disease. AJR Am J
in fulminant and severe acute liver failure. Br J Roentgenol 1990; 154:721–724.
Radiol 1999; 72(854):134–143. 72. Herbetko J, Grigg AP, Buckley AR, et al.
58. Ramchandani P, Goldenberg NJ, Soulen RL, et al. Veno-occlusive liver disease after bone marrow
Isobutyl 2-cyanoacrylate embolization of a transplantation: findings at duplex sonography.
hepatoportal fistula. AJR Am J Roentgenol 1983; AJR Am J Roentgenol 1992; 158:1001–1005.
140:137–140. 73. Abu-Yousef MM. Duplex Doppler sonography
59. Falkoff GE, Taylor KJW, Morse S. Hepatic artery of the hepatic vein in tricuspid regurgitation.
pseudoaneurysm: diagnosis with real-time and AJR Am J Roentgenol 1991; 156:79–83.
pulsed Doppler ultrasound. Radiology 1986; 74. Dodd GD 3rd, Miller WJ, Baron RL, et al.
158:55–56. Detection of malignant tumours in end-stage
60. Cloogman HM, DiCapo RD. Hereditary cirrhotic livers: efficacy of sonography as a
hemorrhagic telangiectasis: sonographic findings screening technique. AJR Am J Roentgenol 1992;
in the liver. Radiology 1984; 150:521–522. 159:727–733.
61. Ralls PW, Johnson MB, Radin DR, et al. 75. Taylor KJW, Ramos I, Morse SS, et al. Focal liver
Hereditary hemorrhagic telangiectasis: findings in masses: differential diagnosis with pulsed
the liver with colour Doppler sonography. AJR Am Doppler ultrasound. Radiology 1987; 164:643–647.
J Roentgenol 1992; 159:59–61. 76. Tanaka S, Kitamura T, Fujita M, et al. Colour
62. Ocran K, Rickes S, Heukamp I, et al. Sonographic Doppler flow imaging of liver tumours. AJR Am J
findings in hepatic involvement of hereditary Roentgenol 1990; 154:509–514.
haemorrhagic telangiectasia. Ultraschall Med 77. Suzuki M, Takahashi T, Sato T. Medial regression
2004; 25(3):191–194. and its functional significance in tumour-supplying
63. Ludwig J, Hashimoto E, McGill DB, et al. host arteries: a morphometric study of hepatic
Classification of hepatic venous outflow arteries in human livers with hepatocellular
obstruction: ambiguous terminology of the carcinoma. Cancer 1987; 59:444–450.
Budd-Chiari syndrome. Mayo Clin Proc 1990; 78. Miller MA, Balfe DM, Middleton WD. Peripheral
65:51–55. portal venous blood flow alterations induced by
64. Hosoki T, Kuroda C, Tokunaga K, et al. Hepatic hepatic masses: Evaluation with colour and
venous outflow obstruction: evaluation with pulsed Doppler sonography. J Ultrasound Med
pulsed duplex sonography. Radiology 1989; 1996; 15:707–713.
170:733–737. 79. Gaiani S, Casali A, Serra C, et al. Assessment of
65. Singh V, Sinha SK, Nain CK, et al. Budd-Chiari vascular patterns of small liver mass lesions:
syndrome: our experience of 71 patients. value and limitation of the different Doppler
J Gastroenterol Hepatol 2000; 15(5):550–554. ultrasound modalities. Am J Gastroenterol 2000;
66. Lee DH, Ko YT, Yoon Y, et al. Sonography and 95(12):3537–3546.
colour Doppler imaging of Budd-Chiari syndrome 80. Marn CS, Bree RL, Silver TM. Ultrasonography of
of membranous obstruction of the inferior vena liver: technique and focal and diffuse disease.
cava. J Ultrasound Med 1994; 13:159–163. Radiol Clin North Am 1991; 29:1151–1170.
67. Sakugawa H, Higashionna A, Oyakawa T, et al. 81. Wang S-S, Chiang J-H, Tsai Y-T, et al. Focal hepatic
Ultrasound study in the diagnosis of primary fatty infiltration as a cause of pseudotumours:
Budd-Chiari syndrome (obstruction of the inferior ultrasonographic patterns and clinical differentiation.
vena cava). Gastroenterol Jpn 1992; 27:69. J Clin Ultrasound 1990; 18:401–409.
182
82. Yates CK, Streight RA. Focal fatty infiltration of
Doppler US in determination of patency and

7
References
the liver simulating metastatic disease. Radiology detection of stenoses. Radiology 1996;
1986; 159:83–84. 201:141–147.
83. Golli M, Mathieu D, Anglade M, et al. Focal 95. Foshager MC, Ferral H, Finlay DE, et al. Colour
nodular hyperplasia of the liver: value of colour Doppler sonography of transjugular intrahepatic
Doppler US in association with MR imaging. portosystemic shunts (TIPS). AJR Am J
Radiology 1993; 187:113–117. Roentgenol 1994; 163:105–111.
84. Learch TJ, Ralls PW, Johnson MB, et al. Hepatic 96. Longo JM, Bilbao JI, Rousseau HP, et al.
focal nodular hyperplasia: findings with colour Transjugular intrahepatic portosystemic shunt:
Doppler sonography. J Ultrasound Med 1993; evaluation with Doppler sonography. Radiology
12:541–544. 1993; 188:529–534.
85. Kudo M, Tomita S, Minowa K, et al. Colour 97. Surratt RS, Middleton WD, Darcy MD, et al.
Doppler flow imaging of hepatic focal nodular Morphologic and haemodynamic findings at
hyperplasia. J Ultrasound Med 1992;11:553–557. sonography before and after creation of a
86. Okuda K, Kojiro M, Okuda H. Neoplasms of the transjugular intrahepatic portosystemic shunt.
liver: benign tumours of the liver. In: Schiff L, AJR Am J Roentgenol 1993;160:627–630.
Schiff E, eds. Diseases of the liver. Philadelphia: 98. Foshager MC, Finlay DE, Longley DG, et al.
JB Lippincott; 1993:1236. Duplex and colour Doppler sonography of
87. Dachman AH, Lichtenstein JE, Friedman AC, et al. complications after percutaneous interventional
Infantile hemangioendothelioma of the liver: a vascular procedures. Radiographics 1994;
radiologic-pathologic-clinical correlation. AJR Am J 14:239–253.
Roentgenol 1983; 140:1091–1096. 99. Zemel G, Katzen B, Grubbs G, et al. Sonographic
88. Kew MC. Tumours of the liver: benign hepatic indicators of unsuccessful transjugular
tumours. In: Zakim D, Boyer TD, eds. Hepatology: intrahepatic portosystemic shunts. Radiology
a textbook of liver disease. Philadelphia; WB 1994; 193(P)(suppl):167.
Saunders, 1990:1232. 100. Dodd GD, Zajko AB, Orons PD, et al. Detection of
89. Pardes JG, Bryan PJ, Gauderer MWL. Spontaneous transjugular intrahepatic portosystemic shunt
regression of infantile hemangioendotheliomatosis dysfunction: value of duplex Doppler sonography.
of the liver: demonstration by ultrasound. AJR Am J Roentgenol 1995; 164:1119–1124.
J Ultrasound Med 1982;1:349–353. 101. Mituzani P, Saxon R, Alexander P, et al. Duplex US
90. Boyer TD. Portal hypertension. In: Zakim D, Boyer screening after transjugular intrahepatic
TD, eds. Hepatology: a textbook of liver disease. portosystemic shunt placement. Radiology 1993;
Philadelphia: WB Saunders; 1990:602. 189(P)(suppl):254.
91. LaFortune M, Patriquin H, Pomier G, et al. 102. Nazarian GK, Ferral H, Castaneda-Zuniga WR,
Haemodynamic changes in portal circulation after et al. Development of stenoses in transjugular
portosystemic shunts: use of duplex carotid intrahepatic portosystemic shunts. Radiology
sonography in 43 patients. AJR Am J Roentgenol 1994; 192:231–234.
1987; 149:701–706. 103. Wachsberg RH. Doppler ultrasound evaluation of
92. Rosch J, Hanafee WN, Snow H. Transjugular transjugular intrahepatic portosystemic shunt
portal venography and radiologic portacaval shunt: function: pitfalls and artifacts. Ultrasound Q 2003;
an experimental study. Radiology 1969; 19(3):139–148.
92:1112–1114. 104. Coldwell DM, Ring EJ, Rees CR, et al.
93. Chong WK, Malisch TA, Mazer MJ, et al. Multicenter investigation of the role of
Transjugular intrahepatic portosystemic shunt: transjugular intrahepatic portosystemic shunt in
US assessment with maximum flow velocity. management of portal hypertension. Radiology
Radiology 1993; 189:789–793. 1995; 196:335–340.
94. Feldstein VA, Patel MD, La Berge JM. Transjugular
intrahepatic portosystemic shunts: accuracy of
183
8
The kidney
Paul A. Dubbins
The kidney is a highly vascular organ, receiving to the arc of the anterior abdominal wall.
approximately 20% of the cardiac output. Many Geometrically this favours visualisation but
of the diseases of the kidney have a major vascular compromises the Doppler examination; however,
component, and systemic diseases such as visualisation is also compromised by bowel
hypertension are mediated through the vascular which overlies much of the course of both renal
control system of the juxtaglomerular apparatus. vessels.
The kidney would therefore appear to be a No single imaging approach can circumvent
fertile field for examination by Doppler. Renal the difficulties relating to geometry, vessel course
and renovascular disease might be expected to and the anterior abdominal relations of gas-
produce changes in the vascular supply, in the filled bowel. Each renal artery must therefore be
microvascular circulation and in venous return. examined with a flexible technique. Anatomical
variations in particular are difficult to image
with ultrasound, although supplemental arteries
ANATOMY AND TECHNIQUE
are being identified increasingly with improved
Examination of the renal arteries requires knowledge equipment resolution and improved sensitivity
of vascular anatomy, anatomical relations, of colour and power Doppler.
anatomical variations and surface anatomy. It The renal veins follow a course parallel to the
also requires an understanding of the effects of renal arteries. The right renal vein is the shorter,
geometry on the ability to record a Doppler signal. coursing anteriorly, medially and towards the
The right renal artery is an anterolateral branch head in front of the right renal artery (Fig. 8.4).
of the aorta and courses laterally and posteriorly It normally joins the inferior vena cava just
behind the inferior vena cava and then behind cephalad to the right renal artery as it crosses
the ipsilateral renal vein to the hilum of the kidney beneath the cava. On the left, the renal vein travels
(Fig. 8.1). Its anterior relations are predominantly medially and anteriorly, (Figs 8.3 and 8.5) usually
bowel: the duodenum, small bowel and trans- superior to the renal artery, passing between the
verse colon. At the hilum of the kidney it divides aorta and the superior mesenteric artery, where
into anterior and posterior branches, then into it may occasionally be compressed by the
interlobular arteries and subsequently the ‘nutcracker effect’ between the two vessels. This
arcuate arteries, which send out striate branches may cause dilatation of the more proximal portion
to the cortex (Fig. 8.2). On the left, the renal of the vein but this is not of any clinical signifi-
artery arises as a posterolateral branch of the aorta cance (Fig. 8.6). The right renal vein usually has
and courses posteriorly laterally and inferiorly no major tributaries but on the left the adrenal,
immediately behind the third and fourth parts of lumbar and gonadal veins usually join the renal
the duodenum, subsequently passing behind the vein; of these, only the gonadal vein is regularly
transverse and descending colon (Fig. 8.3). Both imaged on ultrasound as it enters the inferior
arteries therefore trace an arc that runs parallel aspect of the vein near the aorta. 185
8
The kidney
a
To gut:
coeliac
Gut
superior mesenteric
inferior mesenteric
b
Aorta To glands:
suprarenal
renal
Gland testicular
To walls:
phrenic
lumbar
median sacral
Fig. 8.1 (a) (b) Axial colour flow image, the level
c of the origin of the renal arteries. RRA, right renal
artery. (c) The origin of the right renal artery is
demonstrated as an anterolateral branch of the aorta
passing posterior to the inferior vena cava (IVC).
IVC
RRA
Technique – patient positioning and slim patients it is occasionally possible to follow

scan orientation the renal artery and vein into the hilum of the
With the patient supine, the aorta is localised with kidney by applying compression with the trans-
a longitudinal scan just to the left of the midline. ducer and slightly oblique angulation. Doppler
The probe is rotated through 90° and the superior signal recording is best performed at the origin
mesenteric artery identified (Fig. 8.7). Approxi- of the artery, by slight displacement of the probe
mately 1 cm below the origin of the superior to the right of the midline and angulation to the
mesenteric artery, the right renal artery can be left to identify the course of the first short segment
identified arising from the anterolateral surface of the left renal artery (Figs 8.7 and 8.8). Reversing
of the aorta. Occasionally the left renal artery the angulation to the right approximately 20°
can also be demonstrated at its origin from the from the perpendicular allows signal recording from
posterolateral or lateral surface of the aorta. Both the posterior and lateral coursing right renal
renal veins and their junction with the inferior artery1 (Fig. 8.9).
vena cava can usually be demonstrated in this The patient is then turned into the left posterior
plane (Figs 8.4 and 8.6). On moving the probe oblique position. Scanning just posterior to the
slightly to the right of the midline, the right renal right mid-axillary line in the longitudinal plane
186 artery is seen to turn posteriorly and laterally. In and angling towards the major vessels, both the aorta
a
The kidney
Fig. 8.2 (a) (b) Colour flow

8
Anatomy and technique

Doppler demonstrating
segmental, interlobar, arcuate
Pyramid and interlobular or striate
Cortex vessels.
Papilla
Interlobar
artery Renal
artery
Arcuate artery
Pelvis
of kidney
Striate arteries
Ureter
a
b
Fig. 8.3 (a) The origin of the left renal artery from the lateral aspect of the aorta and its course posterior to the
left renal vein is demonstrated. (b) The origin of the left renal artery in relation to the aorta, left renal vein and
superior mesenteric artery are identified. 187
8
The kidney
Fig. 8.4 The course of the right renal vein is

demonstrated relative to the right renal artery which
Fig. 8.6 The nutcracker appearance of the left renal
lies posteriorly.
vein as it passes between the aorta and the superior
mesenteric artery is identified. There is apparent
a continuity between the left renal vein and the aorta.
Fig. 8.5 (a) The course of the left renal vein between
the aorta and the superior mesenteric artery is Fig 8.7 Representation of the approach to Doppler
demonstrated. (b) Colour flow Doppler demonstrating evaluation of the renal arteries from the anterior
188 the left renal vein. Greater clarity has been achieved approach.
by a stomach distended with fluid.
The kidney
8

a b
Fig. 8.8 (a) The anterior origin of the right renal artery
c is demonstrated. (b) The anterior approach just to the
right of midline allows Doppler signal recording within
the ostium of the artery. (c) Doppler signal recording
within the renal artery as it courses posterior to the
inferior vena cava.
Fig. 8.9 The probe is angled slightly to the left from a

position just to the right of the midline demonstrating
the origin of the left renal artery in colour flow Fig 8.10 Representation of the approach to the renal
Doppler. arteries scanning from the right flank. 189
8
The kidney
and inferior vena cava can be identified lying
parallel to one another (Fig. 8.10). In this plane
the section cuts through both the right and left
renal arteries and the origins of both can usually
be identified (Fig. 8.11). Alteration of the patient
position to the right posterior oblique position
will identify the junction of the renal veins with
the inferior vena cava (Fig. 8.12). Doppler signal
recording is possible in this plane from both
renal arteries (Fig. 8.13), although occasionally,
particularly in heavy or overweight patients, the
left renal artery signal is poorly recorded, either Fig 8.12 Oblique scanning from the right flank,
because of its depth and difficulties with pulse demonstrating the aorta and inferior vena cava,
repetition frequency, or simply because of the origin of the right renal artery and the junction of
attenuation of the beam. Scanning from the the right renal vein with the inferior vena cava.
right flank and angling the probe anteriorly to From Dubbins33, with permission.
align the sonographic plane with the coronal
plane of the kidney allows the vessels at the renal
hilum and within the kidney to be sampled.
These can be observed with colour Doppler
radiating from the hilum (Fig. 8.14).
The right lobe of the liver acts as an acoustic
window to the right renal artery. On the left there
is no such acoustic window and visualisation of
the left renal artery from origin to hilum is extremely
difficult.This may be facilitated by compression,
or by using the coronal approach to the kidney
and following the course of the renal artery in a
retrograde fashion towards its origin, although
Fig. 8.13 Colour flow Doppler in the left posterior

oblique position demonstrating the origin of both
renal arteries.
in this author’s experience the left renal artery is

rarely seen in its entirety.The origin may be seen
in transverse plane, scanning from the anterior
approach and angling slightly to the left (Fig. 8.3
and 8.9), or it may be seen with the patient in
the left posterior oblique position as described
above (Fig. 8.15).The more distal portion of the
Fig 8.11 Oblique view through the aorta scanning main renal artery may be demonstrated just
from the right flank, demonstrating the origin of both proximal to the hilum by the coronal approach,
renal arteries and the adjacent inferior vena cava. but the mid-portion of the left renal artery is only
190 From Dubbins33, with permission. rarely seen except in slim patients. By contrast
The kidney
8

a much of the left renal vein is demonstrated with
the patient in the supine position; its more
anterior and superior location, together with
larger diameter, allowing visualisation using the
left lobe of the liver as an acoustic window
(Fig. 8.16).
Anatomical variation
It has been found that 40% of individuals do not
conform to this simple anatomical arrangement,
with one or more accessory arteries supplying
either or both kidneys. These may arise from the
aorta immediately adjacent to the main renal
b
artery, but their origin may be from anywhere
along the abdominal aorta down to and including
the iliac arteries (Figs 8.17 and 8.18). The polar
artery is another common variant, branching early
from the main renal artery to pass into either the
upper, or more usually the lower, pole without
traversing the renal hilum (Fig. 8.19). Multiple
renal veins are common, occurring in approxi-
mately one-third of the population – the most
common of these being an accessory left renal
vein traversing to the right behind the aorta – a
Fig. 8.14 (a) Colour flow box opened wide to
retroaortic vein (Fig. 8.20).
demonstrate global flow within the kidney.
(b) Colour flow box identifying focal flow to the
upper pole of the left kidney. The intrarenal vessels
Demonstration of the anatomical location and
course of the intrarenal vessels is almost exclu-
sively restricted to colour Doppler, although
pulsations can be seen on real time at the site of
the interlobar vessels, and occasionally at the
bright reflectors at the corticomedullary margin
which represent the arcuate vessels. The arteries
are each accompanied by a vein; they divide into
branches to the upper and lower poles and to the
anterior and posterior parenchyma (Fig. 8.14).
The interlobar vessels course into the renal
parenchyma on either side of the renal papillae,
giving tiny (invisible) branches to the medulla
before arching over the surface of the medulla as
the arcuate arteries, giving off multiple small
striate branches which extend out towards the
cortex.With more modern and sensitive machines,
Fig. 8.15 Origin of the left renal artery and its the capsular artery can occasionally be demon-
proximal course identified through an enlarged strated at the margin of the kidney, curving
spleen. around the surface. 191
8
The kidney
a
Fig. 8.16 (a) (b) Colour flow Doppler demonstrating flow towards the transducer in the left renal vein. A large
gonadal vessel is demonstrated posteriorly (g).
It is important to recognise that in the evalua-

Right renal Left renal tion of the kidneys there are factors other than the
artery artery Doppler study which contribute to the diagnosis.
Kidney Kidney
Both kidneys should be examined carefully with
respect to size, echogenicity and smoothness of
outline, together with an assessment of the corti-
comedullary differentiation. The adrenal areas
should also be carefully examined to exclude any
obvious mass.The time taken to perform a com-
plete vascular examination of the kidney is wasted
if the initial examination has missed a renal mass,
Right an adrenal tumour, or a small or absent kidney.
Accessory
testicular
renal
artery Technique – Doppler examination
artery
Inferior Throughout the course of examination of the
Aorta
mesenteric
artery renal artery, colour Doppler is frequently switched
on to confirm the nature and direction of flow.
The optimum pulse repetition frequency is selected
Right
to detect moderate flow velocities, although it
common may need to be modified to detect high velocities
iliac artery if a stenosis with aliasing of the colour signal is
present. For spectral Doppler, the sample volume
is placed on the selected vessel shown on the colour
display. Recording spectral Doppler is better
Branches of Branches of
right superior left superior performed in duplex rather than triplex mode,
vesical artery vesical artery since the processing required for triplex imaging
reduces both the frame rate and the pulse repetition
Fig 8.17 Origin of the renal arteries including an frequency, compromising further the discrimi-
192 accessory renal artery supplying the lower pole on nation of high-velocity signals at depth.
the left.
The kidney
8

a b
Fig 8.18 (a) Two right renal arteries. Oblique scan

c through the aorta and inferior vena cava demonstrating
the origin of the two renal arteries from the right
lateral wall of the aorta. (b) Transverse scan
demonstrating the larger of the two renal arteries
(in red) beneath the corresponding renal vein (in blue)
coursing towards the hilum of the right kidney.
(c) Similar transverse image of the accessory renal
artery, which is much smaller than the more superior
renal artery, but also beneath the corresponding renal
vein coursing towards the hilum of the kidney.
a b
Fig. 8.19 (a) Coronal colour flow Doppler image of the lower pole of the left kidney. An unusual vessel is
demonstrated at the lower pole. (b) Reducing the colour sensitivity the polar vessel is more clearly identified.
Colour Doppler is invaluable in the assess- the renal hilum, particularly if the angle of inci-
ment of intrarenal vessels.With the system set to dence is optimised to achieve angles of less than
detect low or moderate velocities, flow can be 60° relative to the course of the vessel. Angling
identified in almost all patients in the vessels at the probe medially from the right or left flank 193
8
The kidney
a
Fig. 8.20 (a) (b) A retroaortic left renal vein (ralv) is demonstrated in colour passing posteriorly to the aorta (a)
before it enters the left side of the inferior vena cava (ivc).
will allow assessment of the intrarenal vessels. repetition frequency; the size of the colour box
Peripheral, smaller vessels may be better demon- should therefore be minimised prior to spectral
strated with power Doppler, although in this situa- sampling (Fig. 8.23). It is frequently of value to
tion the directional information is lost.The further activate the zoom function on the machine prior
development of systems with greater colour to interrogation with colour Doppler, as this
Doppler sensitivity has significantly improved allows for greater sensitivity of colour signal
vessel identification and discrimination (Fig. 8.21). recording within the intrarenal vessels. Using
Evaluation of global blood flow requires that this technique, the hilar and interlobular vessels
the colour box be opened to its fullest extent in are demonstrated in all patients, although arcuate
order to visualise relative blood flow distribution and striate arteries are only seen in slimmer
(Fig. 8.22).This is also important in the evaluation patients. In the case of the arcuate vessels, this is
of renal tumours so that flow in the lesion can be partly due to their course, which is usually at
compared with that in normal adjacent renal right angles to the incident beam. Advances in
tissue. However, this compromises temporal
resolution with lower frame rates and pulse
Fig. 8.22 Greater sensitivity of colour flow Doppler

Fig. 8.21 Colour power Doppler demonstrating affords directional information as well as identification
194 perfusion of the renal parenchyma. of the interlobular arteries.
The kidney
8

may be associated with the nutcracker syndrome
and reflect renal venous congestion. However the
findings are not constant in this condition and
similarly may be found in asymptomatic patients
(Fig. 8.25).
Indices used in the assessment of renal

blood flow
A multiplicity of indices is used for evaluation of
Fig 8.23 Colour box reduced in size to demonstrate
renal blood flow. The wide range indicates that
single intrarenal vessel prior to sampling for spectral
no single index provides all the information that
Doppler.
is necessary for adequate evaluation of renal
pathophysiology.The indices used, together with
technology have widened the group in whom the normal values, are shown in Table 8.1, and some
entire renovascular pattern can be identified.2 examples are shown in Figure 8.26. Pulsatility
and resistance indices in the renal arteries increase
Characteristics of the Doppler spectrum with age and in hypertensive patients. This is
The typical spectra from the renal arteries are presumably consequent upon the effect of the
shown in Figure 8.24. There is a rapid systolic juxtaglomerular apparatus producing vasocon-
upstroke, which is occasionally followed by a striction, but also perhaps on the development
secondary slower rise to peak systole (although of hypertensive nephrosclerosis.3–5 A further
this is more frequently seen with advancing age, cause of variability of Doppler indices relates to
hypertension, etc.).There is subsequently a gradual differences of opinion as to the optimum site for
diastolic decay but with persistent forward flow Doppler sampling. Some authors suggest the
in diastole. renal sinus as the optimum site, while others
Renal vein spectra normally reflect the trans- recommend the interlobar and arcuate artery
mitted pulsations from the right atrium. The level.6 There is significant variety in normal
right renal vein is short and exhibits this feature spectral indices but in this author’s view this is
in all cases except where there is renal vein or best minimised by using the same site in order to
caval compromise. The left renal vein, particu- standardise measurement technique.
larly if it is sampled to the left of the superior Ureteric jets are the rhythmic pulses of urine
mesenteric artery, may show only slight variability that occur in the bladder approximately four times
of flow velocities. This pattern of damped flow each minute. These are not necessarily synchro-
b
a
Fig 8.24 (a) Normal Doppler spectrum from the right renal artery. (b) Corresponding normal Doppler spectrum
from the left renal artery using the oblique approach from the right flank. 195
8
The kidney
a b
Fig 8.25 (a) Venous flow within the right renal vein reflecting the pulsatile nature of flow within the inferior vena
cava. (b) Damped flow within the left renal vein, presumably modified by the effect of the superior mesenteric
artery. From Dubbins33, with permission.
a b
Fig 8.26 (a) Measurement of the resistance index.

c
(b) Measurement of the pulsatility index.
(c) Measurement of the acceleration time or time
to maximum systole.
Table 8.1 Normal renal artery Doppler indices
Index Range
Pulsatility index (PI) 0.7–1.4
Resistance index (RI) 0.56–0.7
Peak systolic velocity (PSV) 60–140 cm s-1 (<180)
Diastolic/systolic ratio (D/S) 0.26–0.4
Renal artery/aorta ratio (RAR) <3.5
Time to maximum systole (TMS) 42–57 ms
(systolic rise time)
196 Acceleration index 250–380 cm s-2

The kidney
8
Diseases of the major renal vessels

nous but can be identified in approximately 90% standard imaging and colour settings causes some
of patients of all age groups. Although these jets destruction of the bubbles and compromises the
can be demonstrated on real-time imaging alone, demonstration of certain perfusion characteristics
the frequency of their demonstration is much such as perfusion defects, blood pooling and
improved by the use of colour Doppler.They are early venous filling.
best demonstrated with the patient well hydrated Most manufacturers now provide specific
and with a full but not overfull bladder.The probe contrast imaging packages which rely upon the
is placed transversely in the pelvic midline, angling greater sensitivity of harmonics to the bubble
towards the feet. The trigone of the bladder is oscillation. These may have a low mechanical
identified by the two slightly raised ureteric index (MI), with minimal bubble destruction or
orifices, approximately 3cm apart. Although jets a sequenced imaging algorithm of a flash of high
may be seen in real time they are best appreciated MI to destroy the bubbles followed by an
on colour when bursts of colour signal, directed imaging sequence of low MI to demonstrate the
medially into the bladder are seen (Fig. 8.27). perfusion/reperfusion sequence.
The probe is placed in a position for optimal
Technique – contrast examination visualisation of the kidney (usually coronal to
The role of ultrasound contrast in the evaluation best demonstrate renal vessels) or for best demon-
of the kidney has yet to be evaluated. Initially stration of focal pathology. Following contrast
proposed as is a method of Doppler rescue in injection the probe is maintained in the same
difficult patients, it is now possible to evaluate position and the examination is automatically
renal perfusion and reperfusion. timed (Fig. 8.28).
For the purpose of Doppler rescue, methodology There are now available on-line or off-line time/
is straightforward. Optimum probe positioning intensity packages which allow the selection of
and angulation is selected for the vessel under regions of interest within the kidney and the record-
study and this position is maintained for contrast ing of perfusion/reperfusion graphs (Fig. 8.29).
injection. Colour or power imaging may be used
although it is usual to reduce colour sensitivity
DISEASES OF THE MAJOR RENAL
so as not to cause bloom or flash within the image.
VESSELS
Although this technique may destroy the bubbles
this is not important in the major vessels since General principles
they are immediately reperfused. Evaluation of the renal arteries should include
Renal parenchymal studies were initially an assessment of the aorta, particularly around
performed in the same way. However, the use of the origins of the renal arteries, to look for athero-
a b
Fig. 8.27 (a) Colour flow Doppler of left ureteric jet. (b) Spectral Doppler of the burst of urine into the bladder. 197
8
The kidney
a
Fig. 8.29 Time intensity curves for renal sinus,

b cortex and medullary perfusion.
and that resulting from renal artery fibro-

muscular hyperplasia. The prevalence of the
different types of renal artery disease depends to
a very large degree on the age group; fibro-
muscular dysplasia is predominantly a disease of
younger age groups, whereas atheroma becomes
more common with increasing age. Athero-
matous renal artery stenosis is the cause of the
majority of cases of renal artery stenosis. This is
particularly the case for those patients present-
Fig. 8.28 (a) Colour flow Doppler demonstrating ing with renovascular renal failure, rather than
enhancement of the colour signal within the intrarenal isolated aggressive hypertension. The findings
vessels after the injection of contrast. (b) Pulse on ultrasound imaging are different in the two
inversion harmonic image in bubble destruction
conditions.
reperfusion mode demonstrating parenchymal
In cases of atheromatous disease, the stenosis
perfusion.
tends to be located near the origin of the artery
(Fig. 8.30), although it may occur along the course
matous plaques. The demonstration of athero- of the artery. In fibromuscular hyperplasia, the
matous plaque within the more distal renal artery affected segment tends to lie distal to the origin
by direct visualisation is usually rather more diffi- of the vessel. Doppler findings within the main
ult, and its presence is inferred from alterations renal artery are variable. For a haemodynamically
in the Doppler signal. With careful evaluation significant stenosis of more than 50% diameter
along the course of the renal artery, it is occa- reduction, there is typically an increase in flow
sionally possible to demonstrate the irregular beaded velocity greater than 180 cm s-1 associated with
appearance associated with fibromuscular hyper- spectral broadening (Fig. 8.31); the renal artery/
plasia. As at other vessel sites, focal dilatation of aortic peak systolic ratio (RAR) is increased beyond
the artery may represent a poststenotic dilatation 3.5. Spectral broadening may be the only sign in
associated with a proximal stenotic segment. stenoses of lesser severity and may predominate
in fibromuscular hyperplasia.7 Colour Doppler
Renal artery stenosis findings include narrowing of the luminal diameter
The pathology of narrowing of the renal artery is and aliasing of the colour signal (Fig. 8.32). The
198 largely divided into stenosis produced by atheroma value of using the renal artery/aortic ratio, rather
The kidney
8

a
Fig. 8.30 Colour flow Doppler of the aorta in coronal b

view. There is marked atheroma of the aortic wall.
The origin of the right renal artery is demonstrated
beneath the inferior vena cava. The left renal artery
was occluded.
than absolute measurements of velocity, is that it

takes into account poor inflow; it is important to
remember that these patients are often arterio-
paths and consequently may have poor cardiac
output.8
However, initial enthusiasm for the technique
of direct evaluation of the main renal arteries
for signs of renal stenosis has been markedly
tempered by the results of more recent studies.
Some of these have indicated that the sensitivity
of detection may be as low as 20% in com-
parison to early claims of sensitivities in excess
Fig. 8.31 (a) Right renal artery stenosis. There is
of 80%.9
increased flow velocity with a liasing of the Doppler
Much of the difficulty relates to the poor
spectrum. (b) Right renal artery stenosis. Patient
imaging of the entire length of the renal artery.
presented with flash pulmonary oedema. Only a
Enhanced visualisation of the main renal arteries single heart cycle is demonstrated because of the
resulting from the use of effective ultrasound dyspnoea. Peak flow velocity within the right renal
contrast agents has been shown to improve effi- artery is 240 cm s-1.
cacy of the technique allowing Doppler rescue
but accuracy of diagnosis of renal artery stenosis
remains less than other imaging modalities such tion to the formal structured approach to the
as multislice computed tomography (CT). renal arteries, a coronal scan through the kidney,
Similarly the demonstration of accessory whose plane is directed towards the aorta, will
renal arteries with ultrasound is unreliable. allow visualisation of the renal hilum. Moving
Where a polar artery is stenosed, and is the the colour box towards the aorta, enlarging the
cause of renovascular hypertension this will be box, and/or slightly changing the plane of scan
only rarely demonstrated. Meticulous attention will allow the detection of larger accessory
to technique is vital to optimise the opportunity vessels and occasionally renal artery stenosis
for visualisation of accessory vessels. In addi- (Fig. 8.33). 199
8
The kidney
Fig. 8.33 Coronal view of left kidney. There is

Fig 8.32 Colour flow Doppler of the origin of the significant colour aliasing within a lower polar artery.
right renal artery in a patient with renal artery Diagnosis, polar artery stenosis.
stenosis. There is significant disturbance of the flow
with a mosaic pattern within the colour signal.
renal artery stenosis require visualisation of the
Findings in the intrarenal vessels main renal artery; most reports would indicate
The intrarenal vessels can be imaged in most that this is possible in at best only 70% of cases,
patients. Haemodynamically significant stenoses and that this is only marginally improved by the
will produce significant downstream effects within use of currently available ultrasound contrast,
the intrarenal arteries.These have been described because the problems of obesity and bowel gas
as a ‘tardus parvus’ pattern, with a low- remain. Significant interest has therefore centred
amplitude signal that has a prolonged systolic on the role of the indirect signs from the intrarenal
acceleration time and which appears delayed in vessels, since these measurements are simpler
relation to the QRS complex of the electro- and quicker to perform and are possible in more
cardiogram (ECG). Duplex indices used to charac- than 90% of patients. There are, however, wide
terise this waveform include a systolic acceleration variations of 60–90% in the reported sensitivity
time above 0.07 s, an increase in the acceleration of this technique, with some authors reporting
index in excess of 378 cm s-1, a reduced resist- even lower values.11,12
ance index of less than 0.5 and decreased peak The efficacy of surgical and percutaneous
systolic velocities (Fig. 8.34).10,11 In practice techniques in the treatment of renovascular disease
these changes will only reliably identify stenoses remains controversial. It is therefore important
greater than 70% diameter reduction. Further-
more, diffusely diseased arteries without a
significant focal stenosis may also show these
changes, with a significant false-positive rate.
Considerable care is therefore necessary in the
interpretation of intrarenal waveforms.
Screening for renal artery stenosis

Patients with renovascular hypertension and those
with renal failure consequent upon disease of the
main renal artery may benefit from a number of
possible revascularisation techniques such as Fig 8.34 The tardus parvus pattern of renal artery
angioplasty. Duplex ultrasound is one of a number stenosis. Assessment of the intrarenal vessels
of techniques which have been proposed as demonstrates a delay in the time to maximum
200 possible screening tools. However, direct signs of systole and an increase in the acceleration index.
The kidney
8
Aortic aneurysm

to ensure that patients are selected who are most
likely to benefit from intervention. A renal artery Ultrasound plays a central role screening for and
resistance index of 0.8 or greater reliably the diagnosis of abdominal aortic aneurysm.
identifies patients with renal artery stenosis in Although it is usually possible to demonstrate
whom angioplasty or surgery will not improve the origin of the renal arteries relative to the
renal function, blood pressure or kidney survival. aneurysm the variability of renal vascular anatomy
Where this is combined with a renal length of means that polar vessels may be missed. Where
less than 9 cm, the outcome is universally poor the ostia are involved in the aneurysm this can
(Fig. 8.35). be demonstrated and flow abnormality detected
While we would not, therefore advocate the (Fig. 8.36).
use of renal Doppler as a screening test for renal
artery stenosis the technique retains a role in Aortic dissection
patient selection, in the evaluation of equivocal Ultrasound is of value in assessing dissection of
angiographic findings and follow-up after the abdominal aorta. Although this is significantly
angioplasty.13 less common than thoracic aortic dissection, it
can occur as a consequence of the extension of a
Renal artery aneurysm thoracic dissection into the abdominal aorta, or
Renal artery aneurysms are uncommon. Previously subsequent to interventional procedures performed
on static B-mode imaging, mistaking the dilated via the transfemoral route.The aortic flap can be
left renal vein, produced by the nutcracker effect identified on real-time ultrasound, while flow
on the vein by the superior mesenteric artery, within both the true and false lumen can be docu-
led to false diagnoses of renal artery aneurysm. mented. Flow may be towards the legs in both
With modern real-time equipment an aneurysm channels, or may be reversed in the false lumen
can almost always be identified in continuity with (Fig. 8.37). The course of the dissection is often
the renal artery. Colour Doppler will demonstrate spiral; consequently one renal artery may be
the flow direction and characteristic venous supplied by the true lumen of a dissection and
pulsatility in the case of a prominent vein, or the the other by the false lumen. Documentation of
disordered arterial flow associated with an impaired flow within one or other of the renal
aneurysm. arteries, or in the intrarenal vessels, is of
Fig. 8.36 Distal abdominal aortic aneurysm.

Fig. 8.35 Left renal artery Doppler demonstrating Same case as Figure 8.30. The renal arteries are not
increased systolic flow velocity in excess of 2 m s-1 compromised by the aneurysm, but the origin of the
but reduced diastolic flow and consequent increased left renal artery is identified. There is no colour flow.
resistance index. Renal vascular compromise caused by atheroma. 201
8
The kidney
on colour Doppler as a mosaic of colour extending
into the renal parenchyma around the fistula
(Fig. 8.38).15
Atheroembolic disease
Embolism of atheroma within the renal vascular
tree is a cause of significant renal dysfunction,
producing wedge-shaped infarcts and subsequent
scarring. Theoretically colour Doppler, and in
particular power Doppler with its increased sensi-
tivity to low-flow states, should be able to provide
a global view of renal perfusion (Fig. 8.39).
While it is possible with both colour and power
Fig 8.37 Colour flow Doppler examination of the Doppler to show focal ischaemia and focal infarcts,
aorta in a case of aortic dissection. There is reverse both techniques suffer from limitations of
flow in the false lumen which modifies flow within inadequate penetration to the posterior kidney.
the left renal artery. The reverse flow is demonstrated Therefore focal colour-free segments in this area
(in blue) underneath the linear echogenic dissection cannot reliably be defined as infarcts, as the
flap. From Dubbins33, with permission. features may simply be the result of technical
difficulties with colour sensitivities. Ultrasound
significant prognostic value in cases such as this, contrast has been advocated as a tool to resolve
since it will identify a kidney at risk of ischaemia. some of these difficulties.The blood pool contrast
In such cases, revascularisation may be agents certainly demonstrate tissue perfusion,
considered.14 even in difficult patients, but gas microbubbles
may produce significant sound attenuation, which
Arteriovenous malformation/fistula will degrade the signal from the more deeply
In the native kidney these are almost invariably
post-traumatic following penetrating injury.
Small arteriovenous fistulae following renal biopsy
are usually short lived, although when large they
can compromise renal function and renal survival.
These are described in the section on renal
transplantation (see Ch. 9). Larger arteriovenous
fistulae can be encountered following trauma
such as knife injuries. In these cases, both the
main renal artery and the main renal vein may
enlarge, with increased flow velocities, decreased
pulsatility and resistance indices within the renal
artery, and high flow velocities within the renal
vein.
The site of the arteriovenous communication
is identified by reducing the colour sensitivity, so
that only high velocity is detected. In this situation Fig 8.38 Arteriovenous fistula. Colour flow Doppler
no flow is demonstrated within the normal renal demonstrates increased focal flow within the kidney,
parenchyma but there remains a high-intensity with a tissue bruit of colour out with the renal
colour signal at the site of the fistula. This is vessels caused by vibration of the adjacent renal
202 frequently accompanied by a tissue ‘bruit’ visible substance.
The kidney
8

extrinsic compression or haematological condi-
tions which produce an increased tendency for
thrombosis may predispose to renal vein throm-
bosis. In infants, renal vein thrombosis may
occur as a complication of dehydration.
Colour Doppler may show no flow in the
renal parenchyma, or ‘flash flow’ where the colour
signal is seen only transiently, corresponding to
peak systole (Fig. 8.41). Occasionally a ‘to-and-
fro’ pattern may be seen with alternating red
and blue colour signals visible within the renal
arteries. Spectral Doppler may show absent, or
reversed, diastolic flow. A ‘W’ shape to the
arterial diastolic flow component, in association
Fig 8.39 Colour power Doppler evaluation of a with absent venous flow on colour Doppler, is
patient with focal ischaemia of the kidney. There is thought to be characteristic of renal vein throm-
normal perfusion in the lower and mid-poles but the bosis, although other conditions may produce
anterior surface of the upper pole is avascular. reversed diastolic flow, including acute tubular
necrosis and cardiac abnormalities such as
aortic incompetence (Fig. 8.42).17
situated renal parenchyma with the same potential
for perfusion defect artefacts.With newer contrast Renal infection
agents it appears possible to maximise Doppler The findings in renal infection are extremely
enhancement while minimising the shadowing variable. Infection may be consequent upon the
effect16 (Fig. 8.40). ascending route via the ureter or may be blood
borne. Parenchymal infection has been described
Renal vein thrombosis as global or focal although pathological studies
This is an uncommon condition in adults; it would suggest that these patterns are part of a
occurs most commonly in association with the continuum of disease. Imaging findings include
extension of a renal tumour (see below). Rarely, an enlarged kidney, focal masses of varying
Fig. 8.40 (a) Focal renal infarct. Kidney appears structurally normal. There is possibly slight decrease in
echogenicity of the lower pole. From D.L Cochlin, with permission. (b) Following contrast the lower pole is not
perfused. 203
8
The kidney
and an increase in the Doppler indices. In focal
pyelonephritis, this may allow the demonstration
of a hypoechoic mass with an associated perfu-
sion defect (Fig. 8.43).18 Contrast enhanced
studies using low MI techniques allow real time
demonstration of renal perfusion (Fig. 8.44).
When combined with contrast specific imaging
packages such as those incorporating tissue
harmonics or pulse inversion it is possible to
demonstrate focal perfusion defects which may
correlate with the subsequent development of
renal scars. These techniques remain to be fully
evaluated. They are dependent on body habitus.
Low MI will afford only modest penetration of
the sound beam and although there is significant
Fig 8.41 Flash flow of renal vein thrombosis. signal enhancement by microbubbles, sound
Poverty of flow to the kidney with only brief attenuation by fat and by the bubbles themselves
demonstration of flow within the central vessels. may make it difficult to assess deep structures.
This has particular importance for false-positive
echogenicity, capsular thickening (the renal rind) diagnosis of perfusion defects.
and abscess formation. There may be dilatation
of the collecting system with thickening of the Renal abscess
urothelium. The development of a renal or perirenal abscess
Blood flow findings depend upon the influence is an uncommon complication of renal infection.
of the balance between hyperaemia consequent The appearances are, however, fairly characteristic,
upon infection, and reduced flow consequent upon with a renal mass lesion showing central necrosis
renal swelling and capsular stretching.Thus there subsequently developing into an irregular cystic
may be a global or focal increase in blood flow, lesion with thick walls. Blood flow is usually
with a reduction in the Doppler indices; or, more increased around the margin of the lesion,
commonly, there may be reduction in blood flow producing a ‘colour halo’, although this is indis-
a b
Fig 8.42 (a) Renal vein thrombosis. Spectral pattern

demonstrating the W-shaped flow pattern in diastole
characteristic of renal vein thrombosis. (b) Reversed
flow in diastole consequent upon aortic incompetence.
204 Image courtesy of Dr D.L. Cochlin.
The kidney
8

of the success of colour Doppler in distinguishing
between malignant and benign pathology of the
kidney.19–21 Comparisons with contrast angiography
suggest that the accuracy of the technique is likely
to be less than has been hoped, as both benign
and malignant tumours can be either hypo- or
hypervascular. Angiomyolipomas, for example,
can be extremely vascular and demonstrate blood
pooling while, conversely, transitional cell tumours
are characteristically hypovascular.22 However, in
the evaluation of cystic lesions of the kidney the
assessment of complexity can be improved by the
demonstration of colour flow within septa which
Fig 8.43 Colour flow Doppler image demonstrating
blood flow at the hilum and within the kidney. is a strong predictor of malignancy (Fig. 8.46).
Note there is absent flow to the upper pole in this Examination of renal tumours with colour
patient with focal pyelonephritis. Doppler requires the colour box to be opened to
include the whole of the tumour and a small
tinguishable from a necrotic renal tumour. portion of the adjacent normal renal parenchyma,
Usually a combination of clinical history and so as to demonstrate the difference in size, distri-
grey-scale features, as well as colour Doppler, bution and communication of the abnormal
will allow the correct diagnosis (Fig. 8.45). blood vessels. High- and low-velocity blood flow
patterns have been described, predominantly in
Renal tumour the margin of the tumour but also extending within
Angioneogenesis occurs as part of the disordered the tumour mass. Characteristic colour flow
growth of renal tumours. As a consequence it patterns are of large irregular marginal vessels
was felt that both colour Doppler and spectral extending into the centre of the tumour with an
Doppler might provide a means of assessment of abnormal irregular branching pattern.The pattern
renal tumours. However, reports are varied in terms of flow is extremely variable. Some tumours are
b
a
Fig. 8.44 (a) Focal pyelonephritis. The lower pole of the left kidney is enlarged. There is a possible area of focal
ischaemia at the extreme lower pole. From D.L. Cochlin, with permission. (b) The entire kidney appears poorly
perfused after contrast injection, but with irregular perfusion defects from the equatorial region to the lower
pole. Only the upper pole and parts of the lower pole appear well perfused. 205
8
The kidney
Fig. 8.46 Benign cystic nephroma of the kidney.

There is no evidence of blood flow within the septa
Fig 8.45 Renal abscess. Colour flow Doppler of the of the tumour. From D.L. Cochlin, with permission.
kidney. There is an irregular area within the upper
pole of the kidney. There is no flow within the centre
of the lesion but circumferential flow.
a b
c
Fig. 8.47 (a) Colour flow Doppler of a lower pole
renal mass demonstrating irregular vessels and
colour pooling. (b) Colour power Doppler of renal
tumour with irregular peripheral flow and central
necrosis. (c) Colour flow Doppler of tumour with
relatively avascular appearance and only peripheral
flow.
206
The kidney
8

relatively avascular, while others exhibit areas of vascularisation of septa with improved confidence
necrosis producing largely circumferential, irregular of the diagnosis of malignancy. The pattern of
flow (Fig. 8.47). contrast enhancement allows the differentiation
Contrast enhanced ultrasound affords the of tumour from pseudomasses such as the hyper-
greater characterisation of intratumoral blood trophied column of Bertin, as well as affording the
flow. There is improved sensitivity of demons- characterisation of specific renal masses in some
tration of irregular marginal vessels and of neo- cases (Fig. 8.48). The use of three-dimensional
a b
c
d
Fig. 8.48 (a) Hypertrophied column of Bertin

mimicking a tumour. (b) Following contrast the pattern
e
of enhancement is similar to surrounding normal
renal parenchyma. (c) Lower pole renal tumour
demonstrating irregular enhancement following
contrast. (d) Large tumour at the lower pole of the
kidney demonstrating contrast enhancement in
septa within an otherwise hypovascular tumour.
(e) Angiomyolipoma, dense enhancement following
contrast. The pattern of predominantly peripheral
enhancement is not pathognomonic. (b), (c) and (e)
From D.L. Cochlin, with permission.
207
8
The kidney
power Doppler, with and without the use of (Fig. 8.49). Colour Doppler demonstrates flow
contrast agents, is being explored in an attempt around the thrombus or tumour thrombus. It is
to characterise blood flow patterns specific for only rarely possible to identify blood flow within
benign and malignant tumours, but this work the tumour thrombus itself and thus distinguish
remains in its infancy. Spectral Doppler may between tumour and non-tumour thrombosis.
show scattered high-frequency signals both at In total occlusion of the renal vein, no colour flow
the margin and within the body of the tumour, signal is identified and the spectral flow within
presumably representing flow through vessels the renal artery demonstrates a high-resistance
supplying arteriovenous communications.21 In flow pattern with absent or reversed flow in
very large hypervascular tumours with extensive diastole (Fig. 8.50).
arteriovenous shunting, blood flow within the
main renal artery is abnormal with increased Obstruction
systolic flow velocities and spectral broadening. The intravenous urogram once the mainstay of
Colour Doppler may refine the ability of real- diagnosis of acute obstruction is rapidly being
time scanning to evaluate invasion/thrombosis of replaced by non-contrast enhanced CT. Ultra-
the renal vein.When thrombus is present the renal sound has long had a role in the demonstration
vein is distended and contains low-level echoes of hydronephrosis, particularly in long-standing
Fig 8.49 (a) Transverse scan through the left renal

c vein. The left renal vein is distended and filled with
echogenic material, indicating either extension of
tumour into the vein or renal vein thrombosis.
(b) Same patient – extension of the echogenic
thrombus/tumour into the inferior vena cava.
(c) Distortion of the inferior vena cava and encasement
of the adjacent right renal artery by tumour extending
beneath the inferior vena cava from the adjacent
right kidney.
208
The kidney
8

Fig 8.50 Absent diastolic flow in a patient with renal
vein involvement by tumour. Fig. 8.51 Renal obstruction. There is significant
reduction in the diastolic component flow with a
resistance index of 0.91.
obstruction from whatever cause. Features of
obstruction on real-time imaging include dilata-
tion of the calyces and of the renal pelvis; dilated (Fig. 8.51).23 The findings are time dependant.
ureters can also be identified posterior to the There is usually no detectable Doppler abnor-
urinary bladder. In acute obstruction, however, mality in the first 6 h and the change in Doppler
there is often no evidence of calyceal or pelvic spectra gradually returns to normal after 48 h.
dilatation and the kidney may appear structurally However, not all workers have found Doppler
normal on ultrasound. In this situation, repeat ultrasound to be of value in the diagnosis of
ultrasound after 8–12 h will usually demonstrate obstruction; sensitivities vary from 44 to 92%.24, 25
the development of pelvicalyceal dilatation. This author’s own experience suggests that in
However, there are alterations in intrarenal patients in whom an intravenous urogram is
pressure consequent upon the obstruction and contraindicated, such as pregnancy, the combi-
these may be detected by the use of Doppler nation of Doppler indices with the finding of
techniques.Thus in acute obstruction an increase asymmetric ureteric jets in the bladder is a reliable
in the pulsatility index and in the resistance index discriminator of ureteric obstruction, and is a
has been reported. Resistance indices of greater useful screening test to determine which patients
than 0.7, or with the difference between the two should proceed to contrast radiological studies
sides of greater than 0.05, imply renal obstruction (Fig. 8.52).26
Fig. 8.52 (a) Normal ureteric jet. (b) Poor flow through the right ureteric orifice on colour flow Doppler in a right
ureteric calculus. 209
8
The kidney
Although entry of the urine into the bladder use of an intravenous diuretic, such as frusemide,
is not usually synchronous, the demonstration of may increase the sensitivity for detection of obstruc-
three or more jets on one side without a single tion. If the ureter is obstructed, the administration
jet identified on the other side, or an abnormally of a diuretic will often produce a more marked
directed jet with an abnormal rather diffuse colour alteration in the diastolic flow component of the
signal, implies obstruction of the ipsilateral waveform, and therefore a more marked increase
ureter. in the resistance index.27 In addition, diuresis will
The accuracy of ultrasound findings including increase the frequency with which normal ureteric
Doppler and jet detection is greatest in patients jets are identified. This would therefore enhance
with acute and complete obstruction and is the detection of abnormalities of ureteric jet
diminished in patients where obstruction is production. Similar effects occur in response to
incomplete. a fluid load, and some workers prefer this approach,
particularly in children. Diuresis sonography,
Diuretic enhanced sonography however, is only indicated when there is a strong
In cases of renal obstruction, including pelviureteral contraindication to radiological contrast agents
junction obstruction and calculus obstruction, the or to ionising radiation.
b
a
c d
Fig. 8.53 (a) Diffuse parenchymal renal disease. The patient presented with raised creatinine. There is absent
diastolic flow. (b) and (c) Power Doppler and spectral Doppler in a patient with early diabetic nephropathy.
The kidney is well perfused. (d) Patchy colour perfusion of a patient with diffuse parenchymal renal disease.
210 Current resolution does not afford the clear demonstration of intrarenal shunting.
The kidney
8
Renal trauma

In kidneys with end-stage disease from what-
In cases of renal trauma, CT is normally performed ever cause, it is often difficult to detect any blood
on patients in whom renal injury is likely but flow with colour, power Doppler, or with spectral
ultrasound may be used instead. Colour Doppler Doppler. When detectable, Doppler spectra are
can demonstrate reduced or absent flow to the of low systolic amplitude with poor or absent
renal parenchyma.While this will not differentiate diastolic flow.
between vascular transection and severe spasm, Doppler is therefore of limited use in diffuse
it does suggest the need to consider angiography. renal disease. The specificity of the technique in
It is important to optimise colour Doppler settings differentiating between the various causes of
for the detection of low flow, and therefore renal dysfunction is not sufficient to make this of
institute the lowest high-pass filter together with clinical use and allow avoidance of biopsy.29,30
a low pulse repetition frequency while ensuring There is an association between deteriorating
that flash artefact is minimised. serum creatinine and rising Doppler indices but
this is not sufficiently reliable to allow the use of
Diffuse renal disease Doppler to monitor the disease process.
There are many different causes of diffuse renal In a small kidney, the demonstration of absent
disease which may affect glomerular integrity, diastolic flow within the intrarenal vessels is strong
the tubulointerstitial region of the medulla, or evidence that this is consequent upon diffuse
the vessels directly. For the most part, in purely renal disease, or small vessel disease. This may
glomerular disease of any aetiology, the grey- be important for differential diagnosis. A small
scale image and the Doppler spectrum are unaf- kidney consequent upon renal artery stenosis that
fected until late in the disease process. In renal may respond to revascularisation will exhibit a
disease of tubulointerstitial distribution, or that tardus parvus pattern within the intrarenal vessels
consequent upon a vasculitic process, there is (Fig. 8.34).31
frequently an increase in cortical echogenicity with
accompanying enhancement of corticomedullary Contrast media
differentiation. Concomitant with this is an increase The advent of third-generation ultrasound contrast
in resistance to blood flow with an increase in the media now allows the demonstration of blood
resistance and pulsatility indices (Fig. 8.53).28 pool enhancement in the kidney. Currently, the
Table 8.2 Indications for renal Doppler
Investigation/purpose Possible cause/indication

1. To confirm renal perfusion Renal trauma
Reduction in renal size, etc.
2. Diagnosis of renal vein thrombosis Tumour
Dehydration (infant)
Hypovolaemia
Abnormal clotting
3. Renal obstruction Contrast hypersensitivity
Pregnancy
4. Renal tumour vs pseudotumour
5. Renal artery stenosis Equivocal angiogram
Follow-up to angioplasty,
etc.
(Screening)
6. Arteriovenous abnormalities
7. Aortic aneurysm and aortic dissection
211
8
The kidney
duration of enhancement by commercially
SUMMARY
available ultrasound contrast media is only short
lived, being of the order of 2–3 min after bolus Doppler techniques now allow the assessment of
injection; this may possibly be extended by the intrarenal vessels in most patients (Table 8.2).
use of infusion techniques. No well defined clinical Currently, however, the role of Doppler in the
role has as yet been established for contrast media kidney is limited. It may help to resolve problems
in the enhancement of Doppler signals from the with mass lesions, including the differentiation
kidney, although it is possible to demonstrate of tumours from pseudotumours such as a
perfusion defects in focal pyelonephritis and hypertrophied column of Bertin. It may allow
Doppler rescue is possible in a number of cases the assessment of renal obstruction when there
where the main renal artery is interrogated. is a contraindication to X-ray contrast agents. It
Further roles for ultrasound in evaluation of the may clarify the diagnosis of renal artery stenosis
entire renal vascular supply, the demonstration in the case of an equivocal angiogram, and it will
of abnormal vascular anastomoses in tumours and allow the identification of significant renal venous
in the measurement of global and regional volume disease. The greater sensitivity of modern
flow will depend upon extended enhancement ultrasound systems combined with improved
intervals, improved algorithms for flow detection visualisation with echo-enhancing agents may
and computer packages for flow and perfusion allow visualisation and assessment of the entire
calculation. renal vascular supply in the future.
REFERENCES
1. Dubbins P. Renal artery stenosis – duplex Doppler 9. Berland LL, Koslin DB, Routh WD, et al.
evaluation. Br J Radiol 1986; 59:225–229. Renal artery stenosis: prospective evaluation
2. Taylor GA, Ecklund K, Dunning PS. Renal cortical of diagnosis with colour duplex US compared
perfusion in rabbits: visualization with colour with angiography: work in progress. Radiology
amplitude imaging and an experimental microbubble- 1990; 174:421–423.
based US contrast agent. Radiology 1996; 10. Handa N, Fukunaga R, Etani H, et al. Efficacy of
201(1):125–129. echo Doppler examination for the evaluation of
3. Keogan MT, Kliewer MA, Hertzberg BS, et al. renovascular disease. Ultrasound Med Biol 1988;
Renal resistive indexes: variability in Doppler US 14:1–15.
measurement in a healthy population. Radiology 11. Stavros AT, Parker SH, Yakes WF, et al. Segmental
1996; 199(1):165–169. stenosis of the renal artery: pattern recognition of
4. Gill B, Palmer LS, Koenigsberg M, et al. Distribution parvus and tardus abnormalities with duplex
and variability of resistive index values in undilated sonography. Radiology 1992; 184:487–492.
kidneys in children. Urology 1994; 44(6):897–901. 12. Strunk H, Jaeger U, Teifke A. Intrarenal colour
5. Boddi M, Sacchi S, Lammel RM, et al. Age-related Doppler ultrasound for the exclusion of renal artery
and vasomotor stimuli-induced changes in renal stenosis in cases of multiple renal arteries: analysis
vascular resistance detected by Doppler ultrasound. of the Doppler spectrum and tardus parvus
Am J Hypertens 1996; 9(5):461–466. phenomenon. Ultraschall Med 1995; 16:172–179.
6. Knapp R, Plotzeneder A, Frauscher F, et al. 13. Baxter GM, Aitcheson F, Sheppard D, et al. Colour
Variability of Doppler parameters in the healthy Doppler ultrasound in renal artery stenosis,
kidney: an anatomic-physiologic correlation. intrarenal waveform analysis. Br J Radiol 1996;
J Ultrasound Med 1995; 14(6):427–429. 69:810–815.
7. Taylor DC, Kezzler MD, Moneta GL, et al. Duplex 14. Thomas E, Dubbins PA. Duplex ultrasound of the
ultrasound scanning in the diagnosis of renal artery abdominal aorta – a neglected tool in aortic
stenosis: a prospective evaluation. J Vasc Surg dissection. Clin Radiol 1990; 42:330–334.
1988; 7:363–369. 15. Ozbek SS, Memis A, Killi R, et al. Image-directed
8. Strandness DE. The renal arteries. In: Strandness and colour Doppler ultrasonography in the diagnosis
DE, ed. Duplex scanning in vascular disorders, 2nd of postbiopsy arteriovenous fistulas of native
212 edn. New York: Raven Press; 1993:197–215. kidneys. J Clin Ultrasound 1995; 23(4):239–242.
16. Correas J-M, Kessler D, Worab D, et al. The first
The kidney
25. Tublin ME, Dodd GD, Verdile VP. Acute renal colic:
8
References
phase-shift ultrasound contrast agent: Echogen. diagnosis with duplex Doppler US. Radiology 1994;
In: Goldberg BB, ed. Ultrasound contrast agents. 193:697–701.
London: Martin Dunitz; 1997:101–120. 26. Weston MJ, Dubbins PA. The diagnosis of
17. Tublin ME, Dodd GD. Sonography of renal obstruction: colour Doppler ultrasonography of renal
transplantation. Radiol Clin North Am 1995; blood flow and ureteric jets. Curr Opin Urol 1994;
33:447–459. 4:69–74.
18. Winters WD. Power Doppler sonographic evaluation 27. Renowden SA, Cochlin DL. The effect of intravenous
of acute pyelonephritis in children. J Ultrasound frusemide on the Doppler waveform in normal
Med 1996; 15:91–96. kidneys. J Ultrasound Med 1992; 11:65–68.
19. Hirai T, Ohishi H, Yamada R, et al. Usefulness of 28. Platt JF, Rubin JM, Ellis JH. Intrarenal arterial Doppler
colour Doppler imaging in differential diagnosis sonography in patients with non-obstructive renal
of multilocular cystic lesions of the kidney. disease: correlation of resistive index with biopsy
J Ultrasound Med 1995; 14(10):771–776. findings. Am J Radiol 1990; 154:1223–1227.
20. Erden I, Beduk Y, Karalezli G, et al. Characterization 29. Argalia G, d’Ambrosio F, Mignosi U, et al. Doppler
of renal masses with colour flow Doppler echography and colour Doppler echography in the
ultrasonography. Br J Urol 1993; 71(6):661–663. assessment of the vascular functional aspects of
21. Yamashita Y, Takahashi M, Watanabe O, et al. Small medical nephropathies. Radiol Med 1995;
renal cell carcinoma: pathologic and radiologic 89(4):464–469.
correlation. Radiology 1992; 184(2):493–498. 30. Yoon DY, Kim SH, Kim HD, et al. Doppler sonography
22. Horstman WG, McFarland RM, Gorman JD. Colour in experimentally induced acute renal failure in
Doppler sonographic findings in patients with rabbits. Resistive index versus serum creatinine
transitional cell carcinoma of the bladder and renal levels. Invest Radiol 1995; 30(3):168–172.
pelvis. J Ultrasound Med 1995; 14(2):129–133. 31. Forsberg F, Goldberg BB. New imaging techniques
23. Platt JF, Rubin JM, Ellis JH. Distinction between with ultrasound contrast agents. In: Goldberg BB,
obstructive and non-obstructive pyelocaliectasis ed. Ultrasound contrast agents. London: Martin
with duplex Doppler sonography. Am J Radiol 1989; Dunitz; 1997:177–191.
153:997–1000. 32. Cosgrove D. Ultrasound contrast enhancement of
24. Platt JF. Duplex Doppler evaluation of native kidney tumours. Clin Radiol 1996; 51(suppl. 1):44–49.
dysfunction: obstructive and non-obstructive 33. Dubbins P. Urogenital ultrasound: a text atlas.
diseases. Am J Radiol 1992; 158:1035–1042. London: Martin Dunitz; 1998.
213
9
Doppler ultrasound
evaluation of
transplantation
Myron A. Pozniak
INTRODUCTION Examination of the transplanted kidney

requires careful attention to scan technique and
Graft and patient survival rates following solid an awareness of the potential problems. A com-
organ transplantation are progressively improving plete sonographic examination of the renal trans-
due to refinements in surgical technique, advances plant should cover the points listed in Table 9.1.
in human leukocyte antigen (HLA) typing for The most common abnormal findings, which
recipient/donor matching,1 new and greatly may be demonstrated, are listed in Table 9.2.
improved immunosuppressive agents,2 wide
acceptance of the national coordinated organ Ultrasound anatomy of the renal
sharing system, and advances in non-invasive transplant
transplant monitoring. As the number of trans- In most cases, the transplant kidney is positioned
plant recipients grows, there is a corresponding retroperitoneally in the iliac fossa with an end-
increase in these patients presenting for emergency to-side anastomosis of the renal vasculature to
care to non-transplant centres. The chances for the common or external iliac artery and vein.
graft survival can be significantly improved with The transplanted ureter is implanted directly
timely identification of the aetiology of transplant
dysfunction, allowing prompt medical and/or
surgical intervention when necessary. Table 9.1 Renal transplant sonographic
examination checklist
RENAL TRANSPLANTATION • Review any available prior imaging studies.

Review the surgical record, especially with regard
In the USA 1-year renal allograft survival rates to the vasculature
currently approximate 89% for cadaveric and 95% • Evaluate the renal collecting system. If dilated,
make certain that bladder outflow obstruction is
for living-donor kidney transplants, with 1-year not the underlying cause
patient survival rates at 94% for cadaveric recipients • Measure the renal length. Record any change
and 97% for living-related kidney recipients.3 • Rule out perinephric fluid collections. Record any
When screening laboratory test results indicate change in size if previously present
renal transplant dysfunction, imaging studies are • Rule out lymphocele
often required to evaluate renal morphology and • Verify uniform parenchymal perfusion by colour
Doppler. Rule out tardus parvus waveform by
perfusion. Doppler ultrasound is an ideal tool for examining the interlobar or segmental arterial
this purpose because of its non-invasive nature, waveforms for resistance and delayed systolic
upstroke
ready availability, and ability to detect and distin-
• Examine the main renal artery, particularly near
guish many of the vascular abnormalities that are its anastomosis (especially if a tardus parvus
waveform is observed within the transplanted
found with transplant dysfunction. Occasionally, kidney)
ultrasound can suggest functional problems, such • Verify renal vein patency
as rejection or acute tubular necrosis. 215
9
Doppler ultrasound evaluation of transplantation
Table 9.2 Renal transplant sonographic findings

and possible causes
Increase in size of transplanted kidney

• Hypertrophy of the kidney
• Allograft rejection
• Postoperative infection R
A
• Renal vein thrombosis
Reduction in size of transplanted kidney V
• Ischaemia
• Chronic rejection
Increased renal arterial flow resistance
• Compressive effect by transducer, adjacent mass
or fluid collection
• Infection B
• Advanced stages of rejection
• High-grade obstruction
Fig. 9.1 Artist rendering of a renal transplant (R)
• Acute tubular necrosis
located in the right iliac fossa. The transplant renal
Decreased renal arterial flow resistance
artery is typically anastomosed to the common iliac
• Renal artery stenosis
artery (A). The transplant renal vein anastomosis is
• Severe aortoiliac atherosclerosis
to the common iliac vein (V). The ureter is connected
• Arteriovenous fistula
to the urinary bladder (B).
Renal collecting system dilatation
• Obstructive hydronephrosis
• Ureteral anastomosis stenosis location of anastomoses, and any other atypical
• Chronic distention of flaccid denervated system anatomical information. If the patient has under-
• Sequela of prior obstructive episode gone a simultaneous renal–pancreas transplant,
• Bladder outlet obstruction (neurogenic bladder) then the kidney is usually found in the left iliac
fossa.
into the superior surface of the bladder or to the Doppler ultrasound technique
native ureter (Fig. 9.1). In approximately 20% of Successful Doppler evaluation of the transplanted
transplants, multiple arterial or venous anasto- kidney can only be accomplished when scan
moses may be required. Because numerous quality is optimal.This requires the use of equip-
technical variations exist in the way kidneys are ment that provides high Doppler sensitivity. The
transplanted, it is very important that the examiner must optimise the Doppler settings,
sonologist and sonographer are familiar with the since improper adjustment can result in slow flow
surgical technique common to their institution being overlooked and thrombosis being incorrectly
and the specific anatomical details of the patient diagnosed. Two particularly important scanning
being scanned.4,5 With the numerous possible factors that must be taken into account to ensure
variations, proper documentation and commu- a successful examination of the renal transplant
nication of the surgical record is very important are optimising the angle of insonation relative to
in ensuring correct understanding and interpre- vessel orientation of the renal vascular pedicle,
tation of imaging findings and Doppler flow and maintaining minimal transducer pressure.
profiles in renal transplantation. Ideally, if the Ultrasound examination of the kidney should
transplant has variant vascular anatomy, a draw- first confirm its appropriate location in the
ing is provided which shows the orientation of absence of any significant fluid collections. Colour
216 the kidney and its vasculature, the number and Doppler is then used to identify the renal vascular
9
Renal transplantation
pedicle. Spectral Doppler is applied to the main flow in a vessel. Initially, filtration should be set
renal artery, main renal vein and intrarenal at the lowest possible level and only increased
segmetal or intralobar branches at the mid, incrementally when the low setting does not
upper, and lower poles. If any inflow compro- allow for an effective examination.
mise is suspect, then power Doppler can be
applied to confirm uniform vascular perfusion Optimising angle of insonation relative
throughout the kidney (Fig. 9.2). to vessel orientation
To ensure proper perception of flow by colour
Scale setting (pulse repetition frequency) Doppler, or an accurate display of the spectral
For the initial scan, the colour and spectral velocity, the angle of insonation should be less
Doppler scales should be set as low as possible. than 60°. Finding an appropriate angle can be
By doing so, the examiner will be able to localise especially problematic when examining trans-
the vessel in question with colour Doppler and planted kidneys because their vessels may be
then demonstrate adequate excursion on the extremely tortuous and a committed search for
spectral Doppler tracing. If aliasing occurs, the a suitable Doppler window is required.6
examiner can always increase the scale setting
until the optimal level for that particular vessel is Minimising transducer pressure
achieved. Often the imaging study is limited because inter-
vening adipose tissue increases the distance from
Doppler gain the patient’s skin to the transplanted kidney or
The gain should be set at the highest level there is gas in the overlying bowel. By applying
possible without creating noise in the image or sufficient pressure, fat or bowel loops can be
tracing. displaced. Doing so, however, will compromise
the Doppler examination as the renal parenchyma
Filtration level also becomes compressed and inflow during
The Doppler filter reduces noise in both colour diastole can be impeded (Fig. 9.3). This results
and spectral modes. If the filtration level is set in a perceived elevation of the resistive index.
too high, it can eradicate the display of very slow Thus, care must be taken not to apply pressure
to the kidney or its vessels, so that any diagnosis
made on the basis of the resistive index or
velocity measurement is more accurate.7
Complications of renal
L transplantation
A
I Functional complications
These include hyperacute rejection, perioperative
ischaemia, acute tubular necrosis, acute rejection,
chronic rejection and drug toxicity (most com-
monly immunosuppressive agents).8,9 Imaging
techniques, including ultrasound with Doppler,
are limited in their ability to identify and dis-
Fig. 9.2 Longitudinal power Doppler image of a
normal renal transplant. With the setting set to a very
tinguish these functional complications.10
sensitive level, flow can be perceived in the interlobar With hyperacute rejection (humeral mediated
(I), arcuate (A), and intralobular (L) arteries all the way rejection), graft failure occurs rapidly (within
out to the capsular surface of the kidney. This is a minutes of implantation) secondary to the
normal finding and indicates good renal perfusion presence of preformed circulating antibodies.
with normal resistance to inflow. This condition is typically observed in patients 217
9
a
b
Fig. 9.3 (a) Renal transplant interlobar artery spectral Doppler tracing acquired with gentle transducer contact.
Note the normal waveform and 65% resistive index. (b) When moderate pressure is applied by the transducer,
the tracing of the same artery now exhibits an elevated resistive index of 100%. Transducer pressure alone is
responsible for this increase in vascular impedance and resultant elevation of resistance.
who have been sensitised by a previous trans- are counteracted by intrarenal hormonal auto-
plant organ or a blood transfusion. This diag- regulatory mechanisms. Elevation of resistive
nosis is usually made in the operating suite, within index, therefore, does not manifest until the
minutes of unclamping the vascular pedicle. process of acute rejection is quite severe.11 If a
Since the renal transplant is so short-lived, ultra- scan being performed in anticipation of trans-
sound has little role in the evaluation of this plant biopsy identifies the kidney to be oedema-
condition. Extremely high resistance to inflow tous and swollen with loss of central sinus fat
can be expected. echo and very high resistive indices, thought
Acute rejection is a cellular-mediated process, should be given to deferring the biopsy. Punc-
whereby the immune system attacks the foreign turing the capsule of a tense rejecting kidney
renal allograft. Acute rejection is controlled by may cause it to rupture.
the use of steroids, cyclosporine, tacrolimus, Resistive index is rarely affected in the mild to
sirolimus and other immunosuppressive agents. moderate stages of acute rejection and when it
Occasional elevation in a transplant recipient’s is, its specificity is low.12 It is not until acute
immune status (caused by viral illness or non- rejection progresses to severe levels that the
compliance with immunosuppressive drug resistive index becomes consistently elevated.
therapy) can result in an acceleration of acute Elevation of resistive index, however, can also
rejection to a critical level. The kidney becomes occur from many other causes such as hydro-
oedematous and swollen, intracapsular pressure nephrosis, acute tubular necrosis, infection and
rises, and eventually resistance to vascular perfu- compression of the kidney by an adjacent mass
sion increases (Fig. 9. 4). Although early inves- or fluid collection. Thus, specificity for the diag-
tigators proposed that resistive index elevation nosis of acute rejection by Doppler ultrasound is
was useful in identifying acute rejection as the unacceptably low and renal biopsy is still needed
cause of kidney transplant dysfunction, subse- to establish the diagnosis.11–16
quent laboratory and clinical studies have shown Chronic rejection is primarily a multifactorial
it to be unreliable, and it remains a pathological process including antibody-mediated rejection,
diagnosis. Indeed, in a canine study it has been and the pathophysiology is not entirely under-
found that resistive index actually decreases in stood. Doppler indices rarely show any significant
the mild to moderate stages of acute rejection. alteration in flow profiles with chronic rejection.15
During the early to mid-stages of rejection, the Perioperative ischaemia can result in transient
218 physical effects of increased intrarenal pressure compromise of renal function, particularly at the
9
a
b
c
Fig. 9.4 (a) Longitudinal grey-scale image of a
transplanted kidney. Note the rounded globular
configuration of the kidney. The central hilar space
(arrows) is compressed due to oedema and swelling;
therefore central hilar fat has been displaced.
(b) Spectral Doppler tracing at the main renal artery
level shows a resistive index of 90%. Biopsy confirmed
severe acute rejection. (c) Power Doppler image of
a severely rejecting renal transplant shows the main
central vasculature and a few interlobar vessels but
no flow can be seen out to the periphery of this
rejecting kidney. The high resistance generated by the
rejection process results in this colour Doppler ‘pruned
tree’ appearance. Contrast this to the appearance of
normal renal power Doppler flow in Figure 9.2.
level of the distal tubules which are most sensi- Perinephric fluid is a common sequela of renal
tive to hypoxia.This condition is self-limiting and transplantation and is not considered significant
typically resolves within 1–2 weeks of transplan- if it is crescentic in shape, or decreases in size
tation.The kidney appears oedematous, especially over time. Most fluid collections are haematomas
the medullary pyramids, and Doppler studies or seromas, which result from oozing from the
will show an increase in the resistive index. transplant bed; urinomas are relatively uncom-
Although the imaging and Doppler findings may mon and usually are the result of breakdown at
suggest acute tubular necrosis, they are not the ureteral anastomosis to the bladder. Doppler
specific11,17 (Fig. 9.5). examination is of limited value in these cases.
High pressure collections, such as haematoma
Anatomical complications after biopsy or organ rupture may exert mass
These include haematomas, seromas, urinomas, effect upon the kidney and locally affect haemo-
abscesses, lymphoceles, obstructive hydro- dynamics. In this case, Doppler may reveal a
nephrosis, focal masses, arterial and venous high resistance spectral tracing in proximity to
stenosis or thrombosis, and intrarenal arterio- the fluid pocket (Fig. 9.6).
venous fistula and pseudoaneurysm.8, 9, 18 Unlike A patient with a rounded, expansile collection
functional complications, most anatomical com- with internal debris and associated signs of
plications are readily identified by ultrasound. infection usually has an abscess. It is usually 219
9
a
b
P P
P
Fig. 9.5 (a) Longitudinal grey-scale image obtained with 24 h of implantation. This cadaver organ experienced
prolonged ischaemic time. Note that the kidney is not particularly rounded and the sinus fat is preserved.
The medullary pyramids (P), however, are prominent, hypoechoic and oedematous. (b) Spectral and colour Doppler
image of the same kidney. The resistive index is elevated to over 100% since reversed flow can be perceived in
diastole (arrows). This combination of findings in the appropriate clinical situation is consistent with acute tubular
necrosis. This can be seen in the immediate post-operative state but can also be caused by drug toxicity.
b
a
Fig. 9.6 (a) Longitudinal grey-scale image of a renal

c
transplant within 24 h of an upper pole biopsy.
The biopsy was complicated by haemorrhage (H).
Blood accumulated in the subcapsular space and
severely compressed the upper pole of this kidney
(arrows). (b) Spectral Doppler tracing obtained of an
interlobar artery just adjacent to the high pressure
haematoma reveals extremely high resistive index
with reverse flow in diastole. (c) Spectral Doppler
tracing at the opposite (lower pole) of this same
kidney shows a normal low resistance flow profile.
The compressive haematoma exerts local mass
effect and elevates resistance to flow.
220
9
difficult to diagnose an abscess by sonography around its pedicle. This may result in arterial
alone and computed tomography (CT) is inflow and venous outflow complication (Fig. 9.8).
considered a better imaging study for this Transient dilatation of the collecting system
purpose, especially to determine its extent and as a result of ureteral anastomotic oedema
percutaneous drainage potential. Occasionally frequently occurs immediately after renal trans-
colour Doppler may reveal hyperaemia of the plantation or removal of the ureteral stent. The
tissues surrounding the abscess. presence of a dilated transplant collecting system
Lymphoceles usually manifest at about does not automatically signify an obstructed
6–8 weeks postoperatively and are seen as system under pressure, as the denervated, flaccid
rounded or lobulated collections near the vascular collecting system can become markedly dilated,
anastomoses. They are the result of surgical dis- particularly when the urinary bladder is dis-
ruption of lymphatic channels when the vascular tended.17,19,20 Platt et al19 proposed that the
anastomosis to the transplanted kidney is created. identification of an elevated resistive index was
An expanding lymphocele may cause ureteric useful in distinguishing obstructive hydronephrosis
compression and hydronephrosis. If a lympho- from chronic, low-pressure dilatation of the trans-
cele becomes large enough, it may compress or plant collecting system. Although this observa-
kink the renal vascular pedicle. In this situation, tion may be sensitive, its specificity is very poor
Doppler examination may show findings similar because of the many other factors that similarly
to arterial or venous stenosis (Fig. 9.7). affect renal haemodynamics. At this time, the
Intraperitoneal transplantation, as is common Whitaker test remains the ‘gold standard’ for
with combined pancreas transplantation results differentiating high-pressure obstructive hydro-
in a more mobile kidney. Occasionally ptosis nephrosis from low-pressure distention of a flaccid
or rotation and torsion may occur in the kidney renal transplant collecting system.
L
L
Fig. 9.7 (a) Longitudinal colour Doppler image of this renal transplant shows a large fluid collection medial to
the kidney, surrounding the renal vascular pedicle. This is a lymphocele (L) and it has caused distortion of the
pedicle. Colour aliasing can be seen in the renal artery (arrow). There is also obvious hydronephrosis (H) caused
by compression of the ureter. (b) Spectral Doppler tracing of the renal vein shows marked compression where
it courses past the lymphocele (#1), the measured velocity at this area is 1.1 m s-1; whereas within the kidney
(proximal to the lymphocele (#2)) the velocity is only 0.1m s-1. This 10-fold velocity gradient indicates that this is
truly significant renal venous outflow obstruction.
221
9
a b
Fig. 9.8 (a) Longitudinal colour Doppler image along

the long axis of an intraperitoneal renal transplant
c with a spectral Doppler tracing of an arcuate artery.
The arterial waveform shows a tardus parvus
configuration indicating arterial inflow compromise.
(b) The main renal artery at the hilum reveals a high
velocity jet approximating 3 m s-1. (c) Immediately
adjacent to the main renal artery, the main renal vein
spectral Doppler tracing also shows turbulent high
velocity flow. This finding of high velocity, turbulent
Doppler tracings in both artery and vein is due to
torsion of the renal vascular pedicle and resultant
kink of the main artery and vein. Fortuitously this was
only partial torsion of the kidney and was corrected
with nephropexy. Intraperitoneal torsion of a transplant
kidney has been reported to result in infarction.
Vascular complications vasorum. Stenosis should be suspected if a

Following renal transplantation, vascular com- tardus parvus waveform and relatively low-
plications are observed in less than 10% of resistance flow are noted in the intrarenal
recipients; however, when present, they are asso- branches. A tardus parvus waveform is charac-
ciated with a high morbidity and mortality. terised by a delayed upstroke in systole (pro-
Complications include renal artery or vein stenosis, longed acceleration time >0.07 s), rounding
compression, kinking, thrombosis, intrarenal of the systolic peak and obliteration of the
arteriovenous fistulae and pseudoaneurysms. If early systolic notch. A flow velocity greater than
identified promptly, they can often be success- 2 m s-1 with associated distal turbulence near the
fully repaired prior to transplant failure. Doppler renal artery anastomosis is diagnostic of renal
sonography is a very effective, non-invasive screen- artery stenosis (Fig. 9.9). The examiner should
ing modality for identifying significant vascular conduct a thorough examination from the renal
complications.21–23 hilum to the iliac artery in search of a focal
stenosis, if an intrarenal tardus parvus waveform
Renal transplant artery stenosis is observed24–28 (Fig. 9.10). Occasionally with
This is most often observed within 1–2 cm of severe renal artery stenosis the intraarterial wave-
the anastomosis, usually as a result of vessel wall forms become flattened to the point that it is
222 ischaemia due to surgical disruption of the vasa very difficult to perceive the systolic diastolic
9
a b
Fig. 9.9 (a) Renal transplant arcuate artery spectral

c Doppler tracing in a hypertensive recipient with
elevated serum creatinine. The renal arterial waveform
manifests tardus parvus waveform with a slow systolic
upstroke, and a rounded systolic peak, relatively low
resistance. These findings suggest a more proximal
stenosis. (b) Spectral Doppler tracing just beyond the
anastomosis of the main renal artery to the iliac artery.
Note the high velocity (1.79 m s-1), turbulent flow
characteristic of renal artery stenosis. (c) MR angiogram
of this transplanted kidney actually revealed a double
arterial anastomosis. The lower pole artery shows
a high grade stenosis (arrow) within 1 cm of its
anastomosis.
variation. Subtle pulsatile flow is enough to fused by the affected artery. If thrombosis of this
document the patency of the artery and avoid artery occurs, then the subtended area shows no
misdiagnosis of thrombosis. This should be flow on colour or power Doppler and an arterial
further reinforced by the presence of constant tracing will not be identified by spectral Doppler.
outflow in the renal vein (Fig. 9.11). The area affected will vary depending on the
Approximately 20% of transplanted kidneys anatomical vascular distribution (Fig. 9.12).
require more than one arterial anastomosis due Renal parenchymal scarring secondary to
to the presence of accessory arteries. If one of chronic rejection may result in focal stenosis
these vessels becomes compromised, then perfu- within branch arteries. This should be suspected
sion to the subtended segment of the kidney is if there is irregular distribution of flow on colour
decreased. Again, a tardus parvus waveform may Doppler through the kidney. Segmental or inter-
be seen, this time limited to the segments per- lobar renal artery stenosis can be confirmed by 223
9
a
Fig. 9.10 (a) Spectral Doppler tracing of a transplant main renal artery. Tardus parvus changes were present
in this waveform but no anastomotic stenosis was perceived. (b) Spectral Doppler tracing of the external iliac
artery. A high-velocity jet is identified with velocities of >3 m s-1. This diabetic recipient had an atheromatous
lesion in the iliac artery compromising inflow to the renal transplant.
a
b
Fig. 9.11 Intraoperative spectral and colour Doppler ultrasound of a recent renal transplant recipient being
reoperated because of non-function and a Doppler suggestion of arterial stenosis. This spectral Doppler tracing
of a segmental artery and its adjacent vein shows only a very subtle undulation in the arterial waveform (towards
the transducer). The severe proximal stenosis completely wiped out the expected systolic velocity variation.
The tracing below the baseline is not mirror imaging of the tracing but venous outflow. (b) As the anastomosis
was surgically manipulated a sudden increase in arterial velocity with a more conspicuous arterial tracing
became evident on spectral Doppler.
the presence of intrarenal high-velocity flow. a greater chance of vessel injury. Direct puncture
Because these lesions are typically multiple and of a major artery usually becomes immediately
distal, treatment options are limited.29 manifest as an area of turbulence on colour
A similar appearance can be seen with scarring Doppler, a rapidly expanding haematoma, or
after transplant biopsy. If the biopsy needle is brisk haematuria due to an arteriourteral fistula.
224 guided centrally toward the renal hilum, there is Biopsy in close proximity to a major artery may
9
whereas venous to-and-fro flow has a rounded
appearance.
Renal transplant vein stenosis

Renal vein stenosis is an uncommon complica-
tion following kidney transplantation, but when
present, it is a significant cause of graft dysfunc-
Fig. 9.12 Side-by-side longitudinal grey-scale and tion. Venous stenosis may be seen as a focal
power Doppler image of the low pole of a poorly narrowing with associated dilatation of the
functioning transplant kidney. Note the thin cortex proximal vein. However, to confirm the diag-
(arrow) and complete absence of flow on power nosis of a significant stenosis, there should be at
Doppler at the lower pole. This renal transplant least a fourfold velocity gradient across the
required two arterial anastomoses. Thrombosis of lesion. If the gradient is less than fourfold, it is
the lower pole artery was suggested and confirmed seldom considered clinically significant, even
angiographically. Additionally, ultrasound-guided though it may have a dramatic appearance on
biopsy of the upper pole identified acute rejection. Doppler examination (Figs 9.14 and 9.7b).
Renal transplant vein thrombosis

present with delayed segmental perfusion as This is also a relatively rare post-transplantation
scarring results in progressive compromise of complication. It is usually seen within the first
flow to that portion of kidney. week following surgery. It is more likely to occur
when there is technical difficulty with the venous
Renal transplant artery thrombosis anastomosis. It may occur with preservation
Thrombosis of the main renal transplant artery injury, or it may evolve during an episode of
is a rare event. It is typically due to a technical severe acute allograft rejection. Extremely high
problem with the surgical anastomosis. Doppler resistance (resistive index typically greater than
fails to demonstrate any arterial flow. 100%) will be seen on the renal arterial wave-
Rarely severe acute rejection may cause micro- forms with spectral Doppler. In most cases,
vascular thrombosis. Colour and Power Doppler thrombosis results in partial rather than complete
show no flow within the kidney. Flow may still occlusion of the vein and some renal venous
be present in the segmental or interlobar outflow can be detected with spectral Doppler.
branches, but spectral Doppler shows very high To prevent a false-negative diagnosis, it is
resistance because the flow has nowhere to go important that the examiner conduct a careful
(Fig. 9.13). It may be difficult to differentiate imaging and colour Doppler evaluation of the
the two aetiologies, but the non-rejecting kidney vein when very high arterial resistance is
with arterial thrombosis is typically not as swollen observed31 (Fig. 9.15).
and oedematous as the acutely rejecting throm-
bosed kidney. Intrarenal arteriovenous fistulae and
The spectral Doppler tracing of renal vein pseudoaneurysms
flow in microvascular thrombosis may be very These are typically the result of renal transplant
confusing. The blood within the vein may be biopsy. The true incidence of these complica-
seen to slosh back and forth with cardiac tions varies from centre to centre depending on
periodicity transmitted down the inferior vena biopsy technique. Arteriovenous fistulae mani-
cava (IVC). This to-and-fro flow may mimic fest as a flash of colour, or ‘visible thrill’, in the
high resistance arterial flow, except for the fact adjacent parenchyma when the kidney is
that the upstroke is not brisk. In high resistance examined at normal colour Doppler settings.
arterial flow, systole upstroke is brisk with a peak This phenomenon is caused by vibration of the 225
9
b
C
C
A A A
Fig. 9.13 Longitudinal power, colour and spectral Doppler imaging of this renal transplant recipient with rapidly
rising creatinine levels. (a) Power Doppler image shows central flow within the hilum of the kidney, but no flow
within the cortex (C) or medulla. Microthrombosis was suspect. (b) Spectral Doppler tracing of the artery shows
a brisk spike in systole and essentially no flow during diastole. (c) The renal vein spectral Doppler waveform has
an unusual to-and-fro pattern. In summating the area under the tracing above and below the baseline, it becomes
evident that there is little total antegrade flow. Due to the microvascular thrombosis, the flow within the vein is
simply stagnant moving to-and-fro responding to intrarenal pressure changes between systole and diastole and
pressure changes in the inferior vena cava. Contrast this to a tracing of a normal renal vein (d). The retrograde
component of flow known as the A-wave (A) occurs during atrial systole. But note how relatively little reversed
component of flow there is in comparison to the amount of flow below the baseline that is returning toward
the heart.
surrounding tissues secondary to the rapidly be possible to observe pulsatile flow within the
flowing blood through the fistula. It is often main renal vein (Fig. 9.16).
possible to distinguish the feeding artery and the Pseudoaneurysms are typically the result of
enlarged draining vein by adjusting the colour a biopsy that captured partial thickness of an
Doppler settings to higher velocity. Spectral arterial wall. Therefore, they are extremely rare.
Doppler tracings will demonstrate a high- They usually appear as a simple cystic structure,
velocity, low-resistance flow within the feeding or a small collection of paravascular fluid. Colour
artery. Turbulent, pulsatile (arterialised) flow Doppler, however, immediately reveals that the
will be present in the draining segmental vein. If finding is not a simple cyst (Fig. 9.17). Spectral
226 the arteriovenous fistula is large enough, it may Doppler tracings show to-and-fro blood flow at
9
Liver transplantation
a
Fig. 9.14 Spectral Doppler tracing of the renal

vascular pedicle incorporating both the arterial (A) b
and venous (V) waveform. Note the high velocity in
the vein approximating 1 m s-1. Note the corresponding
high resistance arterial waveform. A renal vein
stenosis due to scarring was identified at surgery.
Fig. 9.16 (a) Power Doppler image of a renal

transplant approximately 2 weeks after biopsy.
A burst of colour is seen overlying the lower pole
(arrow). This flash artefact, due to tissue vibration,
suggests an underlying arteriovenous fistula.
(b) Spectral Doppler tracing obtained at the fistula
site reveals a turbulent, high-velocity arterial waveform;
that differentiates it from pseudoaneurysm that has
to-and-fro flow.
Fig. 9.15 Spectral Doppler tracing of the main renal
transplant artery in a recent recipient with a rapidly
increase in size over a period of time, angiographic
rising creatinine. The arterial waveform reveals to-and-
fro flow with the retrograde component being
embolisation may be necessary.32,33
essentially equal to antegrade flow. The resistive
index measures approximately 140%. No flow could LIVER TRANSPLANTATION
be identified by spectral or colour Doppler in the renal
veins. Complete venous thrombosis was confirmed More than 22 000 liver transplants have been
at angiography. performed in the world since 1988. The current
1-year graft survival rate in the USA is approxi-
mately 82.4% with 1-year patient survival rate of
the neck of the pseudoaneurysm and a distorted, approximately 86%.34 Improved surgical tech-
turbulent, pulsatile waveform can be observed niques, development of effective immunosup-
within the pseudoaneurysm. The vast majority pressive medications, HLA typing for recipient
of intrarenal arteriovenous fistulae and pseudo- matching, and establishment of coordinated
aneurysms resolve spontaneously, but if they transplant sharing systems have greatly improved 227
9
identification of vascular collaterals secondary to
portal hypertension and a search for intra- or
extrahepatic malignancy (Fig. 9.18). There are
many ways to accomplish this including angio-
graphy and ultrasound, but currently CT with
CT angiography in the arterial phase and portal
phase is the favoured method for the adult.
P
Some centres rely on Doppler and magnetic
resonance imaging (MRI), especially for the
paediatric candidate.35 A complete sonographic
examination of the liver transplant candidate
should cover the points listed in Table 9.3.
Fig. 9.17 Power Doppler image of a renal transplant
3 weeks after a biopsy. A rounded area of bright flow
Postoperative assessment
represents a pseudoaneurysm (P). There is a higher
The major complications of liver transplantation
likelihood of this complication with a more centrally
are rejection, vascular stenosis or thrombosis,
directed biopsy, as was the case with this patient.
biliary leak or obstruction, and progression of
unexpected malignant disease. Acute rejection
the success rate in liver transplantation. Graft is clinically monitored by serum liver enzymes,
survival statistics are further enhanced by prompt bilirubin, and ammonia levels and diagnosis is
identification of liver transplant dysfunction and established by biopsy. Ultrasound and Doppler
rapid intervention when appropriate. have little to offer in the diagnosis of acute hepatic
rejection. Doppler ultrasound, however, plays
Preoperative assessment a key role in monitoring for potential vascular
Preoperative assessment consists of confirmation complications. A complete sonographic exami-
of vascular patency, mapping native vascular nation of the liver transplant recipient should
anatomy, quantification of diseased liver volume, cover the points listed in Table 9.4. Evaluation of
Fig. 9.18 Transverse ultrasound image of the left lobe of the liver reveals a focal nodule bulging the capsular
contour (arrows) (a). Ultrasound imaging alone cannot further characterise this lesion. Arterial phase rapidly
enhanced CT (b) shows this lesion to be a hypervascular mass (arrow), most likely a hepatocellular carcinoma.
With the addition of ultrasound contrast, some centres are able to use Doppler to identify malignancy during
228 the transplant work-up.
9
Table 9.3 Checklist for the pre-liver Table 9.4 Checklist for the post liver
transplantation ultrasound examination transplantation ultrasound examination
• Confirm patency of the portal vein. Provide length • Survey the liver parenchyma to rule out any focal
and diameter measurements of the extrahepatic abnormality, specifically fluid collections, areas of
portal vein infarction, and possible neoplasm
• Identify any anatomical variation of the hepatic • Survey and perihepatic recesses, lateral gutters,
artery and pelvis to identify and quantify any
haematoma or fluid collection
• Confirm patency of the inferior vena cava
• Evaluate the biliary system, both intrahepatic and
• Identify and describe collateralisation from the
extrahepatic, to rule out obstruction, intraductal
portal to the systemic circulation
sludge and stone formation
• Identify any hepatic mass that may represent
• Confirm the presence of both intrahepatic and
hepatocellular carcinoma
extrahepatic artery flow, and analyse the
• Quantify the amount of ascites waveforms
• Estimate the size of the diseased liver • Confirm portal vein patency and analyse the
Doppler waveform, particularly across the
• Provide a measurement of spleen size
anastomosis
• Confirm patency of the three hepatic veins and
evaluate their waveforms
the newly transplanted liver requires a precise • Evaluate flow in the inferior vena cava with
attention to the anastomoses
understanding of the surgical anatomy. Many
variations are possible including segmental or
reduced-size transplantation, especially in the
paediatric population.36,37 Variations of the arterial Abnormal findings
anastomoses are necessary when the donor hepatic The most common abnormal findings encoun-
arterial anatomy is anomalous.Variations of venous tered in liver transplantation are listed in Table 9.5.
anastomoses are necessary when the recipient The hepatic artery anastomosis is technically
portal vein is thrombosed. The sonologist must the most difficult to create and problems, such
be aware of any variations so that a thrombosed as stenosis, thrombosis and fistula formation,
accessory hepatic artery or a stenotic jump graft have the most significant impact on liver trans-
is not missed (Fig. 9.19). plant success as they predispose to infarction,
The liver transplant ultrasound examination intrahepatic abscess, biliary stricture and biloma
should include a general survey of the abdomen formation. Doppler findings indicating hepatic
and pelvis in order to identify and quantify any artery stenosis include an intrahepatic tardus
haematomas or fluid collections. The liver parvus waveform with low-resistance flow and
parenchyma is then examined to rule out any a high-velocity jet with turbulence at the point
focal abnormality, specifically any fluid collection, of stenosis. A focal high-velocity jet just beyond
area of infarction or possible neoplasm. The the hepatic artery anastomosis in excess of
biliary system should be evaluated to rule out 200 cm s-1 or greater than three times the velocity
obstruction or sludge accumulation, especially in the prestenotic hepatic artery is highly indica-
in a patient with hepatic artery thrombosis. The tive of a clinically significant stenosis.38 The
intra- and extrahepatic hepatic arteries are identification of an intrahepatic tardus parvus
checked to confirm patency and the waveforms waveform with low resistance (<60% resistive
are analysed to rule out stenosis. Patency of the index) flow, a prolonged upstroke in systole
portal vein is confirmed and the Doppler wave- (>0.08 s)38 and rounding of the systolic peak,
form analysed, particularly across the anasto- should force a careful survey in the anticipated
mosis (Fig. 9.20). Patency of the three hepatic region of the anastomosis for the high velocity
veins is confirmed and their waveforms are eval- jet (Fig. 9.22).38,39 It has been shown that search-
uated. Finally, the IVC is checked with special ing for the tardus parvus waveform pattern is an
attention to the upper anastomosis (Fig. 9.21). excellent screening test for the presence of post- 229
9
a IVC b
HA
IVC
PV
CBD Fig. 9.19 Three artist renderings of liver transplant
variants. (a) An in-line liver transplant has a total of
c
five anastomoses. In the porta hepatis, there are
HV
anastomoses of the hepatic artery (HA), portal vein
(PV), and bile duct (CBD). The recipient inferior vena
cava is resected with the explant liver. The donor
inferior vena cava is implanted in-line with an upper
and lower IVC anastomosis (IVC). (b) The piggyback
Donor technique requires the same three anastomoses in
lliac Vein the porta hepatis. The recipient inferior vena cava
Graft remains in place. The explant native liver is stripped
off of the inferior vena cava. The donor liver is then
AL placed over and hooked up to the recipient inferior
vena cava via a cloaca (C) formed from the recipient
hepatic vein ostia. The lower end of the donor cava is
JGs ligated (arrow) effectively turning it in to an accessory
hepatic vein. (c) The reduced or donor liver transplant
has the most complicated anatomy. The hepatic artery
Superior and portal vein frequently require a jump graft (JGs)
Mesenteric Vein (an interposed length of donor vasculature). The biliary
Recipient system is often drained via an afferent loop (AL) of
Hepatic Artery
bowel. The hepatic vein draining the transplanted
segment is anastomosed to the inferior vena cava
usually via a donor hepatic vein (HV).
liver transplantation hepatic artery stenosis.40 to biliary ischaemia, or may progress to com-
One article suggests that in the paediatric donor plete thrombosis.
population, the finding of hepatic arterial resis- Absence of an arterial signal along the main
tive index <60% is highly predictive of impend- portal vein and its branches on spectral and
ing hepatic artery thrombosis due to stenosis colour Doppler ultrasound indicates hepatic
and thrombectomy and reanastomosis should be artery thrombosis. Since this is a diagnosis based
considered.41 Although an intrahepatic arterial on the absence of flow, great care must be taken
tracing may be demonstrated, it should be to ensure proper Doppler settings. Scanning by
230 remembered that a severe stenosis may still lead a second experienced sonologist is encouraged,
9
a
b
Fig. 9.20 Colour Doppler image of the porta hepatis (a). The anastomosis between donor and recipient portal
vein is evident by the change in calibre which causes the focal aliasing at that level (arrows). Spectral Doppler
tracing (b) across the anastomosis and on either side reveals the velocity changes are relatively insignificant in
this normal anastomosis.
a
b
Fig. 9.21 Transverse colour Doppler image of the hepatic vein/IVC anastomosis (a). Three hepatic veins are
indeed patent. Aliasing however is perceived at the caval anastomosis (arrows). Spectral Doppler tracing (b)
reveals normal cardiac periodicity transmitting into the liver. Since this waveform must propagate against the
direction of flow, its presence within the liver effectively rules out significant hepatic venous outflow obstruction.
since this ultrasound diagnosis routinely leads to Studies performed on patients with hepatic
arteriography. Use of ultrasound echo-enhancing artery stenosis and thrombosis have identified
agents is recommended to improve perception certain factors that place patients at a higher risk
of a weak arterial signal and decrease the rate of and warrant more frequent Doppler screening.
false-positive diagnosis of hepatic artery throm- These factors include bench reconstruction of
bosis and reduce the frequency of hepatic artery anatomical variance, the use of an interposition
arteriography.42,43 graft, and patients undergoing retransplantation.44 231
9
a
Table 9.5 Differential diagnosis of ultrasound
findings in the transplanted liver
Diffusely irregular parenchymal echotexture

• Ischaemia or necrosis secondary to hepatic artery
stenosis or thrombosis
• Infection or abscess
• Recurrent hepatitis
• Post-transplantation lymphoproliferative disorder
or lymphoma
• Geographic or diffuse fatty infiltration
• Hepatocellular carcinoma
• Diffusely metastatic neoplasm
Focal parenchymal abnormality
• Abscess
• Infarction b
• Recurrent neoplasm
• Intraductal air secondary to
choledochojejunostomy or sphincterotomy
• Intraductal sludge or calculi
High-resistance flow in hepatic artery
• Preservation injury – common during the first few
days after transplantation
• Organ compression by adjacent mass or fluid
collection
• Hepatic venous outflow obstruction
• Severe hepatocellular disease
Low-resistance flow in hepatic artery Fig. 9.22 (a) Colour and spectral Doppler tracing of
• Hepatic artery stenosis an intrahepatic arterial flow profile. Tardus parvus
• Severe aortoceliac atherosclerotic disease pattern with a delayed upstroke in systole rounded
• Diaphragmatic crus sling effect systolic curve and relatively low resistance flow is
• Intrahepatic arteriovenous fistula indicative of insufficient hepatic arterial inflow and
• Arteriobiliary fistula suggests stenosis. The examiner should carefully
Flattening of hepatic vein waveform walk the sample volume down the hepatic artery and
• Hepatic parenchymal disease, including rejection out of the liver looking for the anticipated point of
stenosis. In this case, (b) a high velocity jet measuring
• Stenosis or kink of the upper inferior vena cava
anastomosis over 5 m s-1 identified the point of stenosis in the
subhepatic space.
In cases of hepatic artery thrombosis which high resistance flow. This is a common manifes-
are treated conservatively, collaterals will develop tation of ischaemic reperfusion injury (the anoxia
and an intrahepatic arterial signal can be detected and traumatic insult sustained by the liver during
by Doppler ultrasound as early as 2 weeks after recovery, handling, preservation and surgery). It
the thrombosis. This typically manifests as a has been shown to be more common with older
thready tracing with a tardus parvus appearance donor age and a prolonged period of ischaemia46
and can be seen in as many as 40% of patients (Fig. 9.23). The high resistance is due to vaso-
with documented hepatic artery thrombosis, spasm and can be reversed with vasodilatory
especially children.45 agents such as nifedipine. One must be cautious
Within the first few days of transplantation, however to ensure that the transplant recipient
232 the hepatic artery tracing often shows a relatively is stable enough to be given this drug as a diag-
9
b
Fig. 9.23 Transverse ultrasound image of a recently transplanted liver (a). The fat surrounding branches of the
portal triad (arrows) stands out prominently against the background of an oedematous, hypoechoic liver; the
sequela of ischaemic injury related to the surgical procedure. This is very common within the first few days of
transplant. (b) Spectral Doppler tracing of the hepatic artery shows a very high resistance waveform with sharp
spikes in systole and very little flow through diastole. Rarely does this progress to thrombosis, but more often
evolves into a normal resistance waveform as the oedema and vasospasm resolves.
nostic challenge. Augmenting the Doppler exam Arteriovenous fistulas are a rare complication
with this challenge, however, may obviate the of transplantation and are most often the result
need for an arteriogram if hepatic arterial inflow of a biopsy. Imaging rarely reveals an abnormality,
is compromised to the point where it appears but colour Doppler often shows a localised flash
to be occluded (Fig. 9.24). The spasm typically
resolves within a few days of transplantation47
and resistance returns to a normal range.48 A
delayed finding of high resistance, beyond 3–5
days, is a poor prognostic indicator and some
of these patients develop arterial thrombosis.49
The exact cause of thrombosis is not always
apparent and in numerous cases is presumed to be
secondary to immunological causes and rejection.
At the time of implantation, there must be
sufficient length of all of the vasculature to create
the anastomoses. A longer pedicle is easier to
work with, however, if the vessels are too long
a kink may occur as the liver is placed into the
fossa and the abdominal wall is closed. Clinically
Fig. 9.24 Spectral Doppler tracing of the porta
the patient presents with liver dysfunction. A
hepatis on the first postoperative day. The hepatic
stenosis may be suspect on spectral Doppler. artery was extremely difficult to identify (arrows)
Colour or power Doppler may reveal the because of the thready, high resistance flow due to
tortuosity. Three-dimensional CT angiography severe vasospasm. This patient may be considered
may be performed to provide ‘the big picture’ a candidate for ultrasound contrast or a provocative
and determine if a stent of reoperation is the test with vasodilating agents to confidently confirm
best treatment (Fig. 9.25). arterial patency. 233
9
a b
Fig. 9.25 Spectral Doppler tracing of the hepatic

artery in the porta hepatis reveals a high velocity jet
suggesting a stenosis (a). Power Doppler image of
the subhepatic space (b) reveals a markedly tortuous
hepatic artery. The three-dimensional rendering of the
CT angiogram during the arterial phase (c) reveals a
kink (arrow) due to excessive vessel length in the
subhepatic space.
artefact. When Doppler settings are adjusted for relatively rare. When it occurs, it may be due
high velocities, the feeding artery and draining to a mismatch between the diameter of the
vein are better visualised. Spectral Doppler reveals recipient and donor portal veins, to an excessive
a low resistance arterial waveform with high length of vessel causing a kink, or to a stenosis.52
diastolic velocity.50,51 If portal vein stenosis is suspected, the velocity
The donor portal vein is usually anastomosed gradient across the anastomosis should be
end-to-end with the recipient portal vein. measured by spectral Doppler; a velocity
Variations may be required if the recipient portal gradient of less than fourfold is unlikely to be
vein is thrombosed, hypoplastic, or of insuffi- significant (Fig. 9.26).53 Nevertheless, recipient
cient length. Because the vessel is relatively large, donor size mismatch can lead to portal vein
colour Doppler findings can be rather striking thrombosis. This can be treated by surgical
(Fig. 9.26). Not all flow disturbances perceived thrombectomy, angioplasty or thrombolytic
by colour Doppler are haemodynamically signifi- infusion, but in some cases retransplantation
234 cant and compromise of portal vein flow is may be required.54
9
b
a
Fig. 9.26 (a) Colour Doppler image of the portal vein in the region of the anastomosis. Note the area of relative
narrowing at the point of anastomosis (arrows). A high velocity jet with aliasing is seen shooting into the donor
portal vein. Spectral Doppler tracing on either side of this anastomosis (b) had to be set low to project the
relatively slow preanastomotic portal vein flow. As a result, the postanastomotic (poststenotic) tracing shows
a high degree of aliasing. This comparative velocity measurement on one tracing can be achieved by moving
the sample volume across the anastomosis and allows an accurate comparison of pre- and postanastomotic
velocities. Velocity gradients of less than fourfold are not likely to be clinically significant.
Post-transplantation portal vein thrombosis is intrahepatic IVC is excised with the diseased
quite rare and most often attributed to technical liver. The more common surgical technique
factors. It is more likely to occur in the paediatric retains the native IVC of the recipient in place
recipient, especially after split liver transplan- and the upper end of the donor IVC is anasto-
tation.55,56 Prompt detection with frequent mosed end-to-side to the native IVC at the
Doppler evaluation and aggressive surgical treat- confluence of the hepatic veins of the explanted
ment in selected cases are required to reduce the liver. The lower end of the donor IVC is over-
mortality and graft loss. sewn, which functionally converts it into a hepatic
If slow velocity is identified in the portal vein vein. Relative flow volumes through this vessel
(<1 m s–1), it may be due to increased intra- are much less than when it served as the IVC,
hepatic resistance from rejection, or to reduced therefore clot may be seen partially filling the
inflow as can be seen with the collateral steal lumen (Fig. 9.27).This should not cause concern
phenomenon. This occurs when large varices as long as some flow can be perceived. This type
remain unligated, shunting blood from the portal of anastomosis is commonly referred to as a
system to the systemic circulation, bypassing the ‘piggyback’.61 The incidence of hepatic venous
liver.57–59 complications in partial liver transplantation is
A pulsatile waveform in the portal vein may more frequent than that in whole liver trans-
be observed within the first few weeks after plantation.62 Any compromise of the upper caval
transplantation, especially in patients that received anastomosis, from either stenosis or kinking,
small grafts. This pulsatile flow often disappears may cause hepatic venous outflow obstruction.
without any treatment, although it may some- Ultrasound findings include marked damping,
times represent vascular complications such as or complete flattening of the hepatic vein velocity
arterioportal shunt.60 profile with complete loss of periodicity, disten-
The donor IVC has a long intrahepatic course sion of the hepatic veins and a high velocity jet
and is therefore transplanted along with the with turbulence just above the caval anastomosis
liver. The IVC may be inserted in-line with both (Fig. 9.28). Hepatic vein or caval anastomotic
supra- and infrahepatic anastomoses; the native stricture may be treated with balloon dilatation 235
9
or endovascular stent placement. If these fail, then
surgical intervention with a patch venoplasty of
the anastomosis can be performed.62 Ideally,
after a successful procedure hepatic vein calibre
should decrease and cardiac periodicity should
return to the hepatic vein flow profile.63,64 Loss
of periodicity may also be due to compression of
the hepatic veins by the surrounding liver tissue
by oedema in the early postoperative period,
typically due to preservation injury, or by oedema
in the later postoperative period related to rejec-
tion.45,65 Due to the relatively large size of the
IVC and the potential for a size mismatch
Fig. 9.27 Longitudinal colour Doppler image of donor
between the donor and recipient cava, a greater
inferior vena cava. This liver is anastomosed with a than fourfold velocity gradient at the anastomosis
piggyback technique therefore the flow within the cava is required to confidently diagnose a haemo-
is only a fraction of what it was previously. As a result dynamically significant stenosis. A persistent
clot has formed partially occluding the lumen (arrow). monophasic hepatic vein flow profile is highly
a b
Fig. 9.28 (a) Colour Doppler image of the hepatic

c
vein confluence with the IVC. Note the relatively
distended hepatic veins and the colour aliasing at
the anastomosis. (b) Spectral Doppler tracing of the
middle hepatic vein shows complete absence of
RHV
periodicity. Since this tracing is acquired within a few
centimetres of the heart, one should expect to see
normal hepatic vein flow periodicity. Its absence
IVC indicates hepatic venous outflow obstruction either
due to organ shift and kinking or stenosis due to
fibrosis. (c) A longitudinal colour Doppler image at the
junction of the right hepatic vein (RHV) and inferior
vena cava (IVC) shows a distended hepatic vein with
flow backed up behind a focal stenosis (arrow).
236
9
The pancreas transplant

suggestive for hepatic outflow obstruction, but it
THE PANCREAS TRANSPLANT
is not specific. On the other hand, the presence
of periodicity within the hepatic vein tracing on Whole organ pancreaticoduodenal transplanta-
Doppler ultrasound can confidently exclude the tion is the most reliable means of restoring
possibility of significant outflow obstruction.66 normal glucose homeostasis, the ultimate goal in
In those patients with an in-line IVC, com- the treatment of type I diabetes mellitus. Some
promise of the lower anastomosis may present studies have suggested that the restoration of
as lower extremity oedema and renal failure. normoglycemia for extended periods of time
Ultrasound and colour Doppler imaging of the have a beneficial effect on patient quality of life,
anastomosis may reveal a kink or focal stenosis recurrence of diabetic glomerulopathy, micro-
with a relatively high velocity jet. As with a piggy- angiopathy, and progression of peripheral and
back procedure, a size mismatch between the autonomic neuropathy. There has been steady
donor and recipient vessels may produce a rela- improvement in results of pancreas transplan-
tively high velocity jet and a less than threefold tation, gains largely due to the evolution of
velocity increase at the anastomosis is seldom surgical techniques and more potent, selective
clinically significant. A three- to fourfold gradient immunosuppressive medications. In the USA,
is likely to be significant and should be correlated the current 1-year pancreas graft survival rate
with the clinical findings. A greater than fourfold (when combined with a renal transplant) is 86%.70
gradient must be considered haemodynamically Despite the remarkable improvements made in
significant. the graft survival rates for all types of pancreatic
Several authors have studied the possibility transplantation, the high prevalence of graft loss
of predicting acute liver transplant rejection by caused by immunological rejection and surgical
identifying changes in the hepatic vein wave- complications continues to be a problem. The
form.67 During rejection, hepatocellular oedema technical failure rate approximates 8%, with
and inflammatory infiltration increase the pres- vascular thrombosis being the main cause. Infec-
sure within the confining capsule of the liver. tion, pancreatitis, bleeding, anastomotic leaks and
This reduces the compliance of the liver and rejection are other causes of transplant failure.71
results in a dampened hepatic vein waveform. Many sonographic patterns of pancreas trans-
The theory is appropriate but other causes of plant dysfunction have been described, including
hepatocellular oedema, such as cholangitis, focal or diffuse inhomogeneity of the echo texture,
hepatitis and upper IVC anastomotic stenosis increased or decreased overall echogenicity and
produce similar damping, thereby limiting the graft swelling. Unfortunately, none of these
specificity of this finding. The diagnosis of rejec- imaging findings is pathognomonic for any one
tion is best made by needle biopsy. Ultrasound complication of the pancreas transplant. In
and Doppler guidance should be used to guide general, acute inflammatory changes tend to
the biopsy needle into the liver, but away from result in an oedematous swollen pancreas and
the large central vessels.68 a chronic insult tends to result in a small
Hepatocyte transplantation is now performed echogenic gland. These characteristics, however,
at some centres as a temporising bridge until a do not help in solving the immediate clinical
donor liver can be acquired for the candidate. dilemma, specifically, differentiation between
Patients undergo intraportal infusion of cryo- pancreatitis, rejection and vascular compromise.
preserved, matched human allogenic hepatocytes. A complete sonographic examination of the
Portal vein thrombosis with liver failure and pancreas transplant recipient should cover the
death has been reported as a complication of points listed in Table 9.6. When examining a
this treatment. Portal vein Doppler ultrasound pancreas transplant recipient, it is mandatory
during and after cell infusion is mandatory for that the sonologist know the anatomic details of
these patients.69 that patient’s transplant (Fig. 9.29). Urological 237
9
the preferred conduit for managing pancreatic
Table 9.6 Pancreas transplant – sonographic
examination checklist exocrine secretions.
The transplanted pancreas is one of the more
• Review the surgical record, especially with regard
to the vasculature and drainage technique.
difficult organs to monitor by any imaging
Review any available prior imaging studies modality. Ultrasound evaluation of the pancreas
• Evaluate organ echotexture. Make certain it is transplant can be very difficult. Because the
uniform throughout
pancreas does not have a discrete investing capsule,
• Rule out pancreatic duct dilatation
it is difficult to define its margins with sonography.
• Rule out peripancreatic fluid collections
Taken out of its normal anatomical location, it
• Rule out pseudocyst
becomes more difficult to perceive, especially when
• Examine the main artery to the transplant,
particularly near its anastomosis camouflaged among bowel loops. Furthermore,
• Verify venous patency, particularly with high in the absence of an adjacent liver, determining
resistance arterial inflow to the organ its relative echogenicity is difficult. With inflam-
• Verify uniform parenchymal perfusion by colour mation of the pancreas and the associated oedema
and power Doppler
of the adjacent tissues, the already hard to define
pancreas becomes even more indistinct.
The one intrinsic anatomical landmark helpful
in determining that it is indeed the pancreas
being imaged is the pancreatic duct, but this
is not always obvious. Identification of the
pancreas may be improved by use of colour flow
or power Doppler imaging. It identifies flow
within the gland and the adjacent transplanted
splenic artery and vein. This helps it stand out
against the background of bowel loops (Fig. 9.30).
Since the splenic artery and vein end blindly
P
(the donor spleen is removed), there is an increased
DS possibility of them becoming thrombosed. Conse-
quently, Doppler sonography is particularly useful
in searching for vascular complications such as
thrombosis, or anastomotic stricture.71,72 Non-
visualisation of flow within the organ is a disap-
pointing finding (Fig. 9.31). If identified promptly,
Fig. 9.29 Artist rendering of a combined renal
however, thrombolytic therapy alone, or in com-
pancreas transplant. Typically the pancreas (P) is
bination with surgical thrombectomy, has been
transplanted on the right with an interposed duodenal
reported to succeed in restoring transplant perfu-
stump (DS) between pancreas and ileum. The splenic
artery and vein, which serve as the vascular pedicle
sion. Early diagnosis of pancreas transplant vas-
for the pancreas transplant, are anastomosed to the cular complications is of paramount importance
common iliac artery and vein. for appropriate treatment and organ salvage.73
Although the digestive enzymes are typically
drained into the gut via the duodenal stump,
and metabolic complications occur frequently in some may leak around the pancreas vasculature
bladder-drained pancreas transplant recipients; predisposing to pseudoaneurysm formation, espe-
and approximately 25% of bladder-drained cially in the region of the anastomosis. Arterial
pancreas allografts ultimately undergo enteric flow within a perianastomotic fluid collection, the
conversion for these complications.Therefore, in presence of swirling blood on colour Doppler,
238 the last several years, enteric drainage has become and a to-and-fro spectral Doppler waveform
9
The pancreas transplant

a b
P
P
Fig. 9.30 Power Doppler image of the transplant

pancreas (P). (a) The vascular pedicle to the transplant
c pancreas is a prominent feature that helps to
distinguish it from the surrounding bowel loops.
(b) Setting scan parameters for slower velocities
can depict branch vasculature within the pancreas
transplant. This of course may be difficult when there
is significant peristalsis in adjacent small bowel loops.
(c) If a sufficient sonographic window is available,
then a three-dimensional rendering of the vascular
pedicle (arrow) can document its integrity.
indicate a transplant associated pseudoaneurysm74

(Fig. 9.32). Rupture of transplant associated
pseudoaneurysm has been described and is a life
threatening event.75 The pancreas transplant, as
any other transplanted organ, requires occasional
biopsy and this increases the risk of developing an
arteriovenous fistula (AVF). Doppler findings are
similar to AVFs seen in other locations (Fig. 9.33).
Although ultrasound is excellent in identify-
ing parapancreatic fluid collections, the finding
is frequently non-specific. Abscess, haematoma
or liquefied phlegmon may all have similar ultra-
Fig. 9.31 Power and spectral Doppler image of the sound appearance of a complicated, debris-
pancreas. No colour flow can be identified within the filled, irregular collection (Fig. 9.34). The true
gland. A spectral Doppler of the feeding artery shows extent of a fluid collection is better evaluated by
extremely high resistance with spike of systolic inflow CT in which bowel gas does not limit the field
and reversed outflow throughout diastole. of view. Furthermore, follow-up CT examinations 239
9
Fig. 9.32 Colour Doppler image near the anastomosis

of the pancreatic vascular pedicle. A rounded collection
of colour flow with the classic swirl pattern identifies
pseudoaneurysm of the anastomosed splenic artery. Fig. 9.33 Colour and spectral Doppler image of a
pancreas several days after a biopsy. High velocity,
turbulent, low resistance arterial flow is present in
can be more reliably compared than ultrasound an area of flash artifact. This was an arteriovenous
examinations. The source of a fluid collection is fistula; a complication of the biopsy.
best evaluated by a contrast study or by aspiration.
Preliminary reports suggested that elevation needs to identify the pancreatic vascular pedicle
of the resistive index could predict pancreatic which must be avoided.The biopsy is monitored
transplant rejection. Unfortunately this is not in real time, and delayed imaging and Doppler
the case, as many rejecting transplants have are useful to identify any possible complications.
normal resistive indices. Since the pancreatic
graft lacks a capsule, the rejection process will
POST-TRANSPLANTATION
not generate enough intraparenchymal pressure
LYMPHOPROLIFERATIVE
to cause a consistently measurable elevation of
DISORDER
vascular resistance.76 A key role of ultrasound in
rejection is to guide the pancreas transplant Post-transplantation lymphoproliferative disorder
biopsy. It must identify a safe path to the organ, (PTLD) is a rare but serious complication follow-
free of overlying intestine or mesentery. Doppler ing solid organ transplantation. The most
Fig. 9.34 Grey-scale and colour Doppler images of a pancreas transplant reveal an echogenic rounded area
with shadowing from gas. Colour Doppler shows some hyperaemia immediately adjacent to this collection.
240 This was a focal pancreatic abscess.
9
References
commonly accepted theory for the pathophysio- significant vascular stenosis, nevertheless, normal
logy of PTLD is that Epstein–Barr virus (EBV)- MRA findings reliably exclude the possibility of
induced B-cell proliferation, unopposed by the significant stenosis.83 CT and CT angiography
pharmacologically suppressed immune system, (CTA) are excellent at identifying transplant and
causes plasma cell hyperplasia, then premalig- peritransplant fluid collections, abscess forma-
nant polymorphic B-cell proliferation and even- tion, vascular complications, etc. However, due
tually malignant monoclonal lymphoma. If to the need for nephrotoxic contrast and the
untreated, it can be fatal.77 Reduction or cessa- radiation dose, ultrasound is the preferred screen-
tion of immunosuppression is the best treatment ing test, with CT and CTA used for confirma-
and usually results in tumour regression. Early tion and problem solving.
diagnosis prior to the development of frank
lymphoma is very important and these patients
SUMMARY
may have a much better response. In a large
study of PTLD in renal transplant recipients, it These are very rewarding times in the field of
occurred in 2% of 1383 patients and contributed organ transplantation. Advances in organ pro-
to death in over 50% of these cases.78 curement and preservation; better matching of
The liver is the organ most frequently involved donors and recipients; refined surgical tech-
by PTLD. It can appear as a focal mass, diffuse niques; availability of new, more effective
infiltration, or a periportal mass.79 PTLD can immunosuppressive agents; and improved post-
affect the transplanted kidney and may manifest transplant monitoring of organ recipients have
as a focal renal mass or diffuse infiltration.80 The contributed to decreased patient morbidity and
pancreas transplant may be involved, and can improved allograft survival. Although Doppler
present as diffuse enlargement of the allograft or sonography is only able to make a definitive
as a focal mass which may be confused with diagnosis in a small percentage of cases, it is
pancreatitis or acute rejection.81 extremely useful as a screening tool in the
Since ultrasound is often the first imaging study management of transplant complications. All of
performed when laboratory tests suggest trans- this has allowed transplant recipients greater
plant dysfunction, it plays an important role in opportunity to return to a more normal lifestyle
the early diagnosis of PTLD. It may detect urinary after surgery.84,85
or biliary obstruction associated with adenopathy
or may perceive a new ill-defined, typically
ACKNOWLEDGEMENTS
hypoechoic mass. Doppler may show vascular
distortion by the adenopathy.82 I would like to thank Dr Luis Fernandez,
Department of Surgery, University of Wisconsin
for manuscript review, Carrie Poole and Joan
OTHER IMAGING MODALITIES
Palmer for manuscript preparation, and Mike
MR angiography (MRA) has been found to be Ledwidge, RT, RDMS, for image preparation.
sensitive but not specific in the detection of
REFERENCES
1. Takemoto SK. HLA matching in the new 3. US Department of Health and Human Services,
millennium. Clin Transpl 2003; 387-403. Public Health Service. Annual report of the US
2. Gabardi S, Cerio J. Future immunosuppressive Scientific Registry of Transplant Recipients and the
agents in solid-organ transplantation. Prog Organ Procurement and Transplantation Network,
Transplant 2004; 14(2):148–156. 1995-2002. (2004 report, kidney transplant 241
9
survivors).Online. Available: http://www.optn.org/ 19. Platt JF, Rubin JM, Ellis JH. Distinction between
AR2004/default.htm July 27 2005. obstructive and non-obstructive pyelocaliectasis
4. Belzer FO. Transplantation of the right kidney: surgical with duplex Doppler sonography. Am J Radiol 1989;
technique revisited. Surgery 1991; 110:113–115. 153:997–1000.
5. Solinger HW, Ploeg RJ, Eckhoff DE, et al. Two 20. Platt JF, Ellis JH, Rubin JM. Renal transplant
hundred consecutive simultaneous pancreas-kidney pyelocaliectasis: role of duplex Doppler US in
transplants with bladder drainage. Surgery 1993; evaluation. Radiology 1991; 179:425–428.
114:736–743. 21. Dodd GD, Tublin ME, Shah A, et al. Imaging of
6. Pozniak MA, Zagzebski J, Scanlan KA. Spectral and vascular complications associated with renal
colour Doppler artifacts. Radiographics 1992; transplantation. Am J Radiol 1991; 157:449–459.
12:35–44. 22. Pozniak MA, Dodd GD, Kelcz F. Ultrasonographic
7. Pozniak MA, Kelcz F, Stratta R, et al. Extraneous evaluation of renal transplantation. Radiol Clin North
factors affecting resistive index. Invest Radiol 1988; Am 1992; 30:1053–1066.
23:899–904. 23. Grenier N, Douws C, Morel D, et al. Detection of
8. Benoit G, Blanchet P, Moukarzel M, et al. Surgical vascular complications in renal allografts with colour
complications in kidney transplantation. Transplant Doppler flow imaging. Radiology 1991; 178:217–223.
Proc 1994; 26:287–288. 24. Baxter GM, Ireland H, Moss JG, et al. Colour Doppler
9. Hashimoto Y, Nagano S, Ohsima S, et al. Surgical ultrasound in renal transplant artery stenosis: which
complications in kidney transplantation: experience Doppler index? Clin Radiol 1995; 50:618–622.
from 1200 transplants performed over 20 years at 25. Roberts JP, Ascher NL, Fryd DS, et al. Transplant renal
six hospitals in central Japan. Transplant Proc 1996; artery stenosis. Transplantation 1989; 4:580–583.
28:1465–1467. 26. Patriquin HB, Lafortune M, Jequier JC, et al.
10. Drudi FM, Cascone F, Pretagostini R, et al. Role of Stenosis of the renal artery: assessment of slowed
colour Doppler US in the evaluation of renal systole in the downstream circulation with Doppler
transplant. Radiol Med (Torino) 2001; 101(4):243–250. sonography. Radiology 1992; 184:479–485.
[Article in Italian] 27. Saarinen O, Salmela K, Edgren J. Doppler ultrasound
11. Pozniak MA, Kelcz F, D’Alessandro A, et al. in the diagnosis of renal transplant artery stenosis-
Sonography of renal transplants: the effect of acute value of resistive index. Acta Radiol 1994; 35:586–589.
tubular necrosis, cyclosporine nephrotoxicity, and 28. Handa N, Fukunaga R, Etani H, et al. Efficacy of
acute rejection on resistive index and renal length. echo-Doppler examination for the evaluation of
Am J Radiol 1992; 158:791–797. renovascular disease. Ultrasound Med Biol 1988;
12. Genkins SM, Sanfilippo FP, Carroll BA. Duplex 14:1–5.
Doppler sonography of renal transplants: lack of 29. Stavros AT, Parker SH, Yakes WF, et al. Segmental
sensitivity and specificity in establishing pathologic stenosis of the renal artery: pattern recognition of
diagnosis. Am J Radiol 1989; 152:535–539. tardus and parvus abnormalities with duplex
13. Kelcz F, Pozniak MA, Pirsch JD, et al. Pyramidal sonography. Radiology 1992; 184:487–492.
appearance and resistive index: insensitive and non- 30. Kribs SW, Rankin RN. Doppler ultrasonography after
specific sonographic indicators of renal transplant renal transplantation: value of reversed diastolic flow
rejection. Am J Radiol 1990; 155:531–535. in diagnosing renal vein obstruction. Can Assoc
14. Perrella RR, Duerincky AJ, Tessler FN, et al. Radiol J 1993; 44:434–438.
Evaluation of renal transplant dysfunction by duplex 31. Baxter GM, Morley P, Dall B. Acute renal vein
Doppler sonography: a prospective study and review thrombosis in renal allografts: new Doppler
of the literature. Am J Kidney Dis 1990; 15:544–550. ultrasonic findings. Clin Radiol 1991; 43(2):125–127.
15. Akiyama T, Ikegami M, Hara Y, et al. Haemodynamic 32. Middleton WD, Kellman GM, Melson GL, et al.
study of renal transplant chronic rejection using Postbiopsy renal transplant arteriovenous fistulas:
power Doppler sonography. Transplant Proc 1996; colour Doppler versus US characteristics. Radiology
28:1458–1460. 1989; 171:253–257.
16. Saarinen O. Diagnostic value of resistive index of 33. Hubsch PJS, Mostbeck G, Barton PP, et al.
renal transplants in the early postoperative period. Evaluation of arteriovenous fistulas and
Acta Radiol 1991; 32:166–169. pseudoaneurysms in renal allografts following
17. Koga S, Tanabe K, Yagisawa TT, et al. Urologic percutaneous needle biopsy: colour coded
complications in renal transplantation. Transplant Doppler sonography vs duplex Doppler sonography.
Proc 1996; 28:1472–1473. J Ultrasound Med 1990; 9:95–100.
18. Friedewald SM, Molmenti EP, Friedewald JJ, et al. 34. US Department of Health and Human Services,
Vascular and nonvascular complications of renal Public Health Service. Annual report of the US
transplants: sonographic evaluation and correlation Scientific Registry of Transplant Recipients and the
with other imaging modalities, surgery, and Organ Procurement and Transplantation Network,
pathology. J Clin Ultrasound 2005; 33(3):127–139. 1995–2002. (2004 report, liver transplant survivors).
242
Online. Available: http://www.optn.org/AR2004/ 48. Stell D, Downey D, Marotta P, et al. Prospective

9
References
default.htm July 27, 2005. evaluation of the role of quantitative Doppler
35. Ravindra KV, Guthrie JA, Woodley H, et al. ultrasound surveillance in liver transplantation.
Preoperative vascular imaging in pediatric liver Liver Transpl 2004; 10(9):1183–1188.
transplantation. J Pediatr Surg 2005; 40(4):643–647. 49. Propeck PA, Scanlan KA. Reversed or absent
36. Neuhaus P, Platz KP. Liver transplantation: newer hepatic arterial diastolic flow in liver transplants
surgical approaches. Baillieres Clin Gastroenterol shown by duplex sonography: a poor predictor of
1994; 8:481–493. subsequent hepatic artery thrombosis. AJR Am J
37. Houssin D, Boillot AO, Soubrane O, et al. Controlled Roentgenol 1992; 159:1199–1201.
liver splitting for transplantation in two recipients: 50. Otobe Y, Hashimoto T, Shimizu Y, et al. Formation of
technique, results and perspectives. Br J Surg a fatal arterioportal fistula following needle liver
1993; 80:75–80. biopsy in a child with a living-related liver transplant:
38. Dodd GD III, Memel DS, Zajko AB, et al. Hepatic report of a case. Surg Today 1995; 25:916–919.
artery stenosis and thrombosis in transplant 51. Jabbour N, Reyes J, Zajko A, et al. Arterioportal
recipients: Doppler diagnosis with resistive index fistula following liver biopsy: three cases occurring
and systolic acceleration time. Radiology 1994; in liver transplant recipients. Dig Dis Sci 1995;
192:657–661. 40:1041–1044.
39. Platt JF, Yotzy GG, Bude RO, et al. Use of Doppler 52. Doria C, Marino IR. Acute portal vein thrombosis
sonography for revealing hepatic artery stenosis in secondary to donor/recipient portal vein diameter
liver transplant recipients. AJR Am J Roentgenol mismatch after orthotopic liver transplantation:
1997; 168:473–476. a case report. Int Surg 2003; 88(4):184–187.
40. Sidhu PS, Ellis SM, Karani JB, et al. Hepatic artery 53. Davidson BR, Gibson M, Dick R, et al. Incidence,
stenosis following liver transplantation: significance risk factors, management, and outcome of portal
of the tardus parvus waveform and the role of vein abnormalities of orthotopic liver transplantation.
microbubble contrast media in the detection of a Transplantation 1994; 57(8):1174–1177.
focal stenosis. Clin Radiol 2002; 57(9):789–799. 54. Doria C, Marino IR. Acute portal vein thrombosis
41. Kaneko J, Sugawara Y, Akamatsu N, et al. secondary to donor/recipient portal vein diameter
Prediction of hepatic artery thrombosis by protocol mismatch after orthotopic liver transplantation:
Doppler ultrasonography in pediatric living donor a case report. Int Surg 2003; 88(4):184–187.
liver transplantation. Abdom Imaging 2004; 55. Cheng YF, Chen CL, Huang TL, et al. Risk factors
29(5):603–605. for intraoperative portal vein thrombosis in pediatric
42. Sidhu PS, Shaw AS, Ellis SM, et al. Microbubble living donor liver transplantation. Clin Transplant
ultrasound contrast in the assessment of hepatic 2004; 18(4):390–394.
artery patency following liver transplantation: role in 56. Corno V, Torri E, Bertani A, et al. Early portal vein
reducing frequency of hepatic artery arteriography. thrombosis after pediatric split liver transplantation
Eur Radiol 2004; 14(1):21–30. with left lateral segment graft. Transplant Proc
43. Herold C, Reck T, Ott R, et al. Contrast-enhanced 2005; 37(2):1141–1142.
ultrasound improves hepatic vessel visualization 57. Durham JD, LaBerge JM, Kam I, et al. Portal vein
after orthotopic liver transplantation. Abdom Imaging thrombolysis and closure of competitive shunts
2001; 26(6):597–600. following liver transplantation. J Vasc Interv Radiol
44. Vivarelli M, Cucchetti A, La Barba G, et al. Ischemic 1994; 5:611–616.
arterial complications after liver transplantation in 58. Fujimoto M, Moriyasu F, Someda H, et al. Evaluation
the adult: multivariate analysis of risk factors. of portal haemodynamics with Doppler ultrasound
Arch Surg 2004; 139(10):1069–1074. in living related donor liver transplantation in children:
45. Kok T, Haagsma EB, Klompmaker IJ, et al. Doppler implications for ligation of spontaneous portosystemic
ultrasound of the hepatic artery and vein performed collateral pathways. Transplant Proc 995; 27:1174–1176.
daily in the first two weeks after orthotopic liver 59. Nishida S, Kadono J, DeFaria W, et al. Gastroduodenal
transplantation. Invest Radiol 1996; 31:173–179. artery steal syndrome during liver transplantation:
46. Garcia-Criado A, Gilabert R, Salmeron JM, et al. intraoperative diagnosis with Doppler ultrasound
Significance of and contributing factors for a high and management. Transpl Int 2005; 18(3):350–353.
resistive index on Doppler sonography of the 60. Tang SS, Shimizu T, Kishimoto R, et al. Analysis of
hepatic artery immediately after surgery: prognostic portal venous waveform after living-related liver
implications for liver transplant recipients. AJR Am transplantation with pulsed Doppler ultrasound.
J Roentgenol 2003; 181(3):831–838. Clin Transplant 2001; 15(6):380–387.
47. Hall TR, McDiarmid SV, Grant EG, et al. False- 61. Salizzoni M, Andorno E, Bossuto E, et al. Piggyback
negative duplex Doppler studies in children with techniques versus classical technique in orthotopic
hepatic artery thrombosis after liver transplantation. liver transplantation: a review of 75 cases. Transplant
AJR Am J Roentgenol 1990; 154:573–575. Proc 1994; 26:3552.
243
9
62. Akamatsu N, Sugawara Y, Kaneko J, et al. Surgical 74. Tobben PJ, Zajko AB, Sumkin JH, et al.
repair for late-onset hepatic venous outflow block Pseudoaneurysms complicating organ transplantation:
after living-donor liver transplantation. Transplantation roles of CT, duplex sonography, and angiography.
2004; 77(11):1768–1770. Radiology 1988; 169:65–70.
63. Carnevale FC, Borges MV, Pinto RA, et al. 75. Green BT, Tuttle-Newhall J, Suhocki P, et al.
Endovascular treatment of stenosis between Massive gastrointestinal hemorrhage due to rupture
hepatic vein and inferior vena cava following liver of a donor pancreatic artery pseudoaneurysm in a
transplantation in a child: a case report. Pediatr pancreas transplant patient. Clin Transplant 2004;
Transplant 2004; 8(6):576–580. 18(1):108–111.
64. Totsuka E, Hakamada K, Narumi S, et al. Hepatic 76. Wong JJ, Krebs TL, Klassen DK, et al. Sonographic
vein anastomotic stricture after living donor liver evaluation of acute pancreatic transplant rejection:
transplantation. Transplant Proc 2004; 36(8):2252–2254. morphology - Doppler analysis versus guided
65. Rossi AR, Pozniak MA, Zarvan NP. Upper inferior percutaneous biopsy. AJR Am J Roentgenol 1996;
vena caval anastomotic stenosis in liver transplant 166:803–807.
recipients: Doppler US diagnosis. Radiology 1993; 77. Nalesnik MA. The diverse pathology of post-
187:387. transplant lymphoproliferative disorders: the
66. Ko EY, Kim TK, Kim PN, et al. Hepatic vein stenosis importance of a standardized approach. Transpl
after living donor liver transplantation: evaluation Infect Dis 2001; 3:88–96.
with Doppler US. Radiology 2003; 229(3):806–810. 78. Bates WD, Gray DWR, Dada MA, et al.
67. Jequier S, Jequier JC, Hanquinet S, et al. Lymphoproliferative disorders in Oxford renal
Orthotopic liver transplants in children: change in transplant recipients. J Clin Pathol 2003; 56:439–446.
hepatic venous Doppler wave pattern as an 79. Pickhardt PJ, Siegel MJ. Posttransplantation
indicator of acute rejection. Radiology 2003; lymphoproliferative disorder of the abdomen:
226(1):105–112. CT evaluation in 51 patients. Radiology 1999;
68. Van Thiel DH, Gavaler JS, Wright H, et al. Liver 213:73–78.
biopsy: its safety and complications as seen at a 80. Vrachliotis TG, Vaswani KK, Davies EA, et al. CT
liver transplant center. Transplantation 1993; findings in posttransplantation lymphoproliferative
55:1087–1090. disorder of renal transplants. AJR Am J Roentgenol
69. Baccarani U, Adani GL, Sanna A, et al. Portal vein 2000; 175:183–188.
thrombosis after intraportal hepatocytes 81. Meador TL, Krebs TL, Wong You Cheong JJ,
transplantation in a liver transplant recipient. et al. Imaging features of posttransplantation
Transpl Int 2005; 18(6):750–754. lymphoproliferative disorder in pancreas transplant
70. US Department of Health and Human Services, recipients. AJR Am J Roentgenol 2000; 174:121–124.
Public Health Service. Annual report of the US 82. Scarsbrooka AF, Warakaullea DR, Dattania M, et al.
Scientific Registry of Transplant Recipients and the Post-transplantation lymphoproliferative disorder:
Organ Procurement and Transplantation Network, the spectrum of imaging appearances. Clin Radiol
1995–2002. (2004 report, pancreas transplant 2005; 60:47–55.
survivors).Online. Available: http://www.optn.org/ 83. Kim BS, Kim TK, Jung DJ, et al. Vascular complications
AR2004/default.htm July 27 2005. after living related liver transplantation: evaluation
71. Nikolaidis P, Amin RS, Hwang CM, et al. Role of with gadolinium-enhanced three-dimensional MR
sonography in pancreatic transplantation. angiography. AJR Am J Roentgenol 2003;
Radiographics 2003; 23:939–949. 181(2):467–474.
72. Foshager MC, Hedlund LJ, Troppmann C, et al. 84. Lee HM. Quality of life after renal transplantation.
Venous thrombosis of pancreatic transplants: Transpl Proc 1996; 28:1171.
diagnosis by duplex sonography. AJR Am J 85. Park IH, Yoo HJ, Han DJ, et al. Changes in the
Roentgenol 1997; 169:1269–1273. quality of life before and after renal transplantation
73. Spiros D, Christos D, John B, et al. Vascular and comparison of the quality of life between
complications of pancreas transplantation. kidney transplant recipients, dialysis patients, and
Pancreas 2004; 28(4):413–420. normal controls. Transpl Proc 1996; 28:1937–1938.
244
10
Doppler imaging of the
prostate
Fred T. Lee, Jr
INDICATIONS Prostate vascular anatomy

The prostate is supplied from two arterial sources:
The most important use of colour Doppler the prostatic arteries and the inferior vesical
imaging of the prostate remains as an aid in arteries, both arising from the internal iliac
cancer detection. This is particularly relevant in system. The prostatic arteries enter the prostate
patients in whom cancer is suspected based on from an anterolateral location on each side, and
prostate specific antigen (PSA) elevation without give off capsular branches as well as urethral
obvious tumour on grey-scale imaging. Other branches. Capsular arteries course along the
uses for Doppler imaging are largely confined to lateral margin of the prostate, and give off
detection of prostatitis and inflammatory condi- numerous perforating branches which penetrate
tions. Controversy continues surrounding diag- the capsule and supply approximately two-thirds
nosis and treatment of prostate cancer. This is of the total glandular tissue. The areas of pene-
largely attributable to the wide range of bio- tration into the capsule are commonly referred
logical behaviour found with this disease. Up to to as the neurovascular bundles (Fig. 10.2). The
30% of 80-year-old males will have histological inferior vesical arteries run along the inferior
evidence of prostate cancer, yet most will die surface of the bladder and also provide urethral
from other causes. Unfortunately, a more aggres- branches. In addition to supplying the central
sive subset remains an important cause of mortality portion of the prostate, the inferior vesical
among men, with 30350 deaths expected in the arteries also give off branches which supply the
USA in 2005.1 bladder base, seminal vesicles and distal ureters
(Fig. 10.3).3,4 Both the capsular and urethral
branches can be visualised with colour Doppler
ANATOMY
ultrasound. In the absence of inflammation,
The prostate lies immediately anterior to the neoplasm or hypertrophy the normal prostate
rectum and inferior to the bladder. Prostatic is expected to have low level periurethral and
zonal anatomy has been extensively described by pericapsular flow, with only a low level of flow in
McNeal et al.2 In summary, the prostate is the prostatic parenchyma.5
composed of three major zonal areas; the
peripheral zone, the central zone and the transi-
EQUIPMENT AND TECHNIQUE
tion zone (Fig. 10.1). The peripheral zone is the
most posterior, and the central zone is a contin- Examination of the prostate by ultrasound
uation of the peripheral zone cephalad. The requires a high-frequency (5–7.5 MHz) end-fire
transition zone is the most central area of the or biplane transrectal transducer. For the
prostate, and surrounds the urethra as it courses purposes of this chapter, conventional colour
through the prostate.The anterior fibromuscular Doppler and power Doppler are considered
stroma lines the prostate anteriorally. simultaneously. For most general applications, 245
10
Doppler imaging of the prostate
an end-fire transducer is favoured due to the
ease of switching between axial (coronal) and
longitudinal imaging planes, as well as the more
favourable angle for transrectal prostatic biopsies.
For specialised applications such as prostatic
volumetry and cryosurgery, a true biplane trans-
ducer is necessary.
No specific patient preparation is required
∗ ∗ although some centres will give the patient a
pre-examination enema and have them empty
Fig. 10.1 Axial ultrasound of the prostate in a their bladder. The patient is generally placed in
normal patient. Note peripheral zone (*) separated the left lateral decubitus position, and the knees
from the more centrally oriented, periurethral, brought up to the chest. A digital rectal exami-
transition zone by the surgical capsule (arrows). nation is recommended prior to probe insertion
to rule out any obstructing pathology and also to
allow the examiner to evaluate the prostate by
digital examination. The probe is covered with
a condom into which coupling gel has been
placed, and the probe lubricated and gently
inserted into the rectal canal. Examination of the
prostate by grey-scale imaging is first performed,
and the length, width and height of the gland
measured. The prostatic volume is calculated
based on the formula for a prolate ellipsoid
(length × width × height × 0.523); this allows
correlation of the measured PSA with a predicted
PSA based on gland volume. Normal prostatic
Fig. 10.2 Axial image of the left neurovascular tissue produces approximately 0.3 ng cc-1 of PSA,
bundle. Note left neurovascular bundle (arrow) with whereas cancerous tissue produces approximately
perforating branches penetrating into the prostate 3.0 ng cc-1 of tumour. Normal levels for polyclonal
(arrowheads). assays are typically defined as <4.0 ng mL-1;
unfortunately, up to 20% of prostatic cancers
present in patients with ‘normal’ levels of PSA.
A ‘predicted’ PSA can be generated based on
the patient’s gland volume × 0.2 for polyclonal
assays (or gland volume × 0.1 for monoclonal
∗ assays). A level of measured PSA that exceeds
predicted PSA increases the suspicion of cancer
and increases the positive predictive value of
prostatic biopsy.6
Most prostate cancers (70%) arise in the
peripheral zone, with a minority originating in
the central (10–15%) and transition zones
Fig. 10.3 Sagittal image of the prostate at the level (10–15%). Because of this, it is very important
of the seminal vesicle (*) demonstrates periurethral that the sonographer carefully examine the
flow (arrows) originating from the inferior vesicle peripheral zone for signs of tumour. Virtually all
246 artery. prostate cancers will be hypoechoic in relation to
10
Colour Doppler of prostate cancer

normal peripheral zone tissues (Fig. 10.4),
although a minority of cribriform carcinomas
can demonstrate punctate calcifications.Tumours
in the peripheral zone have ready access to sites
of anatomical weakness, including the neurovas-
cular bundles, ejaculatory ducts and apex of the
gland. This results in more aggressive clinical
behaviour of peripheral zone tumours when
compared to other locations. Transition zone Fig. 10.4 Prostate cancer. Axial image of the prostate
tumours tend to behave in a clinically more at the mid-gland. Hypoechoic tumour (+) originates in
the left neurovascular bundle area.
benign manner because they are distant from
sites of anatomical weakness, and thus need to
grow quite large before spreading outside of the
gland. The main problem with the diagnosis of examination of the transition zone should be
transition zone tumours is the heterogeneous undertaken. It is in the search for transition zone
echotexture of the normal transition zone. tumours that colour Doppler ultrasound has
Because normal transition zone tissue can be proven to be most useful. Prostate cancer is
hypoechoic, hyperechoic or contain calcifications generally hypervascular when compared to
or cysts, it is extremely difficult to diagnose subtle normal prostatic tissue and this is manifested as
changes in echogenicity that may be associated increased colour encoding at sensitive instru-
with neoplasia.Therefore, colour Doppler can play ment settings (Fig. 10.5). These can be targeted
a crucial role in the diagnosis of transition zone for biopsy with increased positive biopsy rates
tumours by identifying areas of abnormal flow. compared to blinded sextant biopsies. It is
controversial as to whether targeted biopsies
using color Doppler alone can replace sextant
COLOUR DOPPLER OF PROSTATE
biopsies.7 Early work using contrast enhanced
CANCER
color Doppler ultrasound demonstrates an
Knowledge of the excess PSA for a particular increased sensitivity for the detection of prostate
gland is most important from an ultrasound cancer,8, 9 but the exact type of contrast material,
standpoint when an obvious peripheral zone imaging algorithm and time after injection has
tumour is not found in the face of an elevated not yet been standardised. Additionally, ultra-
measured PSA. As previously mentioned, sound contrast agents have not yet been approved
transition zone tumours are difficult to visualise for use in the USA. Spectral Doppler plays a
by grey-scale criteria due to the in homogeneous limited role in the specific diagnosis of prostate
nature of normal transition zone tissue. Once cancer. Tumours tend to have low resistance
it is established that the patient is at high risk (high diastolic) flow, although the exact role and
for prostate cancer by PSA criteria, and no specificity of this finding has yet to be fully
peripheral zone cancer has been found, a careful elucidated.
Fig. 10.5 Transition zone prostate

cancer (biopsy proven). Axial (left)
and sagittal (right) images
demonstrate a hypervascular
tumour in the transition zone
(arrows).
247
10
The use of colour Doppler in the diagnosis of titis, ranging from bacterial to non-bacterial
peripheral zone tumours is more controversial. causes. In the case of bacterial prostatitis, the
Several authors have found increased colour offending organism is usually Escherichia coli or
Doppler flow to have no significant correlation other urinary tract pathogens.
with the presence or absence of tumour at Grey-scale findings of acute prostatitis include
histology. In addition, there has been no colour an hypoechoic rim around the prostate or peri-
Doppler method consistently to discriminate urethral areas, and low level echogenic areas
tumour from focal prostatitis in areas of increased within the prostate.12 Colour Doppler is useful
flow. Others have found biopsy of sites of in cases of diffuse bacterial prostatitis.The severity
increased flow useful in the face of an increased of the inflammatory reaction is mirrored by
measured PSA (greater than predicted) and no focal or diffuse increase in the colour signal in
other obvious sites of tumour.10 This has been the prostatic parenchyma.13 When focally
found to be particularly useful in black males, increased colour signals are seen in cases of
where the positive predictive value for biopsy of prostatitis, there is no reliable non-invasive
a focal area of increased colour encoding has method to differentiate inflammation from
been found to be twice that of white males (32.2% tumour.13 However, cases of grossly increased
vs 13.5% respectively).11 Most authors now feel flow spread diffusely throughout the gland should
that colour Doppler is more of a complementary be considered prostatitis in the appropriate clinical
test to grey-scale ultrasound, PSA and gland setting (Fig. 10.7). When the inflammatory
volume rather than a single factor on which to process continues to suppuration, a prostatic
base biopsy decisions (Fig. 10.6). abscess can develop. On ultrasound, this is seen
as a cavity filled with low-level echoes from
debris (Fig. 10.8).14 Colour Doppler may detect
COLOUR DOPPLER OF PROSTATIC
increased flow around the rim of the cavity,
INFLAMMATORY DISEASE
although this finding is not necessary to make
Prostatitis is a difficult condition to diagnose the diagnosis. Cases of bacterial prostatitis are
and treat.There are several aetiologies of prosta- treated by antibiotics, whereas prostatic abscess
a Fig. 10.6 Peripheral zone

prostate cancer. (a) Axial and
sagittal grey-scale images
demonstrate a subtle hypoechoic
area in the left neurovascular
bundle (arrows). Biopsy through
this area was positive for
adenocarcinoma, Gleason score
Axial Sagittal 6. (b) Axial and sagittal colour
b Doppler images at corresponding
locations demonstrate increased
colour flow in areas involved by
tumour.
248
10
Conclusions
requires transrectal catheter or transurethral
drainage with unroofing of the abscess cavity.
CONCLUSIONS
Doppler ultrasound of the prostate contributes
significantly to the diagnostic value of sonogra-
phy in the assessment of prostatic disease. Colour
and power Doppler identify areas of abnormal
blood flow, which can then be examined more
Fig. 10.7 Prostatitis. Colour Doppler image of closely with grey-scale imaging, or biopsied under
diffuse prostatitis demonstrates grossly increased
ultrasound guidance.
flow throughout the gland.
Fig. 10.8 Prostatic abscess.

Markedly hypoechoic lesion with
subtle through transmission is
present in the peripheral zone of
this patient. Note lack of flow in
the central portion of this lesion,
a finding that would be very
unusual for prostate cancer.
Drainage confirmed the presence
of an abscess.
REFERENCES
1. Jemal A, Murray T, Ward E, et al. CA cancer. J Clin 9. Bogers HA, Sedelaar JP, Beerlage HP, et al.
2005; 55:10–30. Contrast-enhanced three-dimensional power
2. McNeal JE. Regional morphology and pathology of Doppler angiography of the human prostate:
the prostate. Am J Clin Pathol 1968; 49:347–357. correlation with biopsy outcome. Urology 1999;
3. Flocks RH. The arterial distribution within the 54:97–104.
prostate gland: its role in transurethral prostatic 10. DeCarvalho VS, Soto JA, Guidone PL, et al. Role of
resection. J Urol 1937; 37:524–548. colour Doppler in improving the detection of cancer
4. Clegg EJ. The arterial supply of the human prostate in the isoechoic prostate gland (abstr). Radiology
and seminal vesicles. J Anat 1955; 89:209–217. 1995; 197(P):365.
5. Neumaier CE, Martinoli C, Derchi LE, et al. Normal 11. Littrup PJ, Klein RM, Sparschu RA, et al. Colour
prostate gland: examination with colour Doppler Doppler of the prostate: histologic and racial
US. Radiology 1995; 196:453–457. correlations (abstr). Radiology 1995; 197(P):365.
6. Lee F, Littrup PJ, Loft-Christensen L, et al. Predicted 12. Griffiths GJ, Crooks AJR, Roberts EE, et al.
prostate specific antigen results using transrectal Ultrasonic appearances associated with prostatic
ultrasound gland volume: Differentiation of benign inflammation: A preliminary study. Clin Radiol 1984;
prostatic hyperplasia and prostate cancer. Cancer 35:343–345.
1992; 70:211–220. 13. Patel U, Rickards D. The diagnostic value of colour
7. Halpern EJ, Frauscher F, Strup SE, et al. Prostate: Doppler flow in the peripheral zone of the prostate,
high-frequency Doppler US imaging for cancer with histological correlation. Br J Urol 1994;
detection. Radiology 2002; 225:71–77. 74:590–595.
8. Frauscher F, Klauser A, Halpern EJ, et al. Detection 14. Lee FT Jr, Lee F, Solomon MH, et al. Ultrasonic
of prostate cancer with a microbubble ultrasound demonstration of prostatic abscess. J Ultrasound
contrast agent. Lancet 2001; 357:1849–1850. Med 1986; 5:101–102. 249
11
penis
Myron A. Pozniak and Fred T. Lee, Jr
hyperlipidaemia. Most centres will now prescribe

a trial of the phosphodiesterase inhibitors as the
INTRODUCTION
initial diagnostic/therapeutic test, with only non-
Duplex sonography with colour Doppler is the responders being sent on for further evaluation.
modality of choice for evaluating the penis. The Doppler and ultrasound imaging can be used
advantages of Doppler over other imaging modali- to identify arterial or venous injury following acute
ties are ease of performing the study, patient penile trauma.3,9,10 Doppler and ultrasound can
acceptance, versatility, minimally invasive nature, be used to evaluate patients with Peyronie’s disease,
reproducibility, availability and relatively low cost. an idiopathic process which produces fibrous
Colour Doppler allows the examiner to delineate plaques in the penile tunica albuginea,3,11,12
vascular anatomy, display dynamic variations in specifically to determine if these plaques cause
blood flow, measure arterial velocity and infer any significant vascular compromise.13
adequacy of venous drainage. By combining
Doppler with grey-scale imaging and pharma-
PENILE ANATOMY
cological induction of erection, both anatomical
AND PHYSIOLOGY
and physiological abnormalities can be assessed
during the flaccid and erect states.1–5 The penis contains three longitudinal, cylindrical
erectile bodies.Two corpora cavernosa are located
in the dorsal two-thirds of the penile shaft, and
INDICATIONS
a single corpora spongiosum is located in the
The primary application of penile Doppler imaging ventral one-third of the shaft.The corpora caver-
is to assess the vasculature in patients with erectile nosa are enclosed by the tunica albuginea, a
dysfunction (ED), the prevalence and the severity tough, non-distensible fascial layer. The septum
of which increase with advancing age.6 While that divides the corpora cavernosa contains fenes-
impotence may be the result of psychogenic, trations that create multiple connecting anasto-
neurogenic, or hormonal factors, vascular disease motic channels between the sinusoidal spaces,
is one of the leading causes of ED.4,7 With the allowing for free communication across the
introduction of phosphodiesterase inhibitors midline. The dorsal arteries, veins and nerves
[specifically sildenafil citrate (Viagra©),Vardenafil are situated centrally along the penile dorsum,
HCl (Livetra©), and Tadalafil (Cialis©)], the superficial to the tunica albuginea and deep to
frequency of Doppler studies for penile impo- Buck’s fascia. The urethra is contained within
tence has notably decreased.8 However, diagnostic the corpus spongiosium.14
imaging of ED is still considered useful because On ultrasound, the corpora cavernosa are of
ED can be the presenting symptom for a variety uniform hypoechoic echotexture.The tunica can
of diseases, such as diabetes mellitus, coronary be seen as an echogenic envelope around the
artery disease, atherosclerosis, hypertension and corpora. The echogenic walls of the cavernosal 251
11
Doppler imaging of the penis
arteries can be seen centrally within the corpora. associated with a bulbar parenchymal blush in
The corpus spongiosum is of higher echogenicity the early arterial phase.The urethral artery, which
(Fig. 11.1). is of small diameter, arises anterior to the bulbar
artery. After these branches, the internal pudendal
Arterial anatomy artery continues as the common penile artery. It
The internal pudendal artery and its branches are then divides into left and right penile arteries, which
the primary source of arterial supply to the enter the base of the penis and branch into a
penis.The first three branches are the superficial dorsal artery and a cavernosal artery. The dorsal
perineal artery, the bulbar artery and a small artery extends along the dorsal aspect of the penile
urethral artery. The perineal artery is a large and shaft towards the glans and terminates at the level
constant branch that, in 80% of cases, has an of the arterial corona of the glans; it supplies
internal and external branch. The bulbar artery, blood primarily to the skin, subcutaneous tissues
which supplies the proximal penile shaft, is usually and glans. Collateral vessels from the dorsal artery
easily identified during angiography because it is often communicate with the cavernosal artery.
C C
Fig. 11.1 (a) Transverse and (b) longitudinal ultrasound

c images of a normal penis. The corpora cavernosa (C)
have uniform echogenicity throughout. The echogenic
H tunica (open arrow) and echogenic walls of the
cavernosal artery (arrow) are easily identified.
(c) Colour Doppler flow. The cavernosal artery and
its helicine branches (H) are revealed.
252
11
Penile anatomy and physiology

The cavernosal, or deep penile artery, enters the and the cavernous venules are open.4 The emissary
tunica albuginea proximally and extends the length veins drain the sinusoidal spaces and blood circu-
of the corpus cavernosum.The cavernosal arteries lates into the dorsal veins. During this state, there
and their helicine branches are the primary is high resistance to blood flow into the penis.
source of blood flow to the erectile tissue of the Erection starts when an autonomic neurogenic
penis. Just as the cavernosal artery supplies blood impulse relaxes the cavernosal arterioles and
to the corpus cavernosum, the spongiosal artery sinusoidal spaces.2 As erection occurs, there is a
supplies the corpus spongiosum (Fig. 11.2). marked increase in the volume of arterial inflow
into the penis as the cavernous arteries dilate.
Venous anatomy This is accompanied by relaxation of the smooth
Venous drainage of the penile erectile tissue (i.e. muscle of the corpora cavernosa with expansion
the sinusoidal spaces) primarily occurs through and elongation of the cavernous sinusoids as they
emissary (efferent) veins which drain the corpus fill with blood. Compression of the cavernous
cavernosum, penetrate the tunica albuginea and venules between the dilated cavernous sinusoids
empty into the circumflex veins; these then drain and the unyielding peripheral tunica albuginea
into the deep dorsal venous system of the penis. decreases venous outflow. This veno-occlusive
The emissary veins may also drain directly into the mechanism (which depends on neurological
deep dorsal vein. The superficial dorsal vein drains stimuli, a sufficient supply of arterial blood, and
the distal portion of the corpora cavernosa, as normal function of the tunica albuginea) maintains
well as the skin and glans. The deep and super- sinusoidal distension and rigid erection.1,2,15
ficial dorsal veins can be routinely visualised by Five stages of erectile physiology have been
colour Doppler imaging in the midline of the defined by Lue: latent, tumescent, full erection,
penile shaft. The most proximal portions of the rigid erection and detumescent.16 During the latent
corpora cavernosa are drained by the cavernosal phase, the diameters of the cavernosal arteries
veins directly into the periprostatic plexus. are at their greatest and there is maximum inflow
of blood with minimal resistance. During tumes-
Erectile physiology cence, the sinusoidal cavities of the corpora
When the penis is flaccid, its smooth muscle is cavernosa distend with blood.With full erection,
in a tonic state, the cavernous sinusoids are collapsed, blood flow decreases as do the diameters of the
Fig. 11.2 Normal anatomy.

Dorsal The cavernosal arteries are centrally
located in each corpus cavernosus.
The urethra courses through the
corpus spongiosum. The dorsal
Dorsal penile artery supplies the glands
penile artery and does not play a direct role in
erectile function.
Cavernosal
artery
Corpus
cavernosum
Urethra
Corpus
spongiosum
Ventral
253
11
cavernosal arteries. With a rigid erection, blood The diameter of the arteries and blood flow
inflow (and outflow) ceases and the diameters of velocities are measured. Colour Doppler enhances
the cavernosal arteries are at their narrowest. the accuracy of angle correction, which is
Detumescence occurs when the trabeculae and mandatory for flow velocity determination. In
arteries contract in response to a release of norepi- addition, with colour Doppler, blood flow direc-
nephrine. During the five stages of erection, tion can be assessed and the presence of any
different arterial diameters and waveform patterns communications between the cavernosal, dorsal and
are normally present on Doppler examination. spongiosal arteries can be detected (Fig. 11.3).
Ultrasound technique ERECTILE DYSFUNCTION

Various techniques for the sonographic exami-
nation of the penis have been described with The historical evolution of treatment for impo-
changes over time primarily due to advances in tence has gone through several stages. Initially
ultrasound technology.1-4,17,18 A linear trans- treatment involved the use of penile implants. In
ducer operating at 7 MHz or higher frequency the 1980s, implant complication rates were quite
should be employed. Slow flow sensitivity must high and there was only about a 50% chance of
be optimised. Filters are set at their lowest levels long-term success.These devices failed, migrated
and the Doppler gain is set just below the noise or ruptured often necessitating reoperation.This
threshold; the pulse repetition frequency is set to often resulted in stricture and scarring, further
the lowest velocity setting possible. All of these complicating the problem. The risk to benefit
settings can then be adjusted if higher than with the old treatment was high so numerous
expected velocity distorts the Doppler display. imaging and testing studies were performed to
The evaluation should be performed in a quiet, confirm definitively that the patient had an organic
private setting with the room comfortably warm cause of impotence requiring surgery and not
and darkened, so that the patient is relaxed and some other aetiology that did not require surgical
not embarrassed. The scan is performed with intervention. Treatment of impotence then
the patient lying supine and the penis in the evolved through the use of vacuum devices and
anatomical position (lying superiorly against the self-injection of prostaglandins such as alprostadil
anterior abdominal wall). During the examina- (Caverject) and now is quite successfully accom-
tion, the patient may be asked to help keep the plished with phosphodiesterase inhibitors.
penis immobilised by gently holding the corona The initial report of Doppler sonography
just under the glans penis and then stretching combined with pharmacological induction of
the shaft along the anterior abdominal wall. erection to evaluate vasculogenic impotence was
Scanning is usually performed on the ventral by Lue and associates17 in the mid-1980s. With
surface of the penis,1, 3 but the probe can be placed advances in Doppler technology, the haemo-
on the dorsal or lateral surfaces if necessary.2,3 dynamic information available greatly enhances
Imaging is performed in both the longitudinal the ability to determine if a patient’s impotence
and transverse planes from the base of the penis is due to a vascular aetiology, such as arterial
to the glans to visualise anatomical details of the insufficiency, venous incompetence or a combi-
corpora cavernosa, cavernosal arteries and nation of the two.2,4,17–24
surrounding structures, and also to demonstrate Currently the application of Doppler ultrasound
any abnormalities such as fibrosis, scarring, plaques, is primarily in young men who present concerned
calcification, haematoma or tumour. The trans- that they have an organic issue. Testing with
ducer should be applied gently with minimal Doppler is needed to establish the fact that there
penile compression. Firm pressure causing vascular is indeed no vasculogenic aetiology. The patient
compression can resist inflow and affect accuracy is then channelled to psychological counselling.
254 of velocity measurement, especially during diastole. The younger patient is usually reluctant to initially
Fig. 11.3 Ultrasound technique.

11
Pharmacological induction of erection

A linear transducer is placed in a
longitudinal plane along the ventral
surface of the penis. Since the
cavernosal arteries run parallel to
the transducer, electronic steering
Ventral of the Doppler beam is necessary
to interrogate at an appropriate
angle.
Dorsal Cavernosal
penile artery artery
Deep dorsal vein

Superficial
dorsal vein
Dorsal
admit the need for psychological counselling combined with 10g of alprostadil.1–4,25–30 PGE-1
until Doppler ultrasound and other tests indeed is considered to be slightly safer but is considerably
prove that there is no underlying organic cause more expensive than papaverine. Unfortunately
for the impotence. there is a higher incidence of priapism with
papaverine and PGE-E1 use is increasing.
Gontero et al31 suggest performing a two-stage
PHARMACOLOGICAL INDUCTION
assessment with Doppler ultrasound after a
OF ERECTION
starting dose of 20 μg of PGE-1. If erectile
A vasodilating agent is injected into the corpus response is suboptimal and venous leak is
cavernosum near the base of the penis to induce suspect, then an additional 2 mg intracavernosal
erection. A number of different agents are available, phentolamine is administered and measurements
including papaverine, phentolamine, prostaglandin repeated. They claim high accuracy and a low
E1 (PGE-1) and alprostadil. A standard protocol risk of priapism.31
has not been established and a variety of tech- Some patients may have a poor response to
niques for using these agents has been reported, the initial injection of the vasodilating agent and
either singularly or in various combinations. may be unable to achieve satisfactory erection.
Reported protocols include 15 mg to 60 mg In these cases, manual self-stimulation has been
papaverine alone; 30 mg papaverine combined reported to significantly improve erection in many
with 1 mg phentolamine; 0.25 mL of a triple patients.32 For some patients, a second or third
mixture of 75 mg papaverine, 2.5 mg phento- injection may be required before an adequate
lamine and 25 g/4.25 mL PGE-1; a 0.2 mL total response is attained.3,4 However, this may run
volume of 6 mg papaverine, 0.2 mg phento- the risk of priapism, a painful complication of
lamine and 2 g PGE-1; and 12.5 mg papaverine pharmacological induction. 255
11
constitutes normal and abnormal values. Connolly
DOPPLER EVALUATION AFTER
et al4 have suggested that the most important
VASODILATOR INJECTION
diagnostic indicators are arterial diameter, peak
Just as there are differences in the use of vasodi- flow velocity and blood flow acceleration.
lating agents, there is no collective consensus as To obtain accurate readings, it is important
to how soon, how frequently and how long after that the examiner is familiar with the physiology
injection sonographic measurements should be of arterial inflow to the cavernosal arteries after
performed. However, it is now generally recog- pharmacological injection. Particular attention
nised that waiting until 5 min after injection, as must be paid to the Doppler angle, which should
had previously been the practice, can result in be approximately 60°, or less, to the cavernosal
falsely low peak systolic velocity (peak systolic artery.1,3 Measurements are most reliable and
velocity) findings in the cavernosal arteries, since most easily reproduced when taken at the base
normal maximum peak systolic velocity can occur of the penis where the penile vessels angle
in less than 5 min following injection.3 In addition, posteriorly toward the perineum. The arterial
premature termination of post-injection measure- diameter and waveform of each cavernosal artery
ments (i.e. after 5–10 min) can result in false- is individually assessed. Peak systolic and end-
positive diagnoses of arterial insufficiency, or diastolic velocities are measured and recorded.
venous incompetence because of temporal varia- An asymmetric response of the cavernosal arteries
tions in the response to vasodilating agents.2 Early during erection or a lack of arterial dilation may
termination may also result in a false-negative suggest the presence of a significant vascular
diagnosis of venous incompetence. Suggested lesion.33 The examiner should also carefully search
protocols are summarised in Table 11.1. for anatomical penile arterial variants as they
Doppler evaluation following the injection of may also contribute to vasculogenic impotence.
a vasodilating agent is performed so the penile
anatomical vasculature can be visualised and Normal Doppler findings
haemodynamic parameters measured. There is Prior to injection, during the flaccid state, the
considerable debate, however, as to which haemo- systolic waveform is damped and a monophasic
dynamic parameters are significant and what flow with minimal diastolic component is present
Table 11.1 Recommended time protocols for sonographic

evaluation of pharmacologically-induced erection
Investigators Suggested protocols

Hattery et al1 5, 10, 15, and 20 min following injection, with
delayed measurements occasionally obtained after
20 min if the degree of tumescence or rigidity is still
increasing
Fitzgerald et al2 Immediately after injection and at 5 min intervals for
20–30 min or until waveform progression ceases
Herbener et al3 Immediately after injection and at 1–2 min intervals
until there is a plateau of velocity or until velocities
peak and then decrease, with examination normally
not exceeding 20 min
Herbener et al3 3–5 min intervals until maximum cavernosal artery
diameter and peak systolic velocity measurements
are obtained
Connolly et al4 3–5 min intervals throughout latent, tumescent and
full stages of erection for up to 30 min in order to
detect patients who show delayed but eventually
normal response to injection
256
11
Doppler evaluation after vasodilator injection

(Fig. 11.4). Following pharmacological induce-
ment of erection, the normal progression of
haemodynamic events and the associated Doppler
waveform patterns of the cavernosal artery can
be classified into different haemodynamic stages.34
The appearance of the spectral waveform must
be correlated with the status of the erection (i.e.
flaccid, latent, tumescent, full and rigid).
During the initial latent state, there is a sudden
increase in both systolic and diastolic blood flow
in the cavernosal artery, and a rounded systolic
peak is observed. The flow of blood is unidirec-
tional during systole and diastole, with a pronounced
Fig. 11.5 Spectral Doppler tracing during the initial
forward diastolic component present (Fig. 11.5).
latent phase of the erectile process. Brisk flow is
This spectral waveform reflects the low resistance seen during systole, approaching 30 cm s-1. At this
to flow within the sinusoidal spaces. Minimal state after pharmacological enhancement, diastolic
tumescence normally accompanies this stage. flow is still present in the antegrade direction.
Following that, there is an increase in blood flow
to the corpora cavernosa, causing the intracaver-
nosal pressure to increase. As intracavernosal
pressure rises, a dichrotic notch appears at end
systole and diastolic flow diminishes (Fig. 11.6).
When intracavernosal and diastolic pressures
are the same, diastolic flow ceases and there is
only systolic blood flow. The systolic envelope
narrows and systolic velocity may fluctuate.
Increasing tumescence normally occurs during
this stage.
During full erection, intracavernosal pressure
becomes greater than the arterial pressure during
Fig. 11.6 Spectral Doppler tracing in the early

tumescent phase. Systolic waveform is narrowed into
a sharp spike. Diastolic velocity decreases and a
dicrotic notch are present in early diastole (arrow).
diastole because of the veno-occlusive mechanism.

Flow reversal may be seen during diastole and
the systolic waveform narrows. During this stage,
maximal systolic velocity occurs in many patients
and strong pulsations are normally seen (Fig. 11.7).
In the rigid phase of erection, intracavernosal
pressure may equal or exceed arterial systolic
Fig. 11.4 Spectral Doppler tracing during the flaccid pressure, causing additional narrowing of the
state. Flow velocities are damped and relatively systolic envelope, and normally a decrease in systolic
monophasic. velocity (Fig. 11.8). Both systolic and diastolic 257
11
arteriographic confirmation of arterial integrity.
They found that the changes in diameter were not
significant enough to be indicative or diagnostic
of arterial disease.5,22,26,33,35,36
Other authors use cavernosal arterial diameters
as a parameter in determining arterial integrity
of the penis. Various percentages of increase in
vessel diameter after injection have been reported
as indicators of normal vessel compliance (Table
11.2). It is argued1 that as blood flow in the
cavernosal artery is a function of velocity and
cross-sectional area of the vessels, the ability of
Fig. 11.7 Spectral Doppler tracing approaching rigid the artery to dilate after pharmacological injection
erection. Systolic peak is narrowed, velocity is brisk.
is an important reflection of vessel compliance.
Note the reversal of flow throughout diastole.
It is recognised, however, that the degree of change
in diameter following injection may not correlate
well with other physiological or haemodynamic
parameters. Connolly et al4 contend that because
increased blood flow during erection is accom-
panied by an increase in the diameter of the
cavernosal arteries, patients with arterial disease
will have minimal or no vessel dilatation following
injection because of inadequate vessel compliance
or compromised inflow. Thus, they recommend
use of the diameter and suggest that the measure-
ment be taken approximately 5 min after injection,
since the greatest increase in arterial diameter is
observed during the early latent stage of erection
Fig. 11.8 Spectral Doppler tracing at full rigid when blood inflow is at its maximum.
erection. Note further narrowing of the systolic inflow
peak. Outflow continues through diastole but is at a Arterial variants
lesser velocity (compared to Fig.11.7). Variant penile arterial anatomy can be seen in a
high percentage of patients suspected of having
an arteriogenic cause for their impotence.5,35,41,42
flow may cease completely as the pressure in the Communications among the cavernosal, dorsal,
corporal bodies approaches systolic blood pressure.4
Cavernosal artery diameter

Table 11.2 Criteria indicating normal vessel
Obtaining cavernosal arterial diameters can be compliance
time consuming and the accuracy of the measure-
ment is very dependent on the abilities of the Investigators Percentage increase
in vessel diameter
examiner. Considerable controversy exists as to
Lee et al39 70
the value of obtaining measurements of the caver-
Lue et al17 75
nosal artery diameter following pharmacological
Krysiewicz & Mellinger;22 60–75
induction of erection. Some investigators have James40
reported a poor correlation between the degree Collins & Lewandowski24 60–00
258 of increase in cavernosal artery diameter and
11
Vasculogenic impotence
and spongiosal arteries are often found along tion is considered one of the most important
the shaft of the penis (Fig. 11.9). There may be parameters when evaluating patients whose
duplication of the cavernosal artery or cross- impotence may be due to arterial disease, such
communication between left and right cavernosal as focal stenosis, occlusion, or collateral flow
arteries.43 Rarely, collaterals from the urethral between arteries (Fig. 11.10).1
arteries may also be seen. The incidence of However, what constitutes normal and abnormal
communication between dorsal and cavernosal values for arterial insufficiency varies, Reported
arteries (dorsal–cavernosal perforators) has been peak systolic velocity values indicative of normal
reported in as many as 90% of men.42 ‘Spongiosal– arterial function are summarised in Table 11.3;
cavernosal communications, or ‘shunt’ vessels, which and reported abnormal values indicative of arterial
course from the corpus spongiosum into the corpus disease as the cause of vasculogenic impotence
cavernosum are another common variant.35,43 are summarised in Table 11.4.
Any of these collateral pathways may significantly In addition to peak systolic velocity values in
affect arterial Doppler flow profiles during erec- each cavernosal artery, a comparison of the values
tion.4,43 Although these anatomical variations do can help in the diagnosis of arterial disease.
not necessarily result in arterial insufficiency, Asymmetric velocities are considered abnormal
they may cause inaccurate interpretation if they if the difference between right and left cavernosal
are not appreciated. For example, systolic peak arteries is greater than 10 cm s–1,3, 33 or greater
velocities of the cavernosal artery may be signifi- than 10–15 cm s-1 by Hattery and associates.1
cantly lower in men with a full erectile response Underlying arterial disease should be considered
if arterial collateral communications are present.35,43 in the artery with the lower peak systolic velocity
Careful scanning of the entire penis from the value.44
crura to the glans with colour Doppler is essential Other indicators used to increase the sensi-
in identifying these anomalies. tivity of detecting potential arterial disease include
reversal of blood flow during systole, a penile
blood flow index and blood flow (or cavernosal
VASCULOGENIC IMPOTENCE
artery) acceleration. During rigid erection, reversal
Arterial insufficiency
Measurement of peak systolic velocity in the
cavernosal arteries after pharmacological injec-
Fig. 11.10 Spectral Doppler tracing in a patient with

arterial insufficiency. Despite appropriate pharmaceutical
dose, erectile response was suboptimal. Peak systolic
Fig. 11.9 Longitudinal colour Doppler imaging. velocity in a cavernosal artery only approximates
Note the presence of two dorsal–cavernosal 15 cm s-1. This is well below the accepted range of
perforating arteries (arrows). normal. 259
11
Table 11.3 Criteria for normal peak systolic velocity in cavernosal

arteries following pharmacological induction of erection
Investigators Peak systolic velocity

Lue et al;17 Paushter36 ≥25 cm s-1
Porst45 25–30 cm s-1 or greater
Hattery et al;1 Connolly et al;4 ≥30 cm s-1
Lee et al39
Schwartz et al34 Mean peak systolic velocity of
39 cm s-1*
Herbener et al;3 Benson & Vickers33 ≥40 cm s-1
*Mean peak systolic velocity values are the combined calculated average of the
right and left cavernosal arteries.
Table 11.4 Criteria for abnormal peak systolic velocity in cavernosal arteries following pharmacological
induction of erection
Investigators Peak systolic velocity

Fitzgerald et al;2 Quam et al;5 Lue et al;17 Mean peak systolic velocities <25 cm s-1*
Paushter;36 James40
Hattery et al1 <25 cm s-1 abnormal, 25–30 cm s-1 indeterminate with clinical
correlation recommended
Benson & Vickers33 <30 cm s-1 considered significant, 30–40 cm s-1 regarded
borderline to mild
*Mean peak systolic velocity values are the combined calculated average of the right and left cavernosal arteries.
of diastolic blood flow is considered a normal artery acceleration and blood flow acceleration.
finding; however, reversal of arterial blood flow This index is calculated by dividing the peak
direction during systole is always considered systolic velocity by the systolic rise time (cm s-2).
abnormal and may indicate an underlying vascular The systolic rise time is the time from the start
abnormality.1,3 Systolic flow reversal after pharma- of the systolic curve to its maximum value. Proximal
cological inducement of erection has been observed arterial disease would be expected to damp velocity
in patients with significant proximal penile or waveforms and prolong the systolic rise time.4 In
cavernosal artery stenosis and occlusion with a study comprising 30 patients, Valji and
filling of the distal cavernosal artery secondary Bookstein23 reported the index more predictive
to collateral flow. of arterial insufficiency than cavernosal arterial
Lopez et al46 have described a penile blood flow peak systolic velocity by itself. Oates et al47 found
index, which is calculated by adding the percentage that a systolic rise time of 110 m s-1 or greater
increases in the diameters of the right and left caver- had a predictive value of 92% for arteriogenic
nosal arteries to the peak flow velocities of both impotence. Mellinger et al48 reported that blood
arteries. If the total value is less than 285 vasculogenic flow acceleration seemed to correlate well with
impotence is considered likely to be of an arterial subjective evaluation of erections.
nature; this was 97% sensitive and 77% specific
in diagnosing impotence due to arterial disease. Venous incompetence
Another Doppler index for identifying arterial In some men, venous incompetence, or failure
260 disease has been referred to both as cavernosal of the veno-occlusive mechanism, may be the
11
Vasculogenic impotence
primary cause of vasculogenic impotence.2Patients
with normal arterial inflow parameters (e.g. peak
systolic velocity >25 cm s-1) but weak erections
will very likely have some degree of venous leakage.4
Because primary venous leakage is a potentially
treatable cause of erectile dysfunction, Doppler
examination of the penile venous system may be
helpful in identifying these patients who may
benefit from additional more invasive studies.1,49
If surgical or endovascular therapy is considered
for the patient, then cavernosography is generally
still required.50,51
In those patients with arterial inflow issues Fig. 11.11 Spectral Doppler tracing of venous
(peak systolic velocity < 25 cm s-1), however, the insufficiency. Relatively high velocity persistent
veno-occlusive mechanism will not be fully engaged, forward flow is seen during diastole (approximating
persistent end-diastolic flow can be expected 6 cm/sec). This despite appropriate arterial inflow
and venous leak cannot be assessed. velocities.
A correlation has been shown between end-
diastolic blood flow velocity within the cavernosal
arteries and the presence of venous leakage.With criteria exist as to what constitutes abnormal
a normal erectile response, there should be minimal, diastolic velocity (Table 11.5). In association
if any, flow detected with the cavernosal arteries with a normal arterial inflow, most authors now
during the diastolic phase 15–20 min after injec- consider an end diastolic velocity of >5 cm s-1 as
tion. As previously noted, there should be a the velocity above which a venous leak is
decrease and eventual absence or reversal of present.5,52
diastolic flow in the normal spectral Doppler A potential for a false-positive diagnosis of
waveform during rigid erection. If there is veno- venous leak does exist, especially in young men.
occlusive dysfunction (venous leak), then this The anxiety during a penile ultrasound exami-
decrease or reversal of diastolic flow will not nation increases sympathetic drive which results
occur.5, 40, 49 A persistently elevated diastolic flow in inadequate relaxation of sinusoidal smooth
in the cavernosal arteries is highly indicative of muscle and consequent failure of veno-occlusion.
venous leakage out of the corporal tissue even Additional intracavernosal administration of an
after maximum peak systolic velocity has been alpha-adrenergic antagonist such as phentolamine
attained (Fig. 11.11). However, just as there are 2 mg, should be considered. Phentolamine blocks
differences of opinion regarding normal and the increased sympathetic drive and helps avoid
abnormal peak systolic velocity values, various the false-positive diagnosis of venous leak.53
Table 11.5 Criteria for abnormal mean arterial end-diastolic velocity
Investigators End-diastolic velocity

Herbener et al;3 James40 >3 cm s-1
Quam et al;5Benson & Vickers;33 Paushter;36 >5 cm s-1
Fitzgerald et al37
Hattery et al1 >5–8 cm s-1
Montorsi et al38 >10 cm s-1
261
11
flow priapism and colour Doppler sonography is
TRAUMA
very effective in its identification (Fig. 11.12). It
Injury to the penis may be the result of blunt or shows a characteristic arterial colour blush
penetrating trauma; or from acute bending of the consistent with extravasation of blood from the
erect shaft. Depending on severity of the insult lacerated artery. Spectral Doppler displays
subcutaneous or intracorporeal haematoma, turbulent high-velocity flow.3,10,56,57 In long-
tunical disruption or urethral tear may occur. standing priapism, this area may appear more
The main role of sonography is to exclude albug- circumscribed, mimicking a pseudoaneurysm.
ineal tear because extratunical and cavernous Angiographic embolisation of the lacerated artery
haematoma can be treated conservatively. is currently considered the treatment of choice.
Surgery is required, however, when rupture of Colour Doppler US allows confirmation of
the tunica albuginea cannot be excluded, espe- successful embolisation by demonstrating
cially if the patient reports immediate detumes- disappearance or size reduction of the fistula. If
cence associated with a popping sound. Surgical unsuccessful, persistent patency of feeding
repair of an albugineal tear reduces the risk of vessels would be seen.58
post-traumatic curvature, and lowers the inci-
dence of erectile dysfunction. Colour Doppler
PEYRONIE’S DISEASE
sonography can be used to help localise small
albugineal tears. By squeezing the penile shaft, a Peyronie’s disease (induratio penis plastica) is an
flush of blood can be forced from the cavernosal idiopathic disorder of the connective tissue in
bodies through the tear.54 which fibrous plaques form in the tunica albug-
Following pelvic trauma, Doppler ultrasound inea with induration of the corpora cavernosa of
is useful for identifying the presence of injury to the penis or a fibrous cavernositis. Its aetiology
the penile vasculature.3,20 Most non-organic causes is unknown and the incidence is stated to be 1%.
of erectile dysfunction in young men are secondary The disease causes penile deviation and affected
to arterial injury with pelvic fracture. Penile arterial patients typically complain of pain during
inflow may be compromised by development of erection.3,11 Ultrasound imaging can estimate
a post-traumatic stenosis. Doppler ultrasound the extent and depth of the plaques, which may
can be used to look for a tardus parvus wave- be hypo- or hyperechoic and, if long standing,
form of the affected artery. For patients who often contain calcification.
experience post-traumatic impotence, Broderick Doppler examination of the penile vasculature
et al55 found that a systolic velocity less than is of value to determine if any vascular abnor-
25 cm s-1 and asymmetric velocities >10 cm s-1 malities exist because of the fibrous plaques.11,12,59
were helpful parameters in establishing a Doppler ultrasound is helpful in differentiating
compromise of penile arterial inflow. In some between veno-occlusive dysfunction (considered
centres, vascular microsurgery is being performed the primary vascular cause of impotence
on the internal pudendal artery to bypass a focal associated with Peyronie’s disease) and arterial
stenosis. insufficiency.13 By doing so, Doppler helps
With severe trauma, the cavernosal artery may establish the correct treatment option for the
rupture. This results in unrestricted blood flow patient. If Doppler reveals appropriate arterial
into the cavernosal space, creating an arteriosinu- inflow and an appropriate response to injection,
soidal, arteriolacunar, or arteriovenous fistula. then plaque removal with grafting of bovine peri-
Since venous outflow is maintained from the cardium is preferred. However, if the Doppler
cavernosal space, the patient develops only a partial shows abnormal flow profiles or an inadequate
erection that is not acutely painful. Persistent response to injection, in addition to the disfigure-
venous outflow prevents complete erection, stasis ment, then implantation of prosthesis should be
262 and hypoxia. This condition is known as high- considered.60
11
Other imaging tests

a
Fig. 11.12 This patient sustained a penile injury while

c riding a bull 15 days prior to admission. The patient
presented with a persistent painless erection.
(a) Longitudinal colour Doppler image of the mid-
penis. An area of colour flow aliasing within the left
corpse cavernosa is the focus of the arterial lacunar
fistula (arrow). (b) Corresponding spectral Doppler
tracings. There is marked turbulent flow at the site as
would be expected with an arterial–lacunar fistula.
(c) Pre-embolisation arteriography. An early blush
(arrows) corresponds to the site of the fistula.
This was subsequently successfully embolised.
This case is courtesy of Dr Vikram Dogra, Case
Western Reserve University.
the use of the penile/brachial pressure index and

PENILE TUMOURS
pudendal arteriography at many institutions for
Penile masses are rare. Possibilities include the initial evaluation of patients with suspected
haematoma, cavernosal herniation, Peyronie’s vasculogenic impotence.4 The penile brachial
plaques, penile carcinoma, metastases, foreign index is now basically a test of historical interest
bodies or the penile prosthesis. Doppler may show only. Although arteriography is considered the
some interesting flow patterns but rarely adds ‘gold standard’ for assessing the arteries of the penis,
value to the evaluation. unlike Doppler ultrasound, it provides primarily
anatomical rather than functional information.
If Doppler examination indicates the presence
OTHER IMAGING TESTS
of a surgically correctable lesion, though, many
Doppler ultrasound combined with pharma- investigators still believe a confirmatory arte-
cological inducement of erection has replaced riogram (computed tomographic angiography 263
11
or magnetic resonance angiography) should be cavernosometry are used to determine the
performed before surgery is undertaken. presence of veno-occlusive dysfunction and to
Colour Doppler ultrasound has also proven quantify the degree of venous leakage.
helpful in evaluating patients with suspected venous
incompetence. However, for patients in whom
ACKNOWLEDGEMENTS
venous surgery is being considered to correct
their condition, dynamic infusion cavernosography Reginald Bruskewitz, MD; Mike Ledwidge,
is necessary since it provides a more complete RDMS, RVT; Carrie Poole; Carol Mitchel
preoperative picture of the penile vasculature.1–3 RDMS and Joan Palmer
At many institutions, both cavernosography and
REFERENCES
1. Hattery RR, King BF, Lewis RW, et al. Vasculogenic 13. Levine LA, Coogan CL. Penile vascular assessment
impotence: duplex and colour Doppler imaging. using colour duplex sonography in men with
Radiol Clin North Am 1991; 29:629-645. Peyronie’s disease. J Urol 1996; 155:1270–1273.
2. Fitzgerald SW, Erickson SJ, Foley WD, et al. 14. Gray’s Anatomy, Online. Available:
Colour Doppler sonography in the evaluation of http://education.yahoo.com/reference/gray/subjects/
erectile dysfunction. Radiographics 1992; 12:3–17. subject?id=262 March 2005.
3. Herbener TE, Seftel AD, Nehro A, et al. Penile 15. Aboseif SR, Lue TF. Hemodynamics of penile
ultrasound. Semin Urol 1994; 12:320–332. erection. Urol Clin North Am 1988; 15:1–7.
4. Connolly JA, Borirakchanyavat S, Lue TF. Ultrasound 16. Lue TF. Physiology of penile erection and
evaluation of the penis for assessment of pathophysiology of impotence. In: Walsh PC,
impotence. J Clin Ultrasound 1996; 24:481–486. Retik AB, Stamey TA, et al, eds. Campbell’s
5. Quam JP, King BF, James EM, et al. Duplex and urology, 6th edn. Philadelphia: WB Saunders;
colour Doppler sonographic evaluation of vasculogenic 1992:707–728.
impotence. AJR Am J Roentgenol1989; 17. Lue TF, Hricak H, Marich KW, et al. Vasculogenic
153:1141–1147. impotence evaluated by high-resolution
6. Corona G, Mannucci E, Mansani R, et al. Aging and ultrasonography and pulsed Doppler spectrum
pathogenesis of erectile dysfunction. Int J Impot analysis. Radiology 1985; 155:777–781.
Res 2004; 16:395–402. 18. Mueller SC, Lue TF. Evaluation of vasculogenic
7. Krane RJ, Goldstein I, Saenz de Tejada I. Medical impotence. Urol Clin North Am 1988; 15:65–76.
progress: impotence. N Engl J Med 1989; 19. Gall H, Bahren W, Scherb W, et al. Diagnostic
321:1648–1659. accuracy of Doppler ultrasound technique of the
8. Weiske WH. Diagnosis of erectile dysfunction – penile arteries in correlation to selective
what is still needed today? Urologe A 2003; arteriography. Cardiovasc Intervent Radiol 1988;
42:1317–1321. 11:225–229.
9. Armenakas NA, McAninch JW, Lue TF, et al. 20. Lue T, Tanagho E. Physiology of erection and
Posttraumatic impotence: magnetic resonance pharmacological management of impotence.
imaging and duplex ultrasound in diagnosis and J Urol 1987; 137:829–836.
management. J Urol 1993; 149:1272–1275. 21. Desai KM, Gingell JC, Skidmore R, et al.
10. Harding JR, Hollander JB, Bendick PJ. Chronic Application of computerized penile arterial
priapism secondary to a traumatic arteriovenous waveform analysis in the diagnosis of arteriogenic
fistula of the corpus cavernosum. J Urol 1993; impotence: an initial study in potent and impotent
150:1504–1506. men. Br J Urol 1987; 60:450–466.
11. Lopez JA, Jarrow JP. Duplex ultrasound findings in 22. Krysiewicz S, Mellinger BC. The role of imaging in
men with Peyronie’s disease. Urol Radiol 1991; the diagnostic evaluation of impotence. AJR Am J
12:199–202. Roentgenol 1989; 153:1133–1139.
12. Amin Z, Patel U, Friedman EF, et al. Colour Doppler 23. Valji K, Bookstein JJ. Diagnosis of arteriogenic
and duplex ultrasound assessment of Peyronie’s impotence: efficacy of duplex ultrasonography as a
disease in impotent men. Br J Radiol 1993; screening tool. AJR Am J Roentgenol 1993;
264 66:398–402. 160:65–69.
24. Collins JP, Lewandowski BJ. Experience with
39. Lee B, Suresh CS, Randrup ER, et al.

11
References
intracorporal injection of papaverine and duplex Standardization of penile blood flow parameters in
ultrasound scanning for assessment of arteriogenic normal men using intracavernous prostaglandin E1
impotence. Br J Urol 1987; 59: 84–88. and visual sexual stimulation. J Urol 1993;
25. Bookstein JJ, Valji K, Parsons L, et al. 149:49–52.
Pharmacoarteriography in the evaluation of 40. James EM. Penile ultrasound. Syllabus special
impotence. J Urol 1987; 137:333–337. course: Ultrasound 1991. Presented at the 77th
26. Rajfer J, Canan V, Dorey FJ, et al. Correlation Scientific Assembly and Annual Meeting of the
between penile angiography and duplex scanning Radiologic Society of North America. Chicago:
of cavernous arteries in impotent men. J Urol 1990; Radiologic Society of North America;
143:1128–1130. 1991:259–266.
27. Rosen MP, Schwartz AN, Levine FJ, et al. Radiologic 41. Breza J, Aboseif SR, Orvis BR, et al. Detailed
assessment of impotence angiography, sonography, anatomy of penile neurovascular structures: surgical
cavernosography and cinetrigraphy. AJR Am J significance. J Urol 1989; 141:437–443.
Roentgenol 1991; 157:923–931. 42. Bahren W, Gall H, Scherb W, et al. Arterial anatomy
28. Shabsigh R, Fishman IJ, Scott FB. Evaluation of and arteriographic diagnosis of arteriogenic
erectile impotence. Urology 1988; 32:83–90. impotence. Cardiovasc Intervent Radiol 1988;
29. King BF, Hattery RR, James EM, et al. Duplex 11:195–210.
sonography in the evaluation of impotence: current 43. Mancini M, Bartolini M, Maggi M, et al.
techniques. Semin Intervent Radiol 1990; The presence of arterial anatomical variations can
7:215–221. affect the results of duplex sonographic evaluation
30. Meuelman EJ, Bemelmans BH, Doesburg WH, of penile vessels in impotent patients. J Urol 1996;
et al. Penile pharmacological duplex ultrasonography: 155:1919–1923.
a dose-effect study comparing papaverine, 44. Hattery RR, King BF, James EM, et al. Vasculogenic
papaverine/phentolamine, and prostaglandin E1. impotence: duplex and colour Doppler imaging.
J Urol 1992; 148:63–66. AJR Am J Roentgenol 1991; 156:189–195.
31. Gontero P, Sriprasad S, Wilkins CJ, et al. 45. Porst H. Duplex ultrasound of the penis: value of a
Phentolamine re-dosing during penile dynamic new diagnostic procedure based on over 1,000
colour Doppler ultrasound: a practical method to patients [German]. Urologe - Ausgabe A 1993;
abolish a false diagnosis of venous leakage in 32:242–249.
patients with erectile dysfunction. Br J Radiol 2004; 46. Lopez JA, Espeland MA, Jarrow JP. Interpretation
77:922–926. and quantification of penile blood flow studies
32. Donatucci CF, Lue TF. The combined intracavernous using duplex ultrasonography. J Urol 1991;
injection and stimulation test: diagnostic accuracy. 146:1271–1275.
J Urol 1992; 148:61–62. 47. Oates CP, Pickard RS, Powell PH, et al. The use of
33. Benson CB, Vickers MA. Sexual impotence caused duplex ultrasound in the assessment of arterial
by vascular disease: diagnosis with duplex sonography. supply to the penis in vasculogenic impotence.
AJR Am J Roentgenol 1989; 153:1149–1153. J Urol 1995; 153:354–357.
34. Schwartz AN, Wang KY, Mack LA, et al. Evaluation 48. Mellinger BC, Fried JJ, Vaughen ED. Papaverine-
of normal erectile function with colour flow Doppler induced penile blood flow acceleration in impotent
sonography. AJR Am J Roentgenol 1989; men measured by duplex scanning. J Urol 1990;
153:1155–1160. 144:897–899.
35. Jarrow JP, Pugh VW, Routh WD, et al. Comparison 49. Vickers MA, Benson CB, Richie JP. High resolution
of penile duplex ultrasonography to pudendal ultrasonography and pulsed wave Doppler for
arteriography: variant penile arterial anatomy affects detection of coporovenous incompetence in
interpretation of duplex ultrasonography. Invest erectile dysfunction. J Urol 1990; 143:1125–1127.
Radiol 1993; 28:806–810. 50. Frust G, Muller-Mattheis V, Cohnen M, et al. Venous
36. Paushter DM. Role of duplex sonography in the incompetence in erectile dysfunction: evaluation
evaluation of sexual impotence. AJR Am J with colour-coded duplex sonography and
Roentgenol 1989; 153:1161–1163. cavernosometery/-graphy. Eur Radiol 1999; 9:35–41.
37. Fitzgerald SW, Erickson SJ, Foley WD, et al. 51. Fowlis GA, Sidhu PS, Jager HR, et al. Preliminary
Colour Doppler sonography in the evaluation of report: combined surgical and radiological penile
erectile dysfunction: patterns of temporal response vein occlusion for the management of impotence
to papaverine. AJR Am J Roentgenol 1991; caused by venous-sinusoidal incompetence. Br J
157:331–335. Urol 1994; 74:492–496.
38. Montorsi F, Bergamaschi F, Guazzoni G, et al. 52. Bassiouny HS, Levine LA. Penile duplex sonography
Doppler colour echography in the diagnosis of in the diagnosis of venogenic impotence. J Vasc
impotence. Minerva Chir 1993; 48:99–102. Surg 1991; 13:75–82.
265
11
53. Aversa A, Rocchietti-March M, Caprio D, et al. 57. Sadeghi-Nejad H, Dogra V, Seftel AD, et al.
Anxiety-induced failure in erectile response to Priapism. Radiol Clin North Am 2004; 42:427–443.
intracorporeal prostaglandin-E1 in non-organic male 58. Bertolotto M, Quaia E, Mucelli FP, et al. Colour
impotence: a new diagnostic approach. Int J Androl Doppler imaging of posttraumatic priapism before
1996; 19:307–313. and after selective embolization. Radiographics
54. Bertolotto M, Mucelli RP. Nonpenetrating penile 2003; 23:495–503.
traumas: sonographic and Doppler features. 59. Bertolotto M, de Stefani S, Martinoli C, et al. Colour
AJR Am J Roentgenol 2004; 183:1085–1089. Doppler appearance of penile cavernosal-spongiosal
55. Broderick GA, McGahan JP, White RD, et al. communications in patients with severe Peyronie’s
Colour Doppler US: assessment of post-traumatic disease. Eur Radiol 2002; 12:2525–2531.
impotence. Radiology 1990; 177(suppl):130–134. 60. Fornara P, Gerbershagen HP. Ultrasound in patients
56. Hakim LS, Kulaksizoglu H, Mulligan R, et al. Evolved affected with Peyronie’s disease. World J Urol
concepts in the diagnosis and treatment of arterial 2004; 22(5):365–367.
high flow priapism. J Urol 1996; 155:541–548.
266
12
scrotum
Myron A. Pozniak
2–4 cm in width and thickness with a volume of

4 cm3.7 Testicular size varies with age and stage
INTRODUCTION
of sexual development. At birth, the testis measures
Scrotal sonography was first introduced in the approximately 1.5 cm in length and 1 cm in
mid-1970s. Today, ultrasound is the preferred width. Before puberty, testicular volume is about
scrotal imaging technique. Improvements in image 1–2 cm3.8 The surface of the testicle is covered
resolution with higher frequency transducers, by the tunica albuginea, a thin, dense, inelastic
increased Doppler sensitivity and refinements in fibrous capsule. Just within the tunica albuginea
power Doppler technology have greatly enhanced is the tunica vasculosa, through which the branches
perception of testicular anatomy and pathology. of the testicular artery course before entering into
The haemodynamic information acquired with the gland (Fig. 12.1).8,9 Numerous thin septations
Doppler adds to the imaging findings and (septula) arise from the tunica albuginea and
frequently reinforces and occasionally clinches create 250–400 cone-shaped lobules containing
the diagnosis.1–6 In this era, a thorough ultra- the seminiferous tubules (Fig. 12.2).10–12 These
sound evaluation of the scrotum must include tortuous tubules course towards the mediastinum
Doppler imaging of the testes and epididymides. of the testis and progressively merge to form larger
ducts known as tubuli recti (Fig. 12.3).These, in
turn, join with each other to form a network of
INDICATIONS
epithelium lined spaces embedded in the
The two clinical conditions that most commonly fibrostroma of the mediastinum called the rete
warrant an ultrasound examination with Doppler testis. These continue as 10–15 efferent ductules
are evaluation of acute scrotal pain and a perceived which pass into the head of the epididymis.
scrotal mass. Differentiating an acute inflamma- The epididymis is a comma shaped structure
tory process from testicular torsion is now the that runs along the posterolateral aspect of the testis
domain of colour Doppler. Differentiating a mass (Fig. 12.4). Its components are the head (globus
as intratesticular versus extratesticular can be major), body and tail, (globus minor). The head
accomplished by imaging alone, but colour Doppler is located next to the upper pole of the testis and
adds valuable flow information and helps to receives the efferent ductules.The ductules even-
further characterise the abnormality. Other tually converge through the body and tail and
common indications for use of scrotal Doppler form the vas deferens which continues on in the
include evaluation of trauma and infertility. spermatic cord. Along with the vas, the cord also
contains the testicular artery, cremasteric artery,
differential artery, the pampiniform plexus of veins,
TESTICULAR ANATOMY
the genitofemoral nerve and lymphatic channels.
The normal adult testis is an egg-shaped gland The testis and epididymis are enveloped by
which is approximately 3–5 cm in length and the tunica vaginalis, a fascial structure composed 267
12
Doppler imaging of the scrotum
Spermatic cord
Fig. 12.1 The testicular artery
supplies flow to the epididymis and
the testicle. As the testicular artery
Ductus courses through the spermatic
Testicular artery
deferens cord, it is surrounded by the
pampiniform plexus of veins.
The capsular artery, a branch of the
Head of testicular artery, courses just
epididymis
Deferential beneath the tunica albuginea.
artery The centripetal arteries are
branches of the capsular artery
Pampiniform
plexus and course between the septa
supplying the testicular
parenchyma.
Tunica
albuginea
Mediastinum
Centripetal
artery
Tunica
albuginea
Tail of epididymis
Capsular artery
of an outer parietal layer and an inner visceral

H layer which surrounds the entire gland, except
along the posterior aspect where the vessels and
A nerves enter. The extent to which the tunica
vaginalis envelopes the testis directly correlates
with the risk for developing testicular torsion.
The potential space between the parietal and
visceral layers normally contains a small amount
of lubricating fluid, but a larger fluid collection
accumulating here is know as a hydrocele.
Testicular arterial anatomy

The right and left testicular arteries originate
from the aorta just below the renal arteries.They
Fig. 12.2 Transverse view of a normal testicle course through the deep inguinal ring to enter
surrounded by a hydrocele (H). The testicle is of the spermatic cord, accompanied by the cremas-
relatively uniform echogenicity. With high resolution
teric and deferential arteries, which supply the
equipment, the testicular septa can be perceived as
soft tissues of the scrotum, epididymis and vas
fine lines dividing the parenchyma into the individual
deferens. The testicular artery penetrates the
lobules. Note the echogenic tunica albuginea
268 surrounding the testicle (A). tunica albuginea along the posterior aspect of
a
12
Technique
M
Fig. 12.4 Longitudinal view of a normal right

Fig. 12.3 Transverse view of the left testicle (a).
scrotum. The slightly hypoechoic epididymis wraps
Specimen in the region of the mediastinum testis (b).
as a comma shape around the testis. The larger more
Mediastinum testis is an echogenic structure. It is the
prominent area is known as the globus major or the
point of confluence of the efferent ductules and the
head of the epididymis.
septa. Note the triangular echogenic appearance on
the ultrasound (M) and how well it correlates with the
dense white fibrous stroma on the specimen.
Testicular venous anatomy
Testicular venous outflow courses mostly through
the testis and gives off capsular branches which the mediastinum testis, into the spermatic cord,
course through the tunica vasculosa.These capsular and eventually up through the inguinal canal.
branches then give rise to the centripetal arteries As the veins exit the scrotum through the
which carry blood from the capsular surface, spermatic cord, they form a web-like network
centrally towards the mediastinum along the septula that surrounds the testicular artery, known as
(Figs 12.5 and 12.6). Some of the branches of the pampiniform plexus (Fig. 12.8). This plexus
the centripetal arteries course backward towards is thought to function as a heat exchange mech-
the capsular surface; these are known as anism, pulling warmth away from the testicular
recurrent rami. In about 50% of testes, a small arterial inflow, thereby helping to maintain
branch can be seen passing directly from the spermatogenesis at a lower, more optimal
testicular artery into the parenchyma, known as temperature.
the transtesticular artery (Fig. 12.7). Small On the left side, the testicular vein usually
anastomoses do exist between the testicular drains into the left renal vein; on the right side,
artery, cremasteric, and differential arteries. drainage is directly into the inferior vena cava
Branches of the pudendal artery may also supply just below the right renal vein. The testicular
the scrotal wall. veins normally have valves that prevent retro- 269
12
Fig. 12.7 Longitudinal colour Doppler image of a

testicle. A transtesticular artery (arrow) can be seen
coursing obliquely through the testicle along the
mediastinum. It is a branch of the testicular artery,
C but is seen in less than half of normal cases.
Fig. 12.5 Microangiogram of the testicle. A prominent

capsular branch (C) courses from the testicular artery
t
under the capsular surface. Centripetal arterial branches
emanate from it and pass into the lobules of the pp
testicle. With very sensitive Doppler settings, a series
of centripetal arteries is seen across this normal testicle.
Case courtesy of Dr. Thomas Winter.
Fig. 12.8 The pampiniform plexus of veins (PP) is a

web-like collection of vessels surround the testicular
artery (T) as it passes through the spermatic cord.
TECHNIQUE
Prior to any ultrasound examination, the testes
should be examined with a gloved hand, espe-
cially if the sonographic study is being conducted
to evaluate a palpable mass. The examination is
Fig. 12.6 Longitudinal colour Doppler image of a performed with the patient in the supine
testicle. With very sensitive Doppler settings, position. A towel is placed under the scrotum for
the network of centripetal arteries is seen fanning support. If the testis is tender, the patient may
across this normal testicle. be asked to hold it in a position which would
facilitate the exam. This is particularly useful for
grade flow of venous blood to the scrotum, but the evaluation of a small mass. The patient
if they are absent or become incompetent; this should be asked to hold it between thumb and
270 predisposes to development of a varicocele.13,14 forefinger, and the ultrasound transducer is then
12
Technique
gently placed upon it. Some men may have a
vigorous cremasteric response during the exami-
nation resulting in the testis being drawn upward
and puckering the scrotal wall.To avoid shadowing
from trapped air, copious amounts of gel need to A A
be worked into the scrotal skin folds (Fig. 12.9).
Imaging with the patient standing upright, or
while performing a Valsalva manoeuvre, is useful
for evaluating the testicular venous system, in
particular for determining valvular competence
in patients with suspect varicocele, or for
improving detection of an inguinal hernia.
A high-frequency (10 MHz or greater) linear-
array transducer is used for both grey-scale and
Doppler imaging, with direct contact scanning
Fig. 12.9 Longitudinal image of the testicle. Note the
on the scrotal skin. Examination is performed in
layer of gel between transducer and scrotal wall.
both the longitudinal and transverse planes for Two air bubbles are barely perceived, but are enough
each testis to allow assessment of any differences to cause significant artifact within the testicle (A)
in size and echogenicity between the two sides. which mimics a hypoechoic lesion.
The split screen mode is useful to allow a direct
comparison. Oblique imaging of the epididymis
and spermatic cord should also be performed. feres with the examination, colour artefact can
Any extratesticular masses or fluid collections be decreased by properly adjusting gain and scale
should be noted as well. settings. Temporal resolution can be improved
by minimising the overall image size by using a
Doppler technique write zoom or decreasing depth and limiting the
Ultrasound examination of the scrotum is not size of the colour box. Use of appropriate tech-
complete without applying colour Doppler. The nical parameters should assure demonstration of
procedure is a mandatory part of the imaging intratesticular vessels in all normal cases.9 The
evaluation to confirm the presence (or absence) use of power Doppler can be helpful in very low-
of uniform, symmetric vascular perfusion of the flow conditions, such as in paediatric patients,
testes and epididymides (Fig. 12.10).The settings and with its evolution scrotal ultrasound has
for colour Doppler scanning must be optimised essentially replaced nuclear scintigraphy for the
for low-volume and low-velocity flow.15 If colour evaluation of torsion.11,12,16 Spectral Doppler
noise is excessive at low-flow settings and inter- can assess arterial and venous waveforms and
Fig. 12.10 Longitudinal Doppler

ultrasound image of both testicles
using a split screen format.
Note the symmetry of echotexture
and the uniformity and symmetry
of colour flow.
271
12
quantify velocities15 (Fig. 12.11), but its appli- the testes.The size and echogenicity of testes and
cation in the scrotum is limited, except for a few epididymides should be bilaterally symmetric.
conditions such as partial torsion. Then spectral Colour Doppler should reveal bilaterally
Doppler can best identify the high resistance to symmetric and relatively uniform flow through
arterial inflow caused by venous outflow both testes and epididymides (Fig. 12.10). A fan-
compromise. like array of the centripetal arteries should be
Doppler sensitivity varies greatly between ultra- present through the testicles (Fig. 12.6). Venous
sound systems and software levels. Therefore, flow in the epididymides may be seen to fluctuate
the examiner must be familiar with normal flow with respiration.
perception on their equipment. A good ‘rule of Spectral Doppler tracings of testicular arterial
thumb’ is to examine the contralateral side (provided inflow demonstrate relatively low resistance (Fig.
it is normal) to establish a colour flow baseline 12.11); this is in contrast to the cremasteric and
which can then be used as the standard by which deferential arteries which have relatively high
to judge the abnormal testis or epididymis.When resistance to flow. The normal testicular artery
comparing flow between sides, be sure to set resistive indices in adults range from 0.46 to 0.78,
imaging parameters to the non-affected side; then, with a mean of 0.64.9 Similar findings are reported
without changing any settings, image the affected in the intratesticular arteries of postpubescent
side. Some advanced ultrasound machines currently boys, with resistive indices ranging from 0.48 to
adjust imaging parameters automatically as 0.75 (mean, 0.62).15 In prepubertal boys, however,
depth and position of colour box changes with resistance is higher to the point that diastolic
scanning. If possible, override this software feature arterial flow may not be detectable.17 Suprates-
to avoid misperception of asymmetric colour ticular arteries to the vas deferens or cremaster
flow. muscle, on the other hand, have higher impedance
with low-diastolic flow and resistive indices ranging
Normal ultrasound findings from 0.63 and 1.0, with a mean of 0.84.1,9,15
A normal ultrasound examination of the scrotum Pulsed Doppler is relatively insensitive in detect-
reveals uniform, homogeneous echogenicity ing arterial flow in prepubescent patients.15,17,18
throughout both testes.The epididymis is usually In contrast, power Doppler has been shown to
isoechoic or slightly hyperechoic compared with reveal arterial flow in 92% of testes in pre-
pubescent patients, while colour Doppler demon-
strates flow in 83% of cases.4 In postpubescent
patients, both Doppler imaging techniques
demonstrated flow in 100% of cases.
ACUTE SCROTAL PAIN

Inflammatory disease
Acute epididymo-orchitis is the most common
cause of scrotal pain in men over the age of 20
years, accounting for up to 80% of cases of acute
scrotal pain, but it is frequently clinically
indistinguishable from spermatic cord torsion.
Patients usually present with an acutely painful,
tender, swollen scrotum, with associated erythema,
urinary tract symptoms, fever and leukocytosis.
Fig. 12.11 Spectral Doppler waveform of a normal Sometimes, however, the signs and symptoms may
272 testicle. The resistive index is a normal (approx. 70%). be less distinct, making clinical differentiation
Fig. 12.12 Longitudinal grey-scale

12
Acute scrotal pain

and colour Doppler image of the
right testicle. The patient presented
with severe pain. Both the testis
and epididymis are oedematous,
swollen and hypoechoic.
Colour Doppler shows exuberant
flow through both structures.
The diagnosis is epididymo-orchitis.
between infection and torsion extremely difficult. tified by colour Doppler – an asymmetric appear-
The process typically first manifests in the ance with more robust flow (an increased number
epididymis and then ascends to affect the testicle, and prominence of discernible vessels) in asso-
but isolated epididymitis, orchitis, or even focal ciation with an enlarged, painful, hypoechoic epi-
orchitis can be encountered. didymis and/or testis (Fig. 12.12).1,2,5,21,23 A study
The cause of the infection varies with age.19 by Ralls et al6 demonstrated 91% sensitivity and
(In adult patients less than 35 years of age, 100% specificity for the diagnosis of scrotal inflam-
Chlamydia trachomatis and Neisseria gonorrhoeae matory disease by colour Doppler sonography.
(sexually transmitted organisms) are the most Other authors have also reported a sensitivity close
common cause. In prepubertal boys and in men to 100%.9,23 Inflammatory hyperaemia typically
over 35 years of age, the disease is most frequently shows a low-resistance flow pattern on spectral
caused by Escherichia coli and Proteus mirabilis. Doppler;1 although spectral analysis is occasionally
Cytomegalovirus is the most common agent in necessary to confirm the diagnosis (Fig. 12.13).24
the immunocompromised patient. In most normal In cases of severe epididymitis, periepididymal
paediatric patients, a bacterial pathogen is not swelling may obstruct testicular venous outflow,
isolated and the inflammation is presumed to be leading to testicular ischaemia or infarction. An
viral in nature, but paediatric patients who have
an underlying urogenital congenital anomaly are
prone to infection from gram negative bacteria.20
The sonographic appearance of epididymo-
orchitis varies depending on the stage of the process.
The sensitivity of grey-scale sonography for
detecting epididymo-orchitis has been reported
to be about 80%.21 In the early, acute stage, the
epididymis and/or testicle will be enlarged and
hypoechoic. With the onset of tissue breakdown
and haemorrhage, the echogenicity begins to
increase. 5,12,21 There may be reactive thickening
of the scrotal wall. A hydrocele may be present
and it may contain debris. If allowed to progress,
microabscesses may develop and the appearance
becomes more complex and variable. Scarring Fig. 12.13 Spectral Doppler tracing of an area of
and necrosis associated with chronic orchitis typi- testicular inflammation. Flow during diastole is
cally results in a small hyperechoic testis.22 relatively brisk resulting in a resistive index of <50%.
The diagnosis of inflammation can be made This appearance is consistent with inflammatory
with greater confidence when hyperaemia is iden- hyperaemia. 273
12
enlarged, heterogeneous testicle with reduced or
absent colour flow and a contiguous abnormal
epididymis will be seen on grey-scale images
(Fig. 12.14).15 Hyperaemia of the epididymis helps
to differentiate testicular ischaemia following A
inflammation from that caused by torsion. A

high-resistance waveform, along with decreased
or reversed diastolic flow, may be seen on the
spectral Doppler tracing and suggests venous
infarction.25
If left untreated, the process may progress to
abscess formation (Fig. 12.15), which usually
Fig. 12.15 Transverse view of the scrotum. An inflamed

hyperaemic testicle is located centrally within a
complex septated fluid collection. The brisk flow in
E the wall surrounding this collection helps identify it as
T an abscess (A).
manifests as an enlarged testicle with a complex,

septated collection of fluid debris and tissue of
mixed echogenicity. It may be difficult to distinguish
from a testicular neoplasm, as both can present
as complex cystic/solid masses. Older abscesses
may have radiating echogenic septa separating
hypoechoic spaces. Increased blood flow around
the abscess cavity and no internal flow is present
on colour Doppler.1,21 Surgical exploration may
be necessary to rule out the presence of a tumour
and débride the abscess.
Occasionally, testicular inflammation may be
focal, even rounded, with areas of decreased
echogenicity and swelling. Hyperaemia concen-
trated in the abnormal, tender area may suggest
inflammation over neoplasm, but there is overlap
in this appearance with neoplasm, especially
lymphoma (Fig. 12.16).
Torsion
Torsion is divided into two types – extravaginal
Fig. 12.14 Longitudinal view of the left testicle with
which occurs exclusively in newborns and
and without colour Doppler. The epididymis (E) is
enlarged, but the testicle (T) remains of normal size intravaginal. Extravaginal torsion occurs outside
and echotexture. With colour Doppler, there is brisk the tunica vaginalis when the testes are not yet fixed
flow in the epididymis whereas the testicle shows and are free to rotate.26 The testis is typically
appropriate if not slightly reduced flow. These findings necrotic at birth. Ultrasound reveals an enlarged
274 are consistent with epididymitis. heterogeneous testis, a reactive hydrocele, skin
12
Acute scrotal pain

and signs include sudden pain in the scrotum,
lower abdomen or inguinal area (frequently accom-
panied by nausea, vomiting and low-grade fever),
a tender testicle with a transverse orientation,
and a swollen, erythematous hemiscrotum.15
The false-positive rate of nearly 50% for clinical
diagnosis of testicular torsion often results in
unnecessary surgical exploration.28
Grey-scale sonography by itself has a low
sensitivity and specificity when evaluating patients
for suspected torsion. Findings will depend
on the length of time the torsion has been
present.1,2,5,23,29,30 During the first few hours,
Fig. 12.16 Longitudinal colour Doppler image of a testicular appearance is normal, but after about
testicle. This patient complained of severe scrotal 4–6 h, as the veins are obstructed, there is vascular
tenderness. A focal hypoechoic of the lower pole
engorgement and the testis becomes enlarged
shows marked hyperaemia in contrast to the upper
and oedematous, with a hypoechoic appearance
pole. The diagnosis was focal orchitis.
as compared to the contralateral testicle.6 After
24 h, vascular congestion, haemorrhage and
thickening and no colour Doppler flow in the testis infarction will cause the testis to appear hetero-
or spermatic cord.27 Intravaginal torsion occurs geneous (Fig. 12.17). The epididymis may also
most frequent in adolescent boys. A bell-clapper be enlarged and hypoechoic because of decreased
deformity, in which the tunica vaginalis completely blood supply.
encircles the epididymis, distal spermatic cord, The addition of colour Doppler increases
and testis, is a key predisposing factor.This defor- sensitivity in adults to 90–100% with a high
mity is usually bilateral and leaves the testis free specificity. 1,23,31 Unlike grey-scale imaging,
to swing and rotate within the tunica vaginalis. colour and spectral Doppler are almost always
The diagnosis of spermatic cord torsion must abnormal even during the early stages of torsion.15
be established quickly to allow for prompt surgical Instrument settings should be optimised on the
intervention, since obstruction of blood flow may normal side to identify low-velocity flow before
result in the loss of testicular viability within a ascertaining that there is indeed decreased blood
few hours of onset of symptoms. Clinical history flow on the symptomatic side (Fig. 12.10). If arterial
and physical findings, however, overlap with that flow cannot be detected in the symptomatic
of inflammatory disease to such a degree that testicle but can in the contralateral testicle, the
even an experienced urologist may have difficulty diagnosis of torsion can be effectively estab-
in differentiating the two conditions. Symptoms lished.15 The characteristic finding of ischaemia
Fig. 12.17 Transverse imaging

view, power Doppler view, and
resected specimen of a testicle.
Grossly irregular echotexture and
absence of flow on power Doppler
indicate that this is a severely
compromised testicle with torsion.
The specimen reveals a frankly
necrotic testicle.
275
12
is a completely avascular testicle. In the late
stages of torsion, colour Doppler may reveal an
increase in peritesticular blood flow because of
inflammation in the surrounding soft tissues of
the scrotum.1–3,5,32
Evaluation of the small testicles of prepu-
bescent boys, however, can be very difficult
because of inherently low velocity blood flow.15,31
In addition, if the patient has intermittent torsion
and, by the time the patient arrives for the ultra-
sound examination the torsion spontaneously
resolves, flow may appear normal on colour
Doppler. Since absence of flow is demonstrated Fig. 12.18 Longitudinal view of right testicle.
only if torsion is present at the time of sono- This young man complained of intermittent severe
graphic examination, differentiating between pain. The colour Doppler image shows only a few tiny
normal testicles and testicles with intermittent foci of flow, the most prominent at the edge of the
torsion may be difficult to accomplish. testicle. A spectral Doppler tracing shows reversed
With full torsion spectral Doppler adds little flow in diastole. These findings support the diagnosis
other than to confirm the obvious fact that there of partial torsion with occluded venous outflow.
is no flow. With partial torsion, however, the
spermatic cord twists only enough to occlude blood flow is re-established to the testicle and
the venous outflow. Because the artery has a can be detected on Doppler (Fig. 12.19).This non-
thicker wall patency is maintained.The resultant surgical approach, however, is only considered a
spectral tracing presents a high resistance arterial temporising measure. It is not a substitute for
waveform (Fig. 12.18).33 surgical intervention, which is still necessary to
If the spermatic cord spontaneously untwists correct the underlying anatomical deformity
prior to the ultrasound examination, colour Doppler that predisposes to torsion.
will likely reveal diffuse, reactive hyperaemia. The use of echo-enhancing agents has been
Although this finding will mimic epididymo- studied for improving imaging of small testicles
orchitis, resolution of acute scrotal pain concurrent with low-velocity and low-volume flow. In an animal
with increased blood flow is highly indicative of study by Brown et al,34 the authors examined
spontaneous detorsion.2,3,12,23 Colour Doppler induced testicular torsion with grey-scale imaging,
ultrasound can also be used to monitor non- colour Doppler, power Doppler and spectral
surgical detorsion of the testicle as the testis is Doppler analysis. Injection of contrast media did
manually rotated; if this manoeuvre is successful, not enhance grey-scale images, but visualisation
Fig. 12.19 Longitudinal power

Doppler image of the spermatic
cord (SC) and testicle (T). This image
SC is a composite of several individual
frames. Flow is seen within the
cord and suddenly stops with no
T
flow identified in the testicle
consistent with torsion.
276
12
Acute scrotal pain

of all vessels in both normal and rotated testicles
was significantly improved with both colour and
power Doppler. Asymmetry of blood flow was
more obvious. The authors concluded that
diagnosis of testicular ischaemia could be made
with more confidence using an intravenous
ultrasound contrast agent because of improved
demonstration of altered perfusion patterns.
Similar findings were reported by Coley et al35
in an animal study comparing unenhanced and
contrast-enhanced Doppler analysis of acute
testicular torsion. The findings from these
studies indicate that use of contrast ultra-
sonography will likely reduce, and possibly
Fig. 12.20 A longitudinal view of the left testes at
eliminate, the need for scintigraphy in currently the area of maximum tenderness in this patient.
difficult-to-visualise testicles. The appendix testis was conspicuous as the focus
Hand-held continuous-wave Doppler has no of pain. Note the hyperaemia at the base of this
role in the evaluation of testicular torsion structure. It is quite rare to actually identify colour
because of its inability to provide range-gated Doppler flow abnormalities in the appendix testis.
information. Normal or increased blood flow Typically patients present well after the structure has
within the scrotal wall may lead the examiner to necrosed and sloughed.
incorrectly conclude that intratesticular flow is
preserved.
Although testicular torsion may be accurately in which case it becomes known as a scrotal
diagnosed by sonography, this does not guarantee pearl. Twinkle artifact may be present on colour
successful surgical salvage. Success depends on Doppler and should not be confused for true
a timely diagnosis and the duration of ischaemia. flow (Fig. 12.21).
Indeed, the more obvious the ultrasound diag-
nosis, the less likely the chance of salvage. The Trauma
testicle devoid of colour flow but with an appro- Testicular trauma typically results from an athletic
priate ultrasound appearance is much more injury, a direct blow or a straddle injury. Trauma
likely to be salvaged at surgery. The testicle that can result in testicular contusion, haematoma,
appears hypoechoic or worse, showing frank fracture or rupture. Clinical examination of a
degeneration is likely to be removed.36 traumatised, tender testicle can be difficult because
The appendix testis is a Müllerian duct remnant of pain and swelling; however, ultrasound evalua-
that is attached to the upper pole of the testis. tion is able to provide information regarding
Patients can develop torsion of the appendix testicular integrity, and the presence of haematoma,
testis and usually present with acute scrotal haematocele, fracture or rupture. Colour Doppler
pain. Physical examination reveals a small firm, sonography provides excellent delineation of
palpable, tender nodule. Ultrasound evaluation blood flow throughout the testis, differentiating
usually reveals a hyperechoic mass with a hyperaemic, contused regions from devascularised
reactive hydrocele. Color Doppler may reveal or ischaemic areas (Fig. 12.22).
increased flow around the twisted testicular In the case of testicular rupture, sonographic
appendage but more importantly it rules out identification is extremely important because
testicular torsion and acute epididymo-orchitis prompt diagnosis and quick surgical intervention
(Fig. 12.20). After torsion and necrosis the is required successfully to correct the condition
appendix testis may slough and become calcified (Fig. 12.23). If surgery is performed within 72 h 277
12
Fig. 12.21 Image at the base of the right scrotum,

Fig. 12.22 Longitudinal view of patient with acute
distant from the testicle itself. An echogenic structure
trauma. Scrotal swelling and tenderness prevented
with acoustic shadowing is layering dependently.
adequate physical examination. Imaging reveals a
Power Doppler reveals colour emanating from this
large echogenic collection within the tunica vaginalis
structure indicating frequency shift. This is consistent
consistent with an acute haematoma (H). The testicle
with twinkle artifact produced when the Doppler
is hyperaemic, however, integrity is maintained and
beam interacts with crystalline material on this
flow is present throughout.
calcified left scrotal pearl.
of the trauma, approximately 80% of ruptured

testes can be salvaged.37 If rupture is present, ultra-
sound examination may demonstrate a disrupted T
tunica albuginea; a heterogeneous testicle with
asymmetric, poorly defined margins; thicken-
ing of the wall of the scrotum; and/or a large
haematocele.16,38 Perception of blood flow will
be diminished or absent on colour or spectral
Doppler examination.
Unlike rupture, fractures and small haematomas
and haematoceles do not require surgery if the
tunica albuginea has not been interrupted and
Doppler imaging shows normal blood flow to
the testicle. Sonographic findings associated with
a testicular fracture include a linear hypoechoic
band crossing the parenchyma of the testicle, a
smooth well-defined testicular outline, an intact
tunica albuginea, and often an associated
haematocele. Normal Doppler signals indicate
unimpaired blood flow and viable testicular Fig. 12.23 Longitudinal view of the right testicle (T).
Only the upper portion can be clearly identified.
tissue. If Doppler signals are absent, ischaemia is
The ovoid shape is lost inferiorly where the tunica
very likely and surgical intervention is called for.
albuginea has ruptured. Colour confirms flow through
Acute haematomas are usually hyperechoic relative
the upper part of the testicle, hyperaemia in the mid-
278 to adjacent testicular parenchyma.38 Older testicle, and very little flow in the ruptured segment.
12
Scrotal mass
haematomas may have both hyper- and hypoe- The general ultrasound appearance of germ
choic areas, and there may be associated thickening cell tumours is that of a heterogeneous, small
of the scrotal wall. On colour Doppler, haematomas mass with poorly defined margins, anechoic areas
are usually avascular.1,2,3 Acute haematoceles due to cystic necrosis, and echogenic foci of
tend to be echogenic, while low-level echoes or haemorrhage. 44 If the tumour invades the
septations may be seen in chronic haematoceles. tunica, the normal contour of the testicle may be
distorted.11
Teratomas generally appear on ultrasound as
SCROTAL MASS
extremely heterogeneous masses with well-defined
Testicular neoplasm margins and areas of various sizes that may be
Patients with testicular neoplasms usually present either hyperechoic or hypoechoic.11 Dense
with a palpable scrotal nodule. For palpable scrotal echogenic foci caused by calcification, cartilage,
masses, ultrasound is widely considered the immature bone, fibrosis, and non-calcified
imaging modality of choice.The principal role of fibrous tissue can cause acoustic shadowing. Old
ultrasound examination in the diagnosis of haemorrhage and necrosis may result in hypoe-
testicular cancer is to distinguish intra- from choic areas. Cysts are a common characteristic
extratesticular lesions, because the majority of of teratomas and can cause increased thick-walled
extratesticular masses are benign, whereas anechoic areas with through-transmission.45
intratesticular masses are considered malignant Choriocarcinomas generally appear as small
until proven otherwise.39,40 Ultrasound can masses on ultrasound, but haemorrhage may also
differentiate solid and cystic masses and confirm be seen. Areas containing either solid or cystic
their extra- or intratesticular location. The main components associated with viable tissue, necrosis,
application of ultrasound is to identify those and haemorrhage may be observed. If calcifi-
masses which require additional assessment and cation is present, a distinct area of increased
possible surgical intervention.41 echogenicity with posterior acoustic shadowing
On grey-scale images, testicular neoplasms may be present.
usually appear as a discrete mass whose echo Testicular neoplasms have mixed histological
pattern differs from that of the normal testis. components in 40–60% of cases,46–48 with the
Most neoplasms have hypoechoic components most common combination being that of teratoma
although heterogeneity of echotexture is frequently and embryonal carcinoma (teratocarcinoma).
observed (particularly with larger tumours and Ultrasound findings of mixed tumours will vary,
non-seminomatous germ cell tumours). Sono- depending on which cell lines are dominant and
graphic differentiation of seminomas, embryonal if there are no particular ultrasound findings that
cell carcinomas, teratomas and choriocarci- permit differentiation for possible preoperative
nomas can be difficult, especially since 40–60% planning.
of testicular neoplasms have mixed histolog- Testicular carcinoid is extremely rare and
ical elements. usually presents as painless testicular enlargement
Seminomas are the most common single-cell or as a discrete mass.The tumour may be primary,
type testicular tumour in adult males (40–50%).42 associated with teratoma or metastatic. Primary
The ‘classic’ ultrasound appearance of seminomas testicular carcinoid is believed to arise from
has been described as a well-defined, uniformly pluripotential germ cells or from the development
hypoechoic lesion, with no evidence of calcifi- of a simplified teratoma without other teratoma-
cation, haemorrhage, or cystic areas.43 Modern tous elements.49 Only 71 cases of testicular
high-frequency transducers, however, show fine carcinoid have been reported in the literature.
details of the neoplasm’s internal structure and Very few (<3%) have associated carcinoid
seminomas may appear to be less homogeneous syndrome.50 On ultrasound testicular carcinoid
than previously described. appears as a solid well-defined hypoechoic 279
12
a
intratesticular mass with or without dense
calcification. On Doppler it has increased
vascularity similar to seminoma51 (Fig. 12.24).
If a testicular mass is suspected of being a
tumour, the rest of the scrotum should be
examined carefully to exclude any invasion of the
tunica albuginea or epididymis by the neoplasm.
Enlargement of the epididymis usually indicates
epididymo-orchitis or torsion, rather than neoplasm.
An extratesticular fluid collection normally
indicates inflammation, torsion or trauma; although
testicular neoplasms can be associated with hydro-
cele. Because a hypoechoic appearance has also
b
been reported with other testicular conditions
(e.g. epididymo-orchitis, trauma, spermatic cord
torsion, sarcoid), the additional extratesticular
findings may help in the differential diagnosis
(Fig. 12.25).
Colour Doppler and spectral Doppler sono-
graphy are considered to be of minimal benefit
in the evaluation and characterisation of adult
testicular masses and the diagnosis of testicular
neoplasm. This is because vascularity of these
lesions is extremely variable (Figs 12.26 and
12.27).41 Small lesions tend to be hypovas- Fig. 12.24 (a) Right scrotal sonography shows a well-
cular while larger lesions tend toward hyper- defined solid mass without calcification in the right
vascular compared with normal testicular testis. (b) Colour Doppler shows the hypervascularity
parenchyma. An infiltrative neoplasm of the of the right testicular mass. Case courtesy of Kyoung-
testicle, such as leukaemia or lymphoma, typically Sik Cho, MD, Asan Medical Center, University of
presents as an enlarged hypoechoic area on grey- Ulsan, College of Medicine.
scale imaging. Colour Doppler may demonstrate
hyperaemia in the neoplasm with increased flow
in areas of leukaemic or lymphomatous involve- surgical corroboration is required to differen-
ment (Fig. 12.28), or flow only along the tiate these conditions.52,53
periphery of the lesion, but the appearance is It has been suggested that colour Doppler
quite similar to inflammation and clinical or imaging may be more useful in evaluating
Fig. 12.25 Longitudinal images of

the right and left testicles show
two hypoechoic lesions in both.
Colour shows essentially no flow
within the lesions. These findings
are consistent with the patient’s
known history of sarcoidosis and
represent granulomas.
280
Fig. 12.26 Grey-scale and colour

12
Scrotal mass
Doppler longitudinal images of
the left testis being evaluated for
a palpable mass. The normal
testicular echotexture is distorted.
There is a slightly hypoechoic mass.
Colour Doppler shows a rim of
hyperaemia, but the central portion
of the mass is relatively avascular.
Fig. 12.28 Longitudinal colour Doppler image of a

59-year-old male with a palpable scrotal mass.
Overlying a hypoechoic mass colour Doppler reveals
hyperaemia of the capsular artery and centripetal
arteries without much anatomical distortion.
Colour imaging alone would have suggested an
inflammatory process, but the patient is absolutely
asymptomatic and therefore lymphoma was
entertained and proven at surgery.
Fig. 12.27 Longitudinal images of a testis being
evaluated for a palpable mass. The large hypoechoic
irregular mass replaces the majority of this testis. cystic lesions are benign. The list includes cysts
The colour Doppler image shows marked hyperaemia
of the tunica albuginea, simple intratesticular cysts,
throughout this large mass. Note the distorted
epidermoid cyst, tubular ectasia of the rete testis
vascular architecture indicates tumour neovascularity.
(Fig. 12.29), and intratesticular spermatocele.
Colour Doppler ultrasound helps in confirming
paediatric patients in whom testicular neoplasms their benign nature of these since no blood flow
may be difficult to distinguish with grey-scale will be seen within these cysts.55
imaging but tend to be hypervascular on colour Bilateral, eccentrically-located, intratesticular
Doppler.54 adrenal rest tumours may be seen in patients
with congenital adrenal hyperplasia and primary
Benign testicular lesions adrenal insufficiency.56 In most cases, they are
Benign intratesticular masses are rare, but recog- hypoechoic oblong lesions peripherally located
nition is important to avoid unnecessary biopsy, close to the mediastinum.57 They are bilateral in
or worse orchiectomy. Almost all intratesticular 83–100% of cases.58 They may undergo extensive 281
12
Fig. 12.29 Grey-scale and power
Doppler image focused on the
region of the mediastinum testis.
A collection of small tubules is
present. Power Doppler shows the
absence of flow within these
tubules, consistent with dilated rete
testis.
fibrosis and eventually become hyperechoic with accurate in detecting varicocele.13 At rest and
acoustic shadowing.Vascularity on colour Doppler with normal respiration, colour will saturate the
is variable relative to the normal testicle. Vessels tubules at intervals related to respiratory pressure
may be seen entering from the adjacent testis fluctuations (Fig. 12.31). With more vigorous
without change in course or calibre (Fig.12.30). respiration, to-and-fro movement of blood may
Some lesions may exhibit a spoke-like pattern of
converging vessels.59 There are two theories for
a
the origin of these lesions. One says they originate
from hilar pluripotential cells, which proliferate
as a result of the elevated level of adrenocorti-
cotropic hormone.60 The other says they originate
from aberrant adrenal cortical tissue that adheres
to the testes and descends during prenatal life.61
Whatever the origin, these should be recognised
as benign lesions and first treated with adrenal
suppression with dexamethasone.
b
Varicocele
Varicoceles are present in approximately 15% of
men.62 Incompetent or absent valves in the internal
testicular veins predispose to stasis or retrograde
blood flow, resulting in dilatation of the pampini-
form plexus. This is responsible for the majority
of varicoceles.Varicoceles occur more commonly
on the left; this is attributed to the longer course
of the gonadal vein and its direct drainage into
the left renal vein.Varicoceles are important clini-
cally because of their association with infertility.
Fig. 12.30 (a) Transverse panoramic view of both
Diagnosis of varicoceles is important, because
testicles. Hypoechoic masses are identified adjacent
treatment improves sperm quality in over half of
to the mediastinum of both testes. (b) Colour Doppler
cases.63 image of the left paramediastinal mass shows
On ultrasound a varicocele is seen to consist hyperaemia. Vessels appear to be passing through
of dilated (>2 mm diameter), serpiginous channels this region. The patient had a known history of
in the head of the epididymis and spermatic congenital adrenal hyperplasia and these represent
282 cord. Colour Doppler has been shown to be very benign rests of adrenal tissue.
12
Other imaging procedures

brief burst of reversed flow is common. This is
due to the expulsion of the venous blood in the
gonadal vein of the pelvis below the lowest compe-
tent valve. As soon as this volume of blood is
expressed, then flow usually stops, waiting for the
scrotal venous pressure to rise above that created
by the Valsalva. When the pressure is released as
the patient relaxes, the direction of blood flow
reverts to normal (Fig. 12.32).
Whenever a varicocele is identified, obstruc-
tion by a retroperitoneal mass, such as a left
Fig. 12.31 Longitudinal view of the left spermatic renal malignancy invading the renal vein should
cord in a patient with a palpable thickening and be considered. A brief scan of the upper
infertility. A serpiginous collection of vessels with a abdomen should be performed to assess for
venous flow profile is consistent with that of a this possibility.64
varicocele. Note the laminar flow.
OTHER IMAGING PROCEDURES

manifest as alternating colour in the same vessel,
changing between inspiration and expiration. Computed tomography (CT) is primarily used
Colour Doppler identification of varicocele is for staging and follow-up of testicular tumour
enhanced by having the patient perform a Valsalva metastatic to the retroperitoneum or elsewhere.
manoeuvre or by having the patient stand up. Magnetic resonance imaging (MRI), because of
This increases the abdominal pressure and results its high cost and limited availability, is reserved
in the reversal of blood flow into the pampini- for problem solving of difficult cases. MRI has
form plexus, thereby causing further distention. been shown to be diagnostic and cost effective
When the Valsalva manoeuvre is performed, a following equivocal scrotal ultrasound, however,
Fig. 12.32 Spectral Doppler tracings of the left spermatic cord in two patients with a known varicocele.
The tracings were obtained as a Valsalva manoeuvre was initiated, maintained, and released. The first patient
has a brief reversal of flow at the start of Valsalva; blood below the distal most valve is forced downward, but
then because of valve patency, the flow reversal stops. Upon relaxation of the Valsalva, flow surges forward
again. The second patient initiates reversal of flow and then maintains persistent flow reversal as renal venous
outflow is shunted down the gonadal vein because of the high pressure in the thorax. As pressure builds within
the testicle, flow progressively slows until the point of relaxation at which time flow surges forward. 283
12
even in an aggressive MRI environment only of prepubescent patients) or when the examiner
1.4% of the scrotal sonograms required the has limited proficiency with colour Doppler
addition of MRI.65 Currently, other than in select evaluation.
paediatric cases and in the evaluation of cryp-
torchidism, MRI has not been found to hold
CONCLUSIONS
significant advantage over ultrasound in the
evaluation of the scrotum; but the modality is The scrotal contents are ideally situated for
evolving.66 examination by ultrasound. Their superficial
Although scintigraphy continues to be a location allows the use of high-frequency trans-
dependable means of imaging testicular blood ducers and yields very high resolution images
flow, it lacks sonography’s ability to provide of the testes and associated structures.The addi-
anatomical information, as well as perfusion tion of haemodynamic information by spectral
status, and it exposes the patient to radiation. and colour Doppler allows the examiner to
Therefore, as suggested by Siegel,15 nuclear make a very specific diagnosis with a high degree
scintigraphy should be reserved for those situa- of confidence in several very important disease
tions when the sensitivity of colour Doppler for states. Thoughtful fusion of the anatomical and
low-velocity, low-volume testicular arterial flow physiological findings will allow many problems
is not satisfactory and there are questions to be clarified and managed without the need for
regarding the findings (e.g. in the small testicles further imaging.
REFERENCES
1. Horstman WG, Middleton WD, Melson GL, et al. 10. Holder LE, Martire JR, Holmes ER, et al. Testicular
Colour Doppler US of the scrotum. Radiographics radionuclide angiography and static imaging:
1991; 11:941–957. anatomy, scintigraphic interpretation, and clinical
2. Learner RM, Mevorach RA, Hulbert WC, et al. indications. Radiology 1977; 125:739–752.
Colour Doppler US in the evaluation of acute scrotal 11. Krone KD, Carroll BA. Scrotal ultrasound. Radiol Clin
disease. Radiology 1990; 176:355–358. North Am 1985; 23:121–139.
3. Luker GD, Siegel MJ. Colour Doppler sonography of 12. Middleton WD. Scrotal sonography in 1991.
the scrotum in children. AJR Am J Roentgenol Ultrasound Q 1991; 9:61–87.
1994; 163: 649–655. 13. Petros JA, Andriole GL, Middleton WD, et al.
4. Luker GD, Siegel MJ. Scrotal US in pediatric Correlation of testicular colour Doppler
patients: comparison of power and standard colour ultrasonography, physical examination and
Doppler US. Radiology 1996; 198:381–385. venography in the detection of varicoceles in men
5. Middleton WD, Siegel BA, Melson GL, et al. with infertility. J Urol 1991; 145:785–788.
Acute scrotal disorders: prospective comparison 14. Giorgio G. Cardiovascular (inferior vena cava).
of colour Doppler US and testicular scintigraphy. In: Williams PL, Bannister LH, Berry MM, et al, eds.
Radiology 1990; 177:177–181. Gray’s anatomy, 38th edn. New York: Churchill
6. Ralls PW, Larsen D, Johnson MB, et al. Colour Livingston; 1995:1600–1601.
Doppler sonography of the scrotum. Semin 15. Siegel MJ. The acute scrotum. Radiol Clin North Am
Ultrasound CT MR 1991; 112:109–114. 1997; 35:959–976.
7. Bannister LH, Dyson M. Reproductive system. 16. Siegel MJ. Male pelvis. In: Siegel MJ,ed. Pediatric
In: Williams PL, Bannister LH, Berry MM, et al, eds. sonography, 2nd edn. New York: Raven Press;
Gray’s anatomy, 38th edn. New York: Churchill 1995:479–512.
Livingston; 1995:1848–1856. 17. Paltiel HJ, Rupich RC, Babcock DS. Maturational
8. Doherty FJ. Ultrasound of the nonacute scrotum. changes in arterial impedance of the normal testis
Semin Ultrasound CT MR 1991; 12:131–156. in boys: Doppler sonographic study. AJR Am J
9. Middleton WD, Thorne DA, Melson GC. Colour Roentgenol 1994; 163:1189–1193.
Doppler ultrasound of the normal testis. AJR Am J 18. Jequier S, Patriquin H, Filiatrault D, et al. Duplex
284 Roentgenol 1989;152:293–297. Doppler sonographic examination of the testis in
prepubertal boys. J Ultrasound Med 1993;
contrast-enhanced power Doppler US, colour

12
References
12:317–322. Doppler US, and radionuclide imaging. Radiology
19. Hermansen MC, Chusid MJ, Sty JR. Bacterial 1996; 199:441–446.
epididymo-orchitis in children and adolescents. 36. Middleton WD, Middleton MA, Dierks M, et al.
Clin Pediatr 1980; 19:812–815. Sonographic prediction of viability in testicular
20. Basekim CC, Kizilkaya E, Pekkafali Z, et al. Mumps torsion: preliminary observations. J Ultrasound Med
epididymo-orchitis: sonography and colour Doppler 1997; 16:23–27.
sonographic findings. Abdom Imaging 2000; 37. Cass AS, Luxenberg M. Testicular injuries. Urology
25:322–325. 1991; 37:528–530.
21. Horstman WG, Middleton WD, Nelson GL. Scrotal 38. Jeffrey RB, Laing FC, Hricak H, et al. Sonography of
inflammatory disease: colour Doppler US findings. testicular trauma. AJR Am J Roentgenol 1983;
Radiology 1991; 179:55–59. 141:993–995.
22. Cook JL, Dewbury K. The changes seen on high- 39. Geraghty MJ, Lee FT, Jr, Bernsten SA, et al.
resolution ultrasound in orchitis. Clin Radiol 2000; Sonography of testicular tumors and tumor-like
55:13–18. conditions: a radiologic-pathologic correlation.
23. Burks DD, Markey BJ, Burkhard TK, et al. Crit Rev Diagn Imaging 1998; 39:1–63.
Suspected testicular torsion and ischaemia: 40. Ulbright TM, Roth LM. Testicular and paratesticular
evaluation with colour Doppler sonography. tumors. In: Sternberg SS, eds. Diagnostic surgical
Radiology 1990; 175:815–821. pathology, 3rd edn. Philadelphia: Saunders;
24. Farriol VG, Comella XP, Agromayor EG, et al. 1999:1973–2033.
Gray-scale and power Doppler sonographic 41. Horstman WG, Melson GL, Middleton WD, et al.
appearances of acute inflammatory diseases of the Testicular tumors: findings with color Doppler US.
scrotum. J Clin Ultrasound 2000; 28:67–72. Radiology 1992; 185:733–737.
25. Sanders LM, Haber S, Dembner A, et al. 42. Mostofi FK. Tumors of the testis. IARC Sci Pub
Significance of reversal of diastolic flow in the 1994; 111:407–429.
acute scrotum. J Ultrasound Med 1994; 43. Cotran RS, Kumar P, Collins T. The male genital tract.
13:137–139. In: Robbins SL, ed. Pathologic basis of disease,
26. Backhouse K. Embryology of testicular descent 6th edn. Philadelphia: Saunders; 1999:1011–1034.
and maldescent. Urol Clin North Am 1982; 44. Kurman RJ, Scardino PT, McIntire KR, et al.
9:315–325. Cellular localization of alpha-fetoprotein and
27. Zerin J, DiPietro M, Grignon A, et al. Testicular human chorionic gonadotropin in germ cell tumours
infarction in the newborn: ultrasound findings. of the testis using an indirect immunoperoxidase
Pediatr Radiol 1990; 20:329–330. technique (a new approach to classification utilizing
28. Dubinsky TJ, Chen P, Maklad N. Colour-flow and tumor markers). Cancer 1977; 40:2136–2151.
power Doppler imaging of the testes. World J Urol 45. Woodward PJ, Sohaey R, O’Donoghue MJ, et al.
1998; 16:35–40. Tumours and tumour like lesions of the testis:
29. Meza MP, Amundson GM, Aquilina JW, et al. radiologic-pathologic correlation. Radiographics
Colour flow imaging in children with clinically 2002; 22:189–216.
suspected testicular torsion. Pediatr Radiol 1992; 46. Muir CS, Nectoux J. Epidemiology of cancer of the
22:370–373. testis and penis. Nat Cancer Inst Monogram 1979;
30. Middleton WD, Melson GL. Testicular ischemia: 53:157–164.
colour Doppler sonographic findings in five patients. 47. Carroll BA, Gross DM. High frequency scrotal
AJR Am J Roentgenol 1989; 152:1237–1239. sonography. AJR Am J Roentgenol 1983;
31. Patriquin HB, Yazbeck S, Trinh B, et al. Testicular 140:511–515.
torsion in infants and children: diagnosis with 48. Mostofi FK. Testicular tumours: epidemiologic,
Doppler sonography. Radiology 1993; 188:781–785. etiologic and pathologic features. Cancer 1973;
32. Leahy PF. Diagnosis of testicular torsion using 32:1186–1201.
Doppler ultrasonic examination. Br J Urol 1986; 49. Terhine DW, Manson AL, Jordon GH, et al.
58:696–697. Pure primary testicular carcinoid: a case report
33. Dogra VS, Sessions A, Mevorach A, et al. Reversal and discussion. J Urol 1988; 4: 255–256.
of diastolic plateau in partial testicular torsion. J Clin 50. Zavala-Pompa A, Ro JY, El-Naggar A, et al. Primary
Ultrasound 2001; 29:105–108. carcinoid tumor of testis. Immunohistochemical,
34. Brown JM, Taylor KJW, Alderman JL, et al. ultrastructural, and DNA flow cytometric study of
Contrast-enhanced ultrasonographic visualization of three cases with a review of the literature. Cancer
gonadal torsion. J Ultrasound Med 1997; 1993; 72: 1726–1732.
16:309–316. 51. Kim BH Scrotum. In: Kim SH, ed. Radiology
35. Coley BD, Frush DP, Babcock DS, et al. Acute illustrated: uroradiology. Philadelphia: WB Saunders;
testicular torsion: comparison of unenhanced and 2003:625–664.
285
12
52. Patriquin HB. Leukemic infiltration of the testis. six cases and review of the literature on testicular
In: Siegel BA, Proto AV, eds. Pediatric disease, 4th masses in patients with adrenocortical disorders.
series. Test and syllabus. Reston: Am Coll Radiol; Am J Surg Pathol 1988; 12(7):503–513.
1993:667–688. 61. Stikkelbroeck NM, Otten BJ, Pasic A, et al.
53. Mazzu D, Jeffrey RB, Ralls PW. Lymphoma and High prevalence of testicular adrenal rest tumors,
leukemia involving the testicles: findings on gray- impaired spermatogenesis, and Leydig cell failure in
scale and colour Doppler sonography. AJR Am J adolescent and adult males with congenital adrenal
Roentgenol 1995; 164:645–647. hyperplasia. J Clin Endocrinol Metab 2001;
54. Luker GD, Siegel MJ. Pediatric testicular tumors: 86:5721–5728.
evaluation with gray-scale and colour Doppler US. 62. Meacham RB, Townsend RR, Rademacher D,
Radiology 1994; 191:561–564. et al. The incidence of varicoceles in the general
55. Dogra VS, Gottlieb RH, Rubens DJ, et al. population when evaluated by physical examination,
Benign intratesticular cystic lesions: US features. gray scale sonography and colour Doppler
Radiographics 2001; 21(spec no.):S273–S281. sonography. J Urol 1994; 151:1535–1538.
56. Seidenwurm D, Smathers R, Kan P, et al. 63. Pierik FH, Dohle GR, van Muiswinkel JM, et al.
Intratesticular adrenal rests diagnosed by US. Is routine scrotal ultrasound advantageous in
Radiology 1985; 155:479–481. infertile men? J Urol 1999; 162:1618–1620.
57. Dogra VS, Gottlieb RH, Oka M, et al. Sonography of 64. Dambro TJ, Stewart RR, Barbara CA. The scrotum.
the scrotum. Radiology 2003; 227:18–36. In: Rumack CM, Wilson SR, Charboneau JW, eds.
58. Proto G, DiDonna A, Grimaldi F, et al. Bilateral Diagnostic ultrasound, 2nd edn. St Louis: Mosby;
testicular adrenal rest tissue in congenital adrenal 1998:791–821.
hyperplasia: US and MR features. J Endocrinol 65. Serra AD, Hricak H, Coakley FV, et al. Inconclusive
Invest 2001; 24:529–531. clinical and ultrasound evaluation of the scrotum:
59. Avila NA, Premkumar A, Shawker TH, et al. impact of magnetic resonance imaging on patient
Testicular adrenal rest tissue in congenital adrenal management and cost. Urology 1998;
hyperplasia: findings at gray-scale and colour 51:1018–1021.
Doppler US. Radiology 1996; 198:99–104. 66. Choyke PL. Dynamic contrast-enhanced MR
60. Rutgers JL, Young RH, Scully RE. The testicular imaging of the scrotum: reality check (editorial).
‘tumor’ of the adrenogenital syndrome. A report of Radiology 2000; 217:14–15.
286
13
Doppler ultrasound of the
female pelvis
Paul A. Dubbins
Ultrasound has assumed a central role in the the pelvic organs, and hitherto our methods of
investigation of gynaecological physiology and discriminating between the uncontrolled and
pathology.The advent of transvaginal probes has irregular angiogenesis of malignant tumour and
further advanced gynaecological applications to the angiogenesis associated with benign tumour,
the extent that it is now an indispensable tool in or even the hyperaemia of inflammatory condi-
the evaluation of the female pelvis. Ultrasound tions, have been insufficiently sophisticated to
predominantly assesses structure, although make a reliable distinction. This chapter presents
physiological changes can be inferred, for example the current state of knowledge for potential appli-
by sequential ultrasound examinations such as cations of Doppler ultrasound in gynaecological
change in the size and the appearance of an ovarian disease, while indicating those areas where the
follicle over time, and the pattern of thickening role of Doppler ultrasound is established and
of the endometrium. considering these in more detail.
The application of Doppler techniques allows
the demonstration of changes in perfusion of the
ANATOMY
uterus and ovaries at different phases of the
menstrual cycle. Abnormalities of perfusion indices Knowledge of the course of the pelvic vessels is
have been shown also to reflect features of important for the proper performance and inter-
subfertility. Similarly there are marked vascular pretation of the Doppler examination. Although
changes known to take place in pathological condi- the attention is drawn to the examination of the
tions. For example, neovascularity is an early ovarian and uterine arteries, the anatomical rela-
and persistent feature of tumour growth with the tions of the pelvic organs to the iliac arteries and
development of random and chaotic vessels, lack veins is also important. Pathological conditions
of hierarchical branching, focal calibre variations affecting the major vessels in the pelvis may
and blind ending lakes. If documentation of angio- complicate uterine or ovarian pathology, such as
genesis by Doppler techniques was achievable, iliac venous thrombosis, or may mimic gynae-
this might be of value in differentiation of benign cological pathology, such as an iliac artery
from malignant tumours. aneurysm.
More recently it has been suggested that quan- The iliac arteries and veins course inferiorly
tification of organ flow may be best achieved and laterally from the aortic bifurcation and venous
using quantitative three-dimensional colour power confluence respectively on the anteromedial surface
Doppler.This technique however requires propri- of the psoas muscle, to become the femoral artery
etary software which is not yet widely available. and vein as they emerge beneath the inguinal
The indications for Doppler ultrasound in the ligament in the groin. Surface markings for the
evaluation of pelvic physiology and pathology in common and external iliac artery and vein are
the female are therefore potentially large. Many approximated by a line drawn from the umbilicus
pathologies will produce increased blood flow to to the site of maximum pulsation in the groin. 287
13
Doppler ultrasound of the female pelvis
The common and external iliac veins lie medial the abdomen it lies medial and posterior to the
and posterior to the artery.The vessels often form ureter, crossing the external iliac artery and vein
a lateral anatomical relationship to the ovary to enter the true pelvis, where it turns medially
(Fig. 13.1). The internal iliac artery arises from in the ovarian suspensory ligament, passing into
the medial aspect of the common iliac artery the broad ligament where its terminal branches
along with the vein approximately 4 cm from the supply the ovary and anastomose with adnexal
aortic bifurcation. It gives rise to two branches, branches of the uterine artery. The left ovarian
an anterior and a posterior trunk. At this point, vein drains in to the left renal vein and the right
just distal to the bifurcation, the anterior trunk into the inferior vena cava.The integrity of venous
lies posterior to the ureter and to the ovaries. return is maintained by venous valves within the
The anterior trunk has several branches, one of upper vessel and if these valves become compro-
which is the uterine artery which runs medially mised, venous congestion within the pelvis may
on the surface of the levator ani muscles, crossing occur.
above the ureter and ascending in a tortuous
fashion lateral to the uterus, giving off uterine
TECHNIQUE
branches. It is accompanied through its course
by the uterine vein. The examination of the pelvic vessels has been
The ovary has a dual arterial blood supply. significantly altered by the transvaginal technique
This has assumed greater importance with the of evaluation of pelvic anatomy and pathology.
advent of minimally invasive procedures for the The course of the uterine artery is particularly
treatment of gynaecological pathology, for suitable for transvaginal assessment, with ideal
example fibroid embolisation. geometry for Doppler signal recording (Fig. 13.2).
The ovarian artery is a branch of the renal Similarly, however, it is possible to record
artery on the left, although on the right it may Doppler signals transabdominally with an empty
arise from the aorta. Throughout its course in bladder, when the uterus is normally anteverted,
for similar reasons of geometry. When the
bladder is full, however, the angle of incidence of
the Doppler beam to the uterine arteries is not
optimised in spite of good visualisation of the body
of the uterus (Figs 13.3 and 13.4). The ovarian
arteries, running a somewhat transverse course
through the pelvis, are more difficult to assess,
although greater sensitivity of signal recording is
almost invariably achieved with transvaginal
scanning.
Demonstration of the spiral arteries within
the uterus and intraovarian vessels requires
colour flow or power Doppler for their iden-
tification. It is difficult to be prescriptive about
Fig. 13.1 Colour flow Doppler study of the left iliac
colour flow settings but the following generali-
fossa using compression. The left ovary is
sations apply.
demonstrated anterior to the external iliac artery.
There is a small amount of intraparenchymal flow
within the ovary. The apparent bidirectional flow Filtration
within the iliac artery is related to the geometry with The lowest possible filtration is needed,
the proximal iliac artery, indicating flow towards the particularly when investigating pelvic venous
transducer while the distal artery shows flow away disease and conditions affecting ‘perfusion’ of
288 from the transducer. tissues.
Fig. 13.2 Uterine and ovarian artery

13
Technique
distribution. The uterine artery is a
branch of the internal iliac artery
and ascends on the lateral border
of the uterus. It sends branches to
the ovary and to the fallopian tube.
The ovary is also supplied by the
ovarian artery, a branch usually of
the renal artery.
Fig. 13.3 Sagittal sections through the pelvis

demonstrating the optimum geometry achieved by a
Fig. 13.4 Sagittal planes through the abdomen
transvaginal scan with an empty bladder.
demonstrating that with a full urinary bladder the
geometry for Doppler assessment of blood flow is
suboptimal.
Colour/B-mode priority
For the most part small vessels are being investi-
gated and it is less critical to exclude ‘colour Velocity range
bleed’ out with the wall of the vessel; high colour Selection of velocity range depends upon whether
priority should therefore be selected. arteries or veins are being interrogated. Arterial
flow within the uterine and ovarian arteries is
Persistence usually within the range of 10–50 cm s-1 peak
A moderate persistence setting should be systolic velocity, although this may be lower,
selected. particularly in postmenopausal ovarian vessels. 289
13
Flow velocity in pelvic veins is in the region of available but may allow the semiquantitative
1–10 cm s-1. Velocity range settings should be assessment of blood flow. Volume acquisition is
chosen to reflect these predicted velocities. undertaken after careful selection of the colour
power box to enclose only the region of interest
Motion artefact (the entire ovary, the follicle, the mass) and the
Most machines now have specific algorithms to volume mode sweep activated while maintaining
reduce motion artefact and these should be the position of the probe entirely stable. Storage
employed. of the 3D data is then subsequently analysed.
Proprietary software will calculate morphological
Spectral Doppler settings features such as volume and echogenicity and
These reflect the colour Doppler settings but the the number of colour voxels in the region of
need to use low filtration is particularly important interest. Derived indices include: the
in order to detect slow flow in veins and diastolic vascularisation index (VI) – the number of colour
flow in arteries. voxels in the region of interest (ROI); flow index
– the mean colour value in the voxels; vascu-
Angle of insonation larisation flow index (VFI) – the mean colour value
As with all Doppler applications, it is important in all the voxels within the sphere. Although there
to optimise the angle of insonation of the Doppler is evidence of different patterns for the dominant
beam on the vessel but this may be compromised follicle and for the right and left ovaries there is
by the direction of flow and the limitations that currently insufficient work to establish normal
transducer position, both abdominally and features to afford clearly defined clinical appli-
intravaginally; provide for variation of this angle. cations for this technique.1
Where possible angles of less than 60° should be
employed, although this is less important if veloci- Technique for evaluating pathology
ties are not being measured. Although Doppler It is important when assessing pelvic pathology
indices such as the resistance index and the with Doppler techniques that optimum signal-
pulsatility index are widely used in gynaecology recording methods are used. Evaluation of vessels
and are angle independent, these are often supple- is straightforward but in the demonstration of
mented in clinical applications by measurements mass lesions, inflammatory processes, etc., careful
of peak systolic velocity, which will require stringent attention to technique will ensure that the vascular
attention to angle optimisation and correction. supply is accurately mapped and Doppler spectra
precisely recorded.
Sample volume When a suspected pathological process is
The vessels under study are small and therefore identified on real-time ultrasound, it is assessed
sample volume size should be at a minimum. with colour Doppler. Flexibility of the use of the
colour box is important: initially for small- to
Volume flow quantification moderate-sized lesions the whole lesion can be
Using proprietary software it is possible to identify contained within the colour box in order to identify
a region of interest in colour power mode. On or regions of increased vascularity, but this large field
off line assessment of vessel distribution and power of view compromises the frame rate and pulse
values allow the development of numerical ‘perfu- repetition rate, affecting the detection of high
sion’ indices. This may be of value in evaluation velocities and potentially also pulsatile flow.
of, for example ovarian masses given that malig- Therefore a smaller colour box is then selected
nant lesions contain vessels of larger size and and individual zones within the lesion are inter-
greater density than benign masses. rogated sequentially. Attention is paid to the size,
Three-dimensional (3D) quantification distribution and communications of the vessels
290 (including 3D power Doppler) is not yet widely together with the vessels of supply and drainage.
13
Normal appearances
Throughout the procedure the machine settings pelvic organs. This has the potential to improve
are varied to optimise the system for very high the visualisation of a wide variety of pelvic
and very low velocities. Sampling with spectral pathology, but it is not yet established whether
Doppler is performed at multiple sites, as some this will improve the differentiation of different
vessels may demonstrate apparently normal flow pathologies and no distinct application has been
patterns, while others will demonstrate abnormal identified.2
flow patterns and Doppler indices. Failure to
record signals in this obsessive way will result in
NORMAL APPEARANCES
a lower accuracy of the technique.
Selection of the ROI for quantification requires Colour flow Doppler will demonstrate the
similar attention to technique with the potential uterine arteries coursing along the lateral
for interobserver variation. Drawing the ROI is aspects of the body of the uterus (Fig. 13.5).The
currently largely achieved using a freehand tech- branches of the uterine artery extending towards
nique. Standardisation of the technique, agreement the ovary and the ovarian artery can be iden-
between operators in respect to what is included tified in the broad ligament and on the superior
in the ROI (exclusion for example of cystic areas) aspect of the ovary, respectively. Because of their
will ensure that the pixel intensity is reproducible. tortuosity, only short segments of the arteries
Ultrasound contrast media will enhance are usually identified in any particular scan
the signal in all vessels including those to the plane (Fig. 13.6).
a b
c d
Fig. 13.5 (a) Sagittal transabdominal scan along the left margin of the uterus demonstrating the uterine artery
and vein. (b) Sagittal transvaginal scan along the margin of the uterus demonstrating the uterine vessels.
(c) Coronal transabdominal scan demonstrating uterine arteries. (d) Spectral Doppler from uterine artery
demonstrating automatic calculation of indices. 291
13
a b
Fig. 13.6 (a) Left ovary demonstrating uterine and

c ovarian branches (arrows). (b) Blood supply to the
fallopian tube and ovary from the uterine artery.
(c) Dual blood supply of the right adnexa.
Colour flow and power Doppler will demon- the age of the patient. The normal development
strate perfusion within the uterus and the ovaries. of the ovarian follicle is accompanied by a marked
In the uterus the spiral arteries are arranged change in the vascular pattern which reflects the
perpendicular to the long axis and can be demon- influence of vascular endothelial growth factor
strated extending from the serosal surface of the on the corpus luteum. A relatively oligaemic follicle
uterus to the endometrium. No colour flow develops may exhibit circumferential flow.
signal is routinely demonstrated in the normal However with the development of the corpus
endometrium or in the subendometrial layer. luteum a rich vascular wreath is seen supplied by
However recent improvements in colour and a helical artery. The vascularisation becomes
particularly colour power sensitivity now allow more extensive as vessels invade the central
the documentation of colour power signals within haemorrhagic zone (Fig. 13.7).
the normal endometrium and subendometrial
layer. 3D power Doppler angiography affords the Pelvic veins
semiquantitative assessment of flow by means of Normal venous structures can be seen within
the indices described above. Early reports suggest the body of the uterus and in both adnexa
that the VI and VFI increase in the proliferative coursing towards the ovarian vein and the
phase reaching a peak 3 days prior to ovulation uterine veins.
subsequently decreasing to a nadir 5 days
following ovulation3. Normal Doppler spectra
The colour flow appearance of the normal Normal Doppler spectra vary throughout the
292 ovary depends upon the phase of the cycle and menstrual cycle. This variation is particularly
13
Normal appearances
a b
Fig 13.7 (a) Colour flow Doppler of circumferential flow in corpus luteum. (b) Power Doppler of the corpus
luteum.
marked in the ovarian artery which demonstrates normal values are given in Table 13.1.There is
a low-resistance pattern at the time of the develop- an increase in volume flow and a reduction in
ment of the corpus luteum. However, there is wide resistance index in the luteal phase when
variation in reported normal values of stromal compared to the follicular phase. Again, therefore,
ovarian and ovarian artery flow. Typically the it appears that cyclical variation is more impor-
resistance index in the early follicular phase is in tant in the assessment of normal reproductive
the region of 0.65–0.7, falling to 0.55–0.6 in the late physiology, rather than using individual values.5,6
follicular phase and returning to early follicular Normal Doppler spectra are illustrated in
phase values in the early luteal phase. However, Figures 13.8–13.11.
published normal values range between 0.4 and Blood flow to the prepubertal uterus demon-
0.8 and pulsatility index values vary from 0.6 to strates a similar pattern to the postmenopausal
2.5. Similar values for ovarian vascular indices status with high impedance and absent diastolic
have also been recorded in postmenopausal ovaries flow. Change to the adult pattern of lower indices
but the cyclical variation is lost.4 Thus an indi- and forward flow throughout diastole heralds the
vidual measurement of ovarian artery impedance onset of the menarche.
is of limited value in assessment of ovarian func- Pelvic venous diameter is extremely variable
tion, although the cyclical change of ovarian flow but pelvic veins are normally less than 5 mm in
does correlate with development of the corpus diameter and flow velocities between 5 and
luteum. 10 cm s-1. In this author’s experience, reverse
Cyclical variation in uterine artery flow is well flow on Valsalva is short lived and of low velocity
defined and also appears to correlate with fertility; (<2 cm s-1).
Table 13.1 Variation in Doppler indices in the uterine and ovarian arteries in the menstrual cycle
RI OVAa PI OVAa VEL OVAa (cm s–1) PI UTAa

Early follicular 0.65–0.7 1.8–2.2 20 1.67 ± 0.22
Late follicular/ovulation 0.55–0.6 1.0–1.3 40 1.89 ± 0.4
Luteal 0.6–0.65 1.3–1.8 2.23 ± 0.67
Non-conception 0.6–0.7 1.8–2.2 3.85 ± 1.1
Postmenopausal 0.6–1.0 1.3–4.0 1.8–3.8
aRI = resistance index; PI = pulsatility index; OVA = ovarian artery; VEL = velocity; UTA = uterine artery.
293
13
a Fig. 13.8 (a) Doppler spectrum
of ovarian blood flow within the
follicular phase. (b) Low-resistance
blood flow at the time of
ovulation.
Fig. 13.9 Doppler spectrum of

normal uterine artery blood flow in
the late follicular phase.
294

13
Errors and artefacts

uterine artery blood flow in the
luteal phase.

uterine artery blood flow in the
postmenopausal uterus.
ERRORS AND ARTEFACTS7 Sample volume artefacts

Because of the small size of many of the vessels
There are a number of factors which will contribute under study, maintenance of the position of the
to alterations in flow and which are not related sample volume within the centre of the vessel is
to pelvic pathology. These are common to all frequently difficult. This has the effect that there
vessels and include the following. may be inaccurate recording of either systolic or
diastolic flow velocities, dependent upon vessel
Hypertension movement relative to the sample volume through-
Indices will be uniformly slightly increased. out the cardiac cycle. This produces variation in
the appearance of the Doppler spectrum and
Abnormalities of heart rate and variation in the Doppler indices.
rhythm Bowel movement may provide significant
Both pulsatility and resistance indices will be imaging artefact, but although this may compro-
decreased in tachycardias and increased in brady- mise Doppler signal recording, it is not usually
cardias. Irregularities such as atrial fibrillation will confused for normal or abnormal blood flow
invalidate the use of any of the Doppler indices. patterns. 295
13
Fertility
Bladder distension also has an effect on Doppler
spectra, producing a significant increase in the The cyclic changes of blood flow characteristics
impedance indices compared to the empty and Doppler indices within the uterine and ovarian
bladder.8 arteries suggest a possible application in the diag-
nosis and management of subfertility.13 However
Twinkling artefact there is lack of consensus on the value of flow
The use of colour and power Doppler may produce indices in predicting pregnancy outcome. While
colour signals even in the absence of flow. This some authors have reported differences between
may occur as a consequence of ‘over-gain’ espe- groups for example a higher mean pulsatility index
cially where there is acoustic enhancement such in the uterine and ovarian arteries of infertile
as occurs through a cyst and also may be conse- women in the mid-luteal phase than that in fertile
quent of motion within adjacent bowel. The women others have found no uniform relationship.
twinkling artefact is a feature usually of focal Furthermore there is no confirmed relationship
calcification or the reverberation that may be seen between uterine pulsatility index, subendometrial
in tiny cysts.This is a typical feature of the ‘micro- peak systolic velocity or endometrial thickness
scopic’ echogenicities seen in normal ovaries and and successful pregnancy14,15 (Fig.13.13). Jarvela
those with endometriosis9 (Fig. 13.12). et al16 have suggested that ovarian peak systolic
velocity correlates with oocyte recovery develop-
ment potential of the oocyte and pregnancy rate
APPLICATIONS
others have found no such correlation.17 Ovarian
Menstrual disorders volume appears to be related to ovarian reserve
The use of Doppler techniques to presage puberty and it may be that 3D depiction of perfusion
has already been described. Similarly in girls with patterns will be more sensitive than an evalua-
features of precocious puberty the presence of a tion of indices and will allow the development of
mature pattern pulsatility index correlates with a vascularisation index which correlates with
those patients who exhibit a pubertal response to successful pregnancy. Yang et al have suggested
growth hormone-releasing hormone (GRH).10 that the same may be true for the evaluation of
Women with primary dysmenorrhea have subendometrial vascularity.18 It is interesting
elevated Doppler indices suggesting that this is a however to note that the changes in flow asso-
more complex condition of the menstrual cycle ciated with infertility are seen not only in primary
than had hitherto been suggested. ovarian failure but also in tubal blockage although
In menorrhagia there is evidence of reduced the cause of this phenomenon is not clear.5
pulsatility index values within the uterine arteries. An elevated uterine artery pulsatility index is
These indices increase with successful treatment also seen in patients with polycystic ovarian
for example with danazol.11,12 syndrome, although the ovarian artery pulsatility
a b
296 Fig. 13.12 Twinkling artefact secondary to ‘microcalcification’ in the ovary: (a) grey-scale; (b) colour flow.

13
Applications
uterine artery blood flow in
infertility.
index is variable although there appears to be an below 0.8 and 2.0, respectively. Indeed Wiener
increase in peak systolic velocity in the stroma. et al20 indicate that the lowest impedance indices
In spite of these limitations a high vascular resist- of all the uterine pathologies are recorded in
ance in the uterine artery and ovarian artery in uterine fibroids. If the fibroid attenuates sound
the luteal phase indicates a poor ‘baby take- significantly then it may be impossible to assess
home rate’. blood flow. In cases where it is difficult to
In summary, the role of ultrasound in the
investigation and management of infertility is
complex. It may be helpful in identifying patients a
for different treatment regimens but has failed to

fulfil the promise of a precise method of assessing
uterine receptivity. The role of 3D Doppler in
the assessment of vessel density, distribution and
flow velocity remains to be established. The best
predictor for outcome may be a combination of
ovarian morphology and blood flow distribution
rather than simple indices although there may
be a role for 3D power Doppler.19
Diseases of the uterine body b
Fibroids
The vascularity of fibroids is variable. In some
patients blood flow to the uterus yields normal
resistance and pulsatility indices with uterine
vessels simply being displaced around the
fibroids. In other cases, the fibroids may demon-
strate marked vascularity with increased number
and size of uterine vessels (Figs 13.14–13.17)
together with a significant reduction in Fig. 13.14 Submucous fibroid: (a) grey-scale image
peripheral resistance to blood flow evidenced by demonstrates typical fibroid pattern; (b) diffuse colour
reduction in pulsatility and resistance indices flow throughout the fibroid. 297
13
a Fig. 13.15 (a) Irregular blood flow both at the
margin and within the uterine fibroids. There was
grey-scale evidence of partial necrosis. (b) Doppler
spectrum of same patient showing increased
systolic and diastolic velocities with reduction in
resistance indices indicating uterine hyperaemia
with vascular fibroids.

uterine artery flow in postmenopausal
uterus with uterine fibroids.
No evidence of increase in blood
flow velocities, particularly in
diastole.
298
13
Applications
Diagnosis of pathology largely depends upon iden-
tifying an endometrial thickness greater than 5 mm.
While this represents a sensitive method it lacks
specificity and does not allow the differentiation
of benign from malignant disease.
In the presence of postmenopausal bleeding
and in the absence of uterine fibroids it has been
suggested that uterine artery and endometrial
Doppler are sensitive in the differentiation of
significant pathology from endometrial atrophy.
Fig. 13.17 Attenuation by fibroid prevents any colour When no pathology of the endometrium is present
signal. then resistance index values are 0.85 ± 0.08,
whereas if there is pathology present these are
0.77 ± 0.03. Furthermore these authors indicate
differentiate a subserous fibroid from an that, in their experience, malignancy has not been
extrauterine mass the demonstration of a found in patients where the resistance index was
marginal vessel is said to be a useful discrimi- recorded as greater than 0.83, perhaps allowing
nating sign (Fig. 13.18). a more conservative approach in this group.21
There is some evidence that vascular fibroids The greater sensitivity of depiction of vessel
respond better to medical suppressive therapy distribution with modern colour Doppler and
and this may be useful in their management. power Doppler affords the demonstration of flow
There is, however, no evidence that malignant patterns within the endometrium. In carcinoma
tumours of the myometrium demonstrate any 81% demonstrate a multiple vessel pattern, 6%
specific Doppler ultrasound features.21 a single vessel pattern and 12% a scattered pattern
whereas 97% of polyps demonstrate a single vessel
Endometrial disease pattern and none a multiple pattern (Fig. 13.21).
Endometrial hyperplasia (Fig. 3.19), endometrial The multiple vessel pattern has been quantified
polyps and endometrial carcinoma (Fig. 13.20) with an endometrial colour score.22 In practice,
can produce similar appearances on transvaginal the demonstration of a single pedicular vessel is
ultrasound. There is broadening and inhomo- of great value in differentiating endometrial polyp
geneity of the thickness of the endometrium. from other pathology.When combined with saline
hysterography it demonstrates high sensitivity and
specificity for diagnosis of endometrial pathology.
Tamoxifen
Tamoxifen is widely used in the treatment of breast
carcinoma and has an oestrogenic effect on the
endometrium.This produces significant endome-
trial thickening and there is an associated increase
in the incidence of endometrial carcinoma.
There is a concomitant increase in blood flow
within the uterine arteries and within the myome-
trial vessels, which may be demonstrated on colour
Fig. 13.18 Transabdominal ultrasound demonstrates Doppler and by a decrease in the blood flow
flow in normal uterine arteries (red) and a junctional indices. However, this is not specific for malignancy
vessel between the uterus and the fibroid said to be and there are no features which will currently
characteristic of fibroid disease. allow a specific diagnosis (Fig. 13.22).23 299
13
b
c Fig. 13.19 (a) Transvaginal

ultrasound demonstrating
endometrial hyperplasia.
(b) Same patient. Colour flow
Doppler demonstrating
endometrial vascularity.
(c) Spectral Doppler within the
vessels in the endometrium
shows resistance indices less
than 0.77.
Hormone replacement therapy

Hormone replacement therapy produces no signifi-
cant change in blood flow characteristics to the
normal postmenopausal uterus.24 However, in
the presence of uterine fibroids there may be
increased blood flow with particular increase in
the diastolic component in patients on replacement
therapy.
Trophoblastic disease
The normal appearance of the postpartum uterus
Fig. 13.20 Transabdominal ultrasound of the uterus is variable, with myometrial heterogeneity and a
with abnormal flow demonstrating gross hyperaemia wide range of appearances of the endometrial
extending to the surface of the uterus in an cavity. The diagnosis of retained products of
endometrial carcinoma with extensive myometrial conception should not therefore rely on ultra-
300 invasion. sound appearances although the technique may
a
13
Applications
b
Fig. 13.21 Endometrial polyp. (a) Transvaginal sonography demonstrates an echogenic lesion within the uterine
cavity. (b) On colour flow Doppler there is a single vessel identifying the pedicle.
a Fig. 13.22 (a) Spectral Doppler

of uterine artery demonstrating
increased uterine blood flow in
patient on tamoxifen. (b) The
return of uterine blood flow to
normal values after withdrawal
of tamoxifen.
301
13
inform the diagnosis. Similarly the vascularity of Hydatidiform mole, invasive mole and chorio-
the postpartum uterus is also variable with areas carcinoma are rare neoplasms of the endometrium.
of enhanced vascularity some even exhibiting a Thickening and inhomogeneity of the endometrium
pedicle25 (Figs 13.23 and 13.24). Ultrasound are characteristic ultrasound features with a
including colour Doppler however may improve varying degree of formation of vesicles within the
decision making in those patients with post- uterine cavity. Ultrasound imaging is relatively
partum bleeding who have not responded to poor at distinguishing the three levels of severity
normal conservative management. of the disease and, unless there is evidence of
The demonstration of abnormal vessels within distant metastases, it is unreliable at the assess-
the endometrium extending into the endometrium ment of degree of invasion. There is a significant
usually following miscarriage or termination of increase in myometrial and endometrial blood
pregnancy was thought to be associated with a flow as evidenced by colour Doppler in all of the
diagnosis of arteriovenous malformation. It is now trophoblastic tumours; this is usually more marked
recognised that, although not common, this feature in the more aggressive tumours. Similarly, Doppler
can be associated with a rather prolonged invo- indices are altered in trophoblastic disease: a mean
lution of the trophoblast and supplying vessels. peak systolic velocity of 57.5 ± 20.4 (normal
In the absence of major symptoms ultrasound 28.3 ± 3. 41) and a resistance index of 0.56 ± 0.19
may be used to monitor resolution (Fig. 13.25). (normal 0.86 ± 0.05) have been demonstrated,
a Fig. 13.23 Retained products of conception:

(a) demonstrates residual trophoblast within the
uterine cavity; (b) colour flow (depicted in black and
white) demonstrates disordered pattern while
spectral Doppler yields a low resistance pattern.
302
13
Applications
Criteria for malignancy include the size and
complexity of lesions. However, not all malignant
lesions demonstrate characteristic ultrasound
features. Furthermore, if ultrasound is to have a
role in screening for ovarian carcinoma, then
features must be sought that would allow the early
identification of a potentially malignant lesion
prior to the development of frank malignant
morphology.
Considerable research work has suggested that
malignant tumours demonstrate higher diastolic
flow and consequently lower resistance and
Fig. 13.24 Colour flow Doppler of flow in trophoblastic
pulsatility indices than those seen in benign
tumour.
tumours. Similarly, the pattern of colour flow in
benign and malignant tumours is reportedly different
(Figs 13.26–13.31). However, the reported accu-
racy of colour flow and spectral Doppler is the
subject of wide variability within the literature.
Chou et al27 and Sengoku et al28 suggest a sensi-
tivity of 100% with a negative predictive value of
100%, while Bromley et al29 and Brown et al30 record
sensitivities of only 66% and 50%, respectively.
The uncertainty about the value of Doppler is
perhaps best reflected in the variation of criteria
used by different authors to indicate the presence
Fig. 13.25 Colour flow Doppler demonstrating or absence of malignancy. The cut-off values for
extensive vascular abnormality extending from the the resistance index range between 0.4 and 0.6,
endometrium deep within the myometrium. This was while others simply use a visual assessment of the
an arteriovenous malformation which required colour Doppler pattern. It is important therefore
embolisation but similar patterns may be seen in to compare the results of Doppler findings with
some cases of trophoblastic involution, particularly
the reported sensitivity of ovarian morphology
following termination of pregnancy.
alone in predicting malignancy, which is in the
region of 91–98%.29,31
although the resistance index may vary from 0.2 More recently there has been renewed interest
to 0.8. Nonetheless, the extent of intratumoral in the assessment of the pattern of distribution
flow correlates with the prognosis: the higher the of blood flow32 and a reassessment of the Doppler
resistance index, the lesser the need for prolonged spectrum.33 This work has concentrated on assess-
treatment cycles.26 Furthermore, the return of the ment of the diastolic component of the spectrum
Doppler indices to normal values is good evidence and the development of a new parameter
of successful surgical or medical treatment. assessing the pattern of flow decay, the end
diastolic velocity distribution slope. In addition
Ovarian disease there has been early work to assess the value of
Ovarian cancer contrast agents to refine the evaluation of the
Detection and characterisation of ovarian pattern of flow distribution within the ovary.34
tumours has depended upon the demonstration Although all of these studies are encouraging, to
of ovarian enlargement and the identification of date there has been no validation by larger
both cystic and solid masses within the ovary. studies. 303
13
Fig. 13.26 Solid ovarian tumour
with colour flow Doppler at the
margin and spectral Doppler
demonstrating low resistance.
Fig. 13.27 Solid ovarian cancer. Power Doppler

demonstrates disordered pattern of flow with
irregular branching vessels, anastomoses and a large
draining vein.
Studies attempting to combine a morpho- Fig. 13.28 Colour flow Doppler at the margin of an
logical score with Doppler features and indices apparently simple cyst but which subsequently
suggest that although diagnostic sensitivities may proved to have borderline malignancy detectable in
the wall.
be unaltered, there may be significant improvement
in positive predictive value from 60% to 94%.35
The situation in premenopausal patients is more
complex still.The presence of low-resistance flow Carter et al37 concluded that colour Doppler
in a corpus luteum requires that any suspicious lesion has neither the sensitivity nor the specificity to
be examined in the early proliferative phase of the distinguish between benign and malignant
menstrual cycle to ensure that the neovascularity disease of the ovary. This may no longer be
is not a physiological response (Fig. 13.32). tenable; studies utilising pattern distribution,
It is not surprising that the early results have new indices, 3D colour power and contrast
not been borne out by subsequent studies. The enhancement all promise to improve the
process of angioneogenesis is a microscopic one assessment of malignant disease. It remains the
and is at least partly dependent on cell type and case, however that the pattern of blood flow
304 morphology.36 within a mass is a single parameter and the
Fig. 13.29 Simple ovarian cyst with

13
Applications
spectral Doppler recorded in the
cyst wall with relatively low
resistance index of 0.64.
a b
c Fig. 13.30 Heterogeneous

ovarian mass with cystic and
solid elements. Flow is
demonstrated in colour within
the solid component (a), and
within the septum (b), although
spectral Doppler, while showing
low resistance, still exhibits a
resistance index of greater than
0.5.
evaluation of ovarian tumours with Doppler The diagnosis of ovarian dermoid is usually
techniques requires very significant further work straightforward on ultrasound.There are a number
if it is to provide a clinically useful tool for wide of cardinal features, the most reliable of which is
application. the echogenic plug. It has been suggested that 305
13
Fig. 13.31 Spectral Doppler of a
postmenopausal ovary showing no
diastolic flow. Some artefactual flow
is demonstrated in both systole and
diastole due to the small size of the
vessels sampled.
at present there is an assumption rather than

proof that this does more than just affect the
‘lead time’ to the diagnosis of ovarian cancer.
Furthermore, the case for cost effectiveness is
far from proven. Screening of first-order relatives
of those with ovarian cancer may be justified,
and in these patients the addition of colour flow
and spectral Doppler may contribute to the security
of a negative screening result. Doppler may also
be of value in the management of patients who
yield masses which appear morphologically benign
on transvaginal sonography.38 There is, however,
a diagnostic dilemma when a morphologically
normal ovary, or a morphologically benign simple
cyst, demonstrates neovascularity. Based on existing
studies, about 6% of women screened with
ultrasound will have cystic or complex masses.
Fig. 13.32 Colour flow around a complex mass in the
ovary of a premenopausal patient. This represented Moreover, in one study an ovarian artery resistance
flow in a corpus luteum. index of <0.4 occurred in 12% of premenopausal
patients and 3% of postmenopausal. In this
study, only one borderline ovarian tumour and
the benign nature of a dermoid can be predicted one endometrial malignancy were discovered.39
by the nature of blood flow and similar indices The scientific basis for ovarian screening in a
to those described for ovarian cancer (Fig. low-risk population is therefore unproven and even
13.33). However this lacks further confirmatory in patients with a strong family history there is a
evidence and there has been at least one case high false-positive detection rate, particularly in
where a struma ovarii has mimicked a malignant premenopausal patients, using both cancer antigen
pattern. 125 and ultrasound. It seems unlikely that Doppler,
using current methods of data processing, will
Screening for ovarian cancer further improve on the specificity of the screening
Although it is clear that transvaginal ultrasound can technique; it could in fact increase the number
306 identify asymptomatic masses postmenopausally, of false-positive laparotomies if abnormal colour
a
Fig. 13.33 Ovarian dermoid. (a) Colour flow

13
Applications
demonstrates marginal flow around the dermoid
plug. (b) Spectral analysis demonstrates low
resistance but still out with malignant range.
flow is to be used as another indicator of ovarian features remain dependent upon clinical history
abnormality.40 and other findings. Adenomyosis of the uterus
may exhibit focal or diffuse areas of increased
Other adnexal lesions echogenicity within the myometrium. The
Inflammatory processes may produce a hyper- normal vertical alignment of the spiral arteries
aemic response, although this is not invariable, may be lost (Fig. 13.36).
and will not allow differentiation of lesions of
differing aetiology.41 Findings in pelvic inflam-
matory disease are variable. There may be a
diffuse increase in the number and size of vessels
in the adnexa (Fig. 13.34). Tubo-ovarian abscess
is the pathology most likely to show evidence of
increased Doppler flow with circumferential
vessels draped around the cystic component and
low resistance indices, often less than 0.5.42
Endometriomas demonstrate a wide variety of
appearances on colour Doppler, sometimes with
circumferential vessels, occasionally ‘spotty’
vessels with resistance indices varying from 0.5 Fig. 13.34 Colour flow Doppler showing adnexal
to 0.7443 (Fig. 13.35). Differential diagnostic hyperaemia in pelvic inflammatory disease. 307
13
surgery.Those patients with demonstrable blood
flow within the ovary would be submitted for laparo-
scopic untorsion, while in those with an absence
of flow, open surgery and oophorectomy would
be the likely procedure, while those with normal
flow patterns could be managed conservatively.
Complications of early pregnancy

The trophoblast is extremely vascular during its
Fig. 13.35 Irregular, somewhat spotty flow in development, with two phases of invasion. The
endometriosis. vessels contributing to this invasion have thin
walls and low peripheral resistance to blood flow.
Sampling within the vascular component of the
trophoblast will therefore produce characteristic
low-impedance signals with resistance indices
less than 0.6. Some workers have suggested that
this may contribute to the early differentiation of
viable from non-viable or complicated pregnancies
(Fig. 13.37a). However, the high power output
of Doppler, particularly in relation to trans-
vaginal applications, must be considered and, in
this author’s view, this prohibits its use in most
early pregnancy applications. Ectopic pregnancy
Fig. 13.36 Adenomyosis of the uterus demonstrating
may be considered an exception to this rule and
unusual short segment myometrial vessels aligned to
the serosal surface.
certain valuable data can be acquired.
Ovarian torsion Ectopic pregnancy

Ultrasound imaging occupies a central role in
Ovarian torsion may produce a variety of the investigation of ectopic pregnancy, with recorded
appearances on ultrasound including a complex sensitivities of 94% for transvaginal ultrasound.47
cystic mass, a solid mass, and a solid mass with Modern methods of management now suggest
peripherally placed cysts. The absence of colour that a conservative approach can be adopted in
signal will not distinguish this from other patholo- those patients without active trophoblast and
gies, since colour Doppler has been demonstrated that when there is active trophoblast, medical
both centrally and peripherally in torsion.44 treatment using either systemic or local methotrexate
There may be reduction or absence of diastolic is the preferred first-treatment option. Using colour
flow and a reduced distribution of intraovarian Doppler as a guide, the vascular trophoblast can
vessels. A particular pattern of torsion described be identified within the adnexal region. Sampling
as the whirlpool sign in which the vessels are the flow within the trophoblast will yield pulsatile
wrapped around the ovary appears to be specific flow of low resistance index in the presence of
although it occurs in only a minority of cases.45 active trophoblast but a resistance index in excess
Where there is clinical suspicion of ovarian torsion of 0.8 in inactive trophoblast.48 This may provide
Doppler studies are reported to be abnormal in all a useful guide for conservative, non-operative
of the cases where torsion is confirmed and in only treatment in these patients (Fig. 13.37b).
2% where torsion was not confirmed.46 Preoperative Ancillary Doppler findings in ectopic pregnancy
assessment with colour Doppler may therefore are similar to those in a normotopic pregnancy,
308 be used to determine the nature and extent of with flow in uterine vessels and to the corpus
a
13
Applications
b
Fig. 13.37 (a) Normal colour flow pattern in the trophoblast of a normal 7-week intrauterine pregnancy.
(This patient subsequently had a termination of the pregnancy.) (b) Vascular trophoblast in the adnexal region.
Colour and spectral Doppler demonstrating low-resistance flow, implying active trophoblast and suggesting the
need either for administration of methotrexate as medical treatment or operative intervention.
luteum displaying comparable impedance indices, noma or pelvic metastatic disease, including ovarian
although peak velocities within the uterine arteries carcinoma. Certain metastases may show increased
are reduced. The reported findings do not, vascularity similar to that of primary ovarian
however, appear sufficiently specific to allow tumours, but there are no data to relate vascu-
differential diagnosis.49 larity to susceptibility for certain forms of treat-
ment, nor to the likely response to treatment
Cervical ectopic pregnancy (Fig. 13.38). Similarly certain cell types in cervical
Although rare, this form of ectopic pregnancy cancer are more likely to demonstrate abundant
presents particular difficulties with management. blood flow but this is of no current clinical value.
Attempts at evacuation of the uterus are usually There is no correlation with stage or degree of
unsuccessful, accompanied by heavy bleeding local invasion.50
and often necessitating hysterectomy. More recently,
direct installation of potassium chloride and Pelvic congestion syndrome
subsequently methotrexate into the gestational The association between chronic pelvic pain,
sac can result in regression of the gestation and dyspareunia and pelvic varices has been termed
either complete resolution, or a more safely pelvic congestion syndrome. Its aetiology is not
performed uterine evacuation. Colour and duplex clear but some have ascribed it as being due to
Doppler ultrasound have two roles in this situation: dilatation of pelvic veins, congestion of the ovaries
first, to identify the position of the uterine arteries with resultant ovarian swelling and cyst formation
joining the uterus, thus confirming the diagnosis (Fig. 13.39). In its most marked form there may
of a cervical pregnancy by its position caudal to be vulval and leg varices. Some have suggested a
the uterine arteries; second, Doppler is used to psychological component for the condition but
monitor the regression of the trophoblast by alter- it is difficult to assess a particular psychological
ation of the pattern of the Doppler signal with profile as cause or effect in women with long-term
time, as this usually mirrors the fall in human disabling pelvic pain. There is an association
chorionic gonadotrophin values. between the symptoms and multiple large
serpiginous pelvic veins, usually of diameter
Other gynaecological tumours greater than 4 mm and, in this author’s expe-
Duplex Doppler ultrasound has no established rience, flow velocities within the pelvic veins of
role in the assessment either of cervical carci- less than 5 cm s-1. Reverse flow within the pelvic 309
13
Fig. 13.38 Echogenic ovarian
metastasis in patient with primary
breast carcinoma demonstrating
high-velocity peripheral flow which
is normally indicative of shunting.
a b
Fig. 13.39 (a) Large collection of serpiginous vessels in the pelvis adjacent to the right ovary which also
contains a cyst. (b) Extension of the distended veins towards the uterine cervix and vaginal vault.
veins during a Valsalva manoeuvre, which exceeds Retroperitoneal mass lesions including
2 cm s-1 peak velocity, appears to be a prominent lymphadenopathy
feature of the syndrome. Further work is
necessary, however, to define this association The relationship of retroperitoneal masses to the
more clearly. iliac vessels can usually be established without
Occasionally it is helpful to use colour the use of colour flow Doppler.
Doppler to differentiate distended pelvic veins
from hydrosalpinx. The latter will demonstrate Iliac artery aneurysm
no flow even at low velocity settings (Fig. Common external and internal iliac artery
13.40). aneurysms can represent a differential diagnostic
finding of a pelvic cystic mass. Common and
Other pathologies external iliac artery aneurysms are usually easily
Non-gynaecological pathologies can be encoun- identified by their relationship to the main arterial
tered during pelvic and transvaginal ultrasound trunk, although internal iliac artery aneurysms
310 examination. may require colour flow Doppler for confirmation
13
Summary
contrast within the tubes and spill into the
peritoneum in similar fashion to that seen with
orthodox hysterosalpingography, although the
correlation between hysterosalpingography and
HyCoSy remains less than 90%. The technique
has therefore not fully replaced formal X-ray
hysterosalpingograms.The use of colour Doppler
in conjunction with ultrasound contrast improves
the sensitivity of the technique in the detection
of tubal spill.36
Fig. 13.40 Hydrosalpinx. Colour flow demonstrates
flow in adjacent adnexal structures but not in the
tube. SUMMARY
Colour and duplex Doppler have yet to find well-
of their nature. Usually a pulsatile but swirling defined applications in gynaecology. The value
colour signal can be observed within the true in ovarian tumour diagnosis is controversial and
lumen with a variable amount of mural clot. in tumours of the endometrium unproven. It
may improve diagnostic security in patients with
Deep venous thrombosis morphologically benign ovarian tumours, but
Transvaginal ultrasound is not the primary hopes that it would enhance the process of
method of diagnosis of deep vein thrombosis but ovarian screening have not been fulfilled. In
the pelvic extension of clot can be demonstrated inflammatory pathology there are no reported
by both transabdominal and transvaginal ultra- features that will allow the distinction between
sound, with distension of the iliac veins by different pathological processes, even torsion of
intraluminal echogenic clot and either distortion the ovary demonstrating intraovarian flow in
of flow pattern around the clot or complete many cases.
occlusion. In the assessment and management of subfer-
tility, Doppler techniques appear to have signifi-
Bowel masses cant value. Ovarian activity and uterine receptivity
Bowel wall thickening is a non-specific feature of can be more accurately monitored than by
bowel pathologies. Similarly, mural hypervascu- ultrasound morphology alone. Cyclical changes
larity is a feature typical of inflammatory bowel can be documented and the effects of ovulation
pathologies including Crohn’s disease, diverticular induction therapy defined.
disease and infective colitides; the degree of In ectopic pregnancy it may have a role in the
vascularity reflects disease activity. operative planning process, while in malignant
trophoblastic disease it appears to be of value
Hysterosalpingo-contrast sonography in determining prognosis and in treatment
(HyCoSy) planning.
The use of echo-enhancing agents in the assess- There is a need for considerable research
ment of tubal patency has been shown to be of particularly in the areas of neovascularity and
value in the investigation of infertility. A fine signal processing before Doppler techniques are
cannula is inserted into the cervical canal under applicable to common gynaecological conditions,
direct vision and between 5 and 10 cc of contrast and before this author could advocate its wide-
are injected gently whilst performing a trans- spread adoption in the investigation of gynae-
vaginal scan. This should show the echogenic cological conditions.
311
13
REFERENCES
1. Jarvela IY, Sladkevicius P, Kelly S, et al. Three of embryo implantation in an ICSI programme.
dimensional sonographic and power Doppler J Assist Reprod Genet 2000; 17 (6):329–334.
characterisation of ovaries in late follicular phase. 15. Schild RL, Knobloch C, Dorn C, et al. Endometrial
Ultrasound Obstet Gynecol 2002; 20:281–285. receptivity in an in vitro fertilisation programme as
2. Suren A, Osmers R, Kulenkampff D, et al. assessed by spiral artery flow, endometrial
Visualization of blood flow in small ovarian tumour thickness, endometrial volume and uterine artery
vessels by transvaginal colour Doppler sonography blood flow. Fertil Steril 2001; 25(2):361–366.
after echo enhancement with injection of Levovist. 16. Jarvela IY, Sladkevicius P, Kelly S, et al. Quantification
Gynaecol Obstet Invest 1994; 38: 210–212. of ovarian power Doppler signal with three
3. Raine-Fenning NJ, Campbell BK, Kendall NR, et al. dimensional ultrasonography to predict response
Quantifying the changes in endometrial vascularity during in vitro fertilization. Obstet Gynecol 2003;
throughout the normal menstrual cycle with 102:816–822.
three dimensional power Doppler angiography. 17. Frates MC, Doubilet PM, Durfee SM, et al.
Hum Reprod 2004; 19 (2):330–338. Sonographic and Doppler characteristics of the
4. Sladkevicius P, Valentin L, Marsal K. Transvaginal corpus luteum: can they predict pregnancy
gray-scale and Doppler ultrasound examinations of outcome? J Ultrasound Med 2001; 20 (8) 821–827.
the uterus and ovaries in healthy postmenopausal 18. Yang JH, Wu MY, Chen CD. Association of
women. Ultrasound Obstet Gynecol 1995; 6: endometrial blood flow as determined by a
81–90. modified colour Doppler technique with subsequent
5. Steer CV, Tan SL, Mason BA, et al. Mid-luteal phase outcome of in vitro fertilization. Hum Reprod 1999;
vaginal color Doppler assessment of uterine artery 14(6):1606–1610.
impedance in a subfertile population. Fertil Steril 19. Sladkevicius P, Campbell S. Advances in
1994; 61: 53–58. ultrasound assessment in the establishment and
6. Kurjak A, Zalud I. Doppler and color flow imaging. development of pregnancy. Br Med Bull 2000;
In: Nyberg DA, Hill LM, Bohm-Velez M, et al, eds. 56(3):691–703.
Transvaginal ultrasound. St Louis: Mosby Year Book; 20. Wiener Z, Beck D, Rottem S, et al. Uterine artery
1992:285–294. flow velocity waveforms and color flow imaging in
7. Meire HB. Doppler – artefacts, errors and pitfalls. women with perimenopausal and postmenopausal
Abdom Gen Ultrasound 1993; 5: 83–93. bleeding. Correlation to endometrial histopathology.
8. Battaglia C, Artini PG, D’Ambrogio G, et al. Uterine Acta Obstet Gynecol Scand 1993; 72:162–166.
and ovarian blood flow measurement. Does the full 21. Carter JR, Lau M, Saltzman AK, et al. Gray-scale
bladder modify the flow resistance? Acta Obstet and color flow Doppler characterisation of uterine
Gynecol Scand 1994; 73: 716–718. tumors. J Ultrasound Med 1994; 13:835–840.
9. Kamaya A, Tuthill T, Rubin JM. Twinkling artifact on 22. Alcazar JL, Castillo G, Minguez JA, et al.
colour Doppler sonography: dependence on Endometrial blood flow mapping using transvaginal
machine parameters and underlying cause. AJR Am power Doppler sonography in women with post
J Roentgenol 2003; 180 (2):215–222. menopausal bleeding and thickened endometrium
10. Ziereisen F, Heinrichs C, Dufour D, et al. The role Ultrasound Obstet Gynecol 2003; 21(6):583–588.
of Doppler evaluation of the uterine artery in 23. Tepper R, Cohen I, Altaras M, et al. Doppler flow
girls around puberty. Pediatr Radiol 2001; 31 evaluation of pathologic endometrial conditions in
(10):712–719. postmenopausal breast cancer patients treated
11. Frajndlich R, von Eye Corleta H, Frantz N. Color with Tamoxifen. J Ultrasound Med 1994;
Doppler sonographic study of the uterine artery in 13:635–640.
patients using intrauterine contraceptive devices. 24. Van den Bosch T, Van Schoubroeck D, Lu C, et al.
J Ultrasound Med 2000; 19 (8):577–579. Color Doppler and gray scale evaluation of the post
12. Pepper J, Dewart PJ, Oyesanya OA. Altered uterine partum uterus. Ultrasound Obstet Gynecol 2002;
artery blood flow impedance after danazol therapy: 20 (6):586–591.
possible mode of action in dysfunctional uterine 25. Zalud I, Conway C, Schulman H, et al. Endometrial
bleeding. Fertil Steril 1999; 72(1):66–70. and myometrial thickness and uterine blood flow in
13. Steer CV, Tan SL, Dillon D, et al. Vaginal color postmenopausal women: the influence of hormonal
Doppler assessment of uterine artery impedance replacement therapy and age. J Ultrasound Med
correlates with immunohistochemical markers of 1993; 12:737–741.
endometrial receptivity required for the implantation 26. Zhou Q, Lei XY, Cardoza JD. Sonographic and
of an embryo. Fertil Steril 1995; 63:101–108. Doppler imaging in the diagnosis and treatment
14. Contart P, Baruffi RL, Coelho J, et al. Power Doppler of gestational trophoblastic disease; a 12 year
312 endometrial evaluation as a method for the prognosis experience. J Ultrasound Med 2005; 24(1):15–24.
27. Hsieh FJ, Wu CC, Chen CA, et al. Correlation of
39. Muto MG, Cramer DW, Brown DL. Screening for

13
References
uterine hemodynamics with chemotherapy ovarian cancer: the preliminary experience of a
response in gestational trophoblastic tumors. familial ovarian cancer center. Gynecol Oncol 1993;
Obstet Gynecol 1994; 83:1021–1025. 51:12–20.
28. Chou CY, Chang CH, Yao BL, et al. Color Doppler 40. Karlan BY, Raffel LJ, Crvenkovic G, et al.
ultrasonography and serum CA125 in the A multidisciplinary approach to the early detection
differentiation of benign and malignant ovarian of ovarian carcinoma: rationale, protocol design,
tumors. J Clin Ultrasound 1994; 22:491–496. and early results. Am J Obstet Gynecol 1993;
29. Sengoku K, Satoh T, Saitoh S, et al. Evaluation of 16:494–501.
transvaginal colour Doppler sonography, 41. Quillin SP, Siegel MJ. Transabdominal color Doppler
transvaginal sonography and CA125 for prediction ultrasonography of the painful adolescent ovary.
of ovarian malignancy. Int J Gynaecol Obstet 1994; J Ultrasound Med 1994; 13:549–555.
46:39–43. 42. Tinkanen H, Kujansuu E. (Doppler ultrasound
30. Bromley B, Goodman H, Benacerraf BR. findings in tubo-ovarian infectious complex).
Comparison between sonographic morphology and J Clin Ultrasound 1993; 21:175–178.
Doppler waveform for the diagnosis of ovarian 43. Aleem F, Pennisi J, Zeitoun K, et al. The role of
malignancy. Obstet Gynecol 1994; 83:434–437. colour Doppler in diagnosis of endometriomas.
31. Brown DL, Frates MC, Laing FC, et al. Ovarian Ultrasound Obstet Gynecol 1995; 5:51–54.
masses: can benign and malignant lesions be 44. Stark JE, Siegel MJ. Ovarian torsion in prepubertal
differentiated with color and pulsed Doppler US? and pubertal girls: sonographic findings. AJR Am J
Radiology 1994; 190:333–336. Roentgenol1994; 163:1479–1482.
32. Itakura T, Kikkawa F, Kajiyama H, et al. Doppler flow 45. Vijayaraghavan SB. Sonographic whirlpool sign in
and arterial location in ovarian tumours. Int J ovarian torsion. J Ultrasound Med 2004;
Gynecol Obstet 2003; 83(3):277–283. 23(12):1653–1659.
33. Shaharabany Y, Akselrod S, Tepper R. A sensitive 46. Ben-Ami M, Perlitz Y, Haddad S. The effectiveness
new indicator for diagnostics of ovarian malignancy, of spectral and colour Doppler in predicting ovarian
based on the Doppler velocity spectrum. torsion. A prospective study. Eur J Obstet Gynecol
Ultrasound Med Biol 2004; 30(3):295–302. Reprod Biol 2002; 104(1):64–66.
34. D’Arcy TJ, Jayaram V, Lynch M, et al. Ovarian cancer 47 de Crespigny LC. Demonstration of ectopic
detected non-invasively by contrast-enhanced pregnancy by transvaginal ultrasound. Br J Obstet
power Doppler ultrasound. Br J Obstet Gynaecol Gynecol 1998; 95:1253–1256.
2004; 111(6):619–622. 48. Tekay A, Jouppila P. Color Doppler flow as an
35. Stein SM, Laifer-Narin S, Johnson MB, et al. indicator of trophoblastic activity in tubal
Differentiation of benign and malignant adnexal pregnancies detected by transvaginal ultrasound.
masses: relative value of gray–scale, color Doppler, Obstet Gynecol 1992; 80:995–999.
and spectral Doppler sonography. AJR Am J 49. Jurkovic D, Bourne TH, Jauniaux E, et al.
Roentgenol 1995; 164:381–386. Transvaginal color Doppler study of blood flow
36. Timor-Tritsch LE, Lerner JP, Monteagudo A, et al. in ectopic pregnancies. Fertil Steril 1992;
Transvaginal ultrasonographic characterisation of 57:68–73.
ovarian masses by means of color flow-directed 50. Acazar JL, Castillo G, Jurado M, et al. Intratumoral
Doppler measurements and a morphologic scoring blood flow in cervical cancer as assessed by
system. Am J Obstet Gynecol 1993; 168:909–913. transvaginal color Doppler ultrasonography:
37. Wu CC, Lee CN, Chen TM, et al. Incremental Correlation with tumor characteristics. Int J Gynecol
angiogenesis assessed by color Doppler ultrasound Cancer 2003; 13(4):510–514.
in the tumorigenesis of ovarian neoplasms. 51. Shlief R, Deichert U. Hysterosalpingo-contrast
Cancer 1994; 73:1251–1256. sonography of the uterus and fallopian tubes:
38. Carter JR, Lau M, Fowler JM, et al. Blood flow results of a clinical trial of a new contrast medium
characteristics of ovarian tumors: implications for in 120 patients. Radiology 1991; 178:213–215.
ovarian cancer screening. Am J Obstet Gynecol
1995; 172:901–907.
313
14
Clinical applications
of Doppler ultrasound
in obstetrics
Imogen Montague and Paul A. Dubbins
The circulatory changes that occur during preg- until 22 weeks, the invasion extends as far as the
nancy involve modification of vascular structure spiral arteries. This invasion of the spiral arteries
within the uterus (spiral arteries), the development affects the resistance to blood flow within the
of a neocirculation (the placenta and the fetus), spiral arteries and thereby in the arcuate and
and a redistribution of blood flow and alteration main uterine arteries.1
in circulating blood volume such that the placenta
in the third trimester receives 20% of the total Method of examination
maternal circulation and maternal blood volume The complexity of the uteroplacental circulation
increases by a similar value. Certain disease makes accurate identification of the vessel under
processes and certain complications of pregnancy study difficult with either continuous wave or duplex
are at least in part mediated by a microvascular Doppler ultrasound. Flow velocity waveforms are
abnormality. Thus, for example, impaired obtained from the lateral lower quadrants of the
trophoblast migration of the spiral arteries is a uterus, angling the transducer on either side of
major component in pre-eclampsia. As a result the uterus towards the cervix. Signals achieved in
there has been considerable interest in the appli- this way are assumed to be originating from the
cation of Doppler techniques to the detection of uterine arteries (see Fig. 13.2).2,3 The uterine
complications of pregnancy, detection and arteries are more accurately identified using colour
characterisation of certain fetal abnormalities, as Doppler: the region lateral to the lower uterus is
well as an assessment of the value of Doppler in the examined and the external iliac artery and the
detection and management of maternal disease. adjacent vein are identified. The uterine artery
crosses the external iliac artery on its course from
the internal iliac artery to the body of the uterus.
THE UTEROPLACENTAL
In this way more accurate identification of the
CIRCULATION
particular vessel being investigated is achieved.
The uterine artery is a branch of the anterior It is important to angle the transducer to improve
division of the internal iliac artery, and divides the angle of insonation whilst maintaining vessel
further into four arcuate arteries, each of which identification on colour Doppler. Spectral wave-
divides into more than 25 spiral arteries. There forms are obtained by placing the pulsed Doppler
are therefore between 100 and 200 spiral arteries range gate within the vessel at this point (Fig.
which enter the intervillus space. During early 14.1a).4 The spectral waveform from the normal
pregnancy, trophoblast cells invade this space uteroplacental system is unidirectional, of low
and disrupt the wall of the spiral arteries as part pulsatility, and demonstrates frequencies throughout
of the process of placental formation. There are the cardiac cycle (Fig. 14.1b). This is a result of
two separate waves of invasion. Between implan- the trophoblastic invasion of the spiral arteries;
tation and 10 weeks, the trophoblastic invasion is end-diastolic frequencies increase to a maximum
limited to the decidual layer. From about 14 weeks at 24–26 weeks of gestation (see Fig. 14.7).5 315
14
Clinical applications of Doppler ultrasound in obstetrics
a It is possible to calculate several indices to
quantify waveform analysis; however, the most
commonly used in the uterine artery is the
resistance index (RI) (Fig. 14.7a). After 26 weeks
of gestation the normal range of the resistance
index is between 0.45 and 0.58. Decreased end-
diastolic flow and consequently raised resistance
indices above 0.58 is considered abnormal, as is
a notch in early diastole in either uterine artery,
suggesting failure of trophoblastic invasion of
the spiral arteries (Fig. 14.2).6,7
Problems and pitfalls

b A comparative histological study of third-trimester
placental biopsy at caesarean section8 has demon-
strated that the uterine artery flow impedance
reflects impaired trophoblast migration. None-
theless, there are a number of reasons for caution.
Impedance to flow varies throughout the utero-
placental circulation, with the lowest value seen
in the arcuate vessels on the placental side of the
uterus and the highest value seen in the uterine
arteries on the non-placental side of the uterus.9
The physiological variations and anatomical
complexities of the uteroplacental vascular tree
make it difficult to obtain accurate and repro-
ducible measurements using continuous wave
Doppler, with interobserver variations ranging
Fig. 14.1 (a) Course of the left uterine artery as it
from 3.9 to 17%.10,11 In later pregnancy, between
crosses over the iliac vessels. (b) Normal uterine
37 and 40 weeks, maternal position may also
artery flow in the early third trimester demonstrating
high diastolic flow.
alter flow patterns, with the umbilical artery
Fig. 14.2 Uterine artery waveform

in intrauterine growth retardation
demonstrating low diastolic flow
and an early diastolic notch.
316
14
The uteroplacental circulation

resistance being higher in the supine position flow.14,15 Most of the antihypertensive drugs
than in the decubitus.12 Furthermore, variations appear to have no effect on fetomaternal blood
in uterine activity (Fig. 14.3), maternal heart rate flow.16 However, nifedipine appears to produce a
and exercise also significantly alter the wave- reduction in umbilical artery resistance, and this
form.13,14 The examination should therefore be has therefore been suggested as a better drug to
performed with the mother resting and only use during pregnancy.17 The effect of smoking
during a period of uterine inactivity. The effect on blood flow has been somewhat controversial
of exercise is more marked in complicated preg- with researchers reporting either no effect or a
nancies, but there are no data evaluating the significant effect from smoking a cigarette.18,19
effect of exercise to improve sensitivity or specificity However, because of the dispute and the potential
of Doppler in predicting fetal outcome.The time for cigarette smoking to alter flow patterns in
of day, or recent eating by the mother, do not chronic smokers, these authors advocate that
appear to have an effect on the uterine artery patients should not smoke for at least 1 h prior
to a Doppler study.
a
Pathology
Although superficially attractive, the use of uterine
artery Doppler even in primiparous mothers where
the risk of hypertensive complications is greater
is neither sensitive nor specific in the detection
of pre-eclampsia or small-for-gestational-age
infants.20 This is typical for results seen in low-
and moderate-risk pregnancies. The routine use
of uterine artery Doppler studies to screen low
risk populations for subsequent development of
pre-eclampsia and intrauterine growth retardation
(IUGR) has a poor positive predictive value.
Furthermore, screening can only be useful if there
b is an intervention available which will improve
the outcome, and in obstetrics treatment options
for both conditions are limited. Early delivery can
only be considered if the pregnancy is sufficiently
mature. At present there is no proven interven-
tion to prevent either pre-eclampsia or IUGR.
However in a high-risk population the presence
of a diastolic notch beyond 24 weeks of gestation
may indicate the need for more intensive maternal
and fetal surveillance to allow early identification
of IUGR or pre-eclamptic toxaemia.This screening
technique could potentially be used to rationalise
patient care pathways dependant upon their
Fig. 14.3 (a) Uterine artery waveform in labour.
uterine artery Doppler studies.
Blood flow characteristics in the uterine artery between
contractions is normal. (b) During a contraction the
overall flow velocity within the uterine artery is Coexistent maternal disease
reduced. However, the diastolic velocity is reduced Uterine artery Doppler appears to be of signifi-
further than the peak systolic velocity and there is an cantly greater utility when there is pre-existing
early systolic notch. maternal disease. In chronic renal disease, for 317
14
example, an abnormal artery waveform predicts
Table 14.1 Vessels examined by Doppler
pre-eclampsia and IUGR with a high degree of
accuracy. Only 8% of patients with negative Doppler Maternal Uterine arteries
Arcuate arteries
findings in one study developed complications Placental vessels
of pregnancy.21 By contrast, although the resist- Fetal Umbilical
ance index is increased in the uterine arteries of Aorta
Middle cerebral artery
diabetics with a morphological vasculopathy, Carotid
there is no relationship with short- or long-term Renal
Ductus venosus
diabetic control and it is not a good predictor of
diabetes-related fetal morbidity, presumably
because these changes are reflecting the risk of
acidosis as a result of hypoxia rather than meta- study suggests that an abnormal uterine artery
bolic acidosis.22,23 In patients with pre-existing, Doppler will identify all those pregnancies with
essential hypertension, uterine artery Doppler an adverse outcome.24
appears to be useful in defining groups of patients
who are at risk of developing complications. If the
DOPPLER EXAMINATION OF THE
systolic blood pressure is greater than 140 mmHg,
FETOPLACENTAL CIRCULATION
then resistance indices in both uterine arteries is
(Table 14.1)
increased. If the systolic blood pressure is less than
140 mmHg, three separate groups may be iden- The placenta rather than the lungs is the organ
tified: those with (a) bilateral or (b) unilateral of gaseous exchange in the fetus. Two umbilical
abnormalities of the waveform within the uterine arteries convey deoxygenated blood from the fetus
arteries; and (c) those with an entirely normal to the placenta, and one umbilical vein returns
uterine artery flow. The prognosis appears to be oxygenated blood to the fetal inferior vena cava.
related to the degree of abnormality of uterine artery The umbilical arteries take origin from the fetal
flow. In systemic lupus erythematosus, one internal iliac arteries coursing alongside the lateral
Fig. 14.4 Normal umbilical artery

flow in the third trimester. In the
left hand side of the trace
coexistent umbilical venous flow is
demonstrated. The resistance index
is 0.53 and the pulsatility index
0.81.
318
14
Doppler examination of the fetoplacental circulation

walls of the bladder in the urachus to the umbilical Colour flow Doppler confers some advantages
insertion. The umbilical vein courses posteriorly on the evaluation of the umbilical artery.
and cephalad to join the left branch of the portal In patients with oligohydramnios it may be
vein. The oxygenated blood is then shunted the only way to identify a loop of cord. Further-
through the liver to the inferior vena cava by the more, in multiple pregnancies it allows the identifi-
ductus venosus. Spectral waveforms from the cation of the individual umbilical arteries.
umbilical artery and vein can usually be easily Ultrasound imaging and colour Doppler allow
obtained using continuous wave Doppler ultra- the optimisation of the angle of insonation to the
sound;25 a simple Doppler probe is placed on umbilical vessels and therefore improve clarity
the abdomen and the beam randomly directed of signal reproduction. However, as the Doppler
into the uterine cavity until the characteristic indices relate systolic to diastolic components of
arterial and venous waveforms are demonstrated flow, optimal angle of incidence is less important.
(Fig. 14.4). Duplex Doppler allows identification Volume flow assessment within the fetoplacental
of a loop of cord and the recording of an unam- circulation was much in vogue in the early 1980s
biguous signal from the umbilical vessels. but has not found widespread clinical utility and
a Fig. 14.5 (a) Relative oligohydramnios. Colour flow

Doppler demonstrating umbilical arteries in the
pelvis on either side of an empty bladder. (b) Relative
oligohydramnios sampling from the umbilical artery
within the pelvis demonstrates normal umbilical
artery flow patterns.
319
14
is not performed in most departments. Early index is more accurate for situations where there
suggestions that different signals were achieved is low, absent or reversed end-diastolic flow.
at the placental end of the cord compared to the Variations in fetal heart rate26 and the presence
umbilical end have not been confirmed, and most of fetal breathing movement27 may significantly
studies relating abnormal umbilical artery Doppler alter the arterial waveforms, although within the
studies to prenatal outcome relate to sampling physiological range of 120–160 beats per minute
of a free loop of umbilical cord. Clinical use of it is not necessary to correct indices for fetal heart
umbilical artery Doppler studies must therefore rate28 (Fig. 14.6). Nonetheless, it is important to
reflect research results and where possible signals examine umbilical artery waveforms during a
should be recorded from a free loop. However in period of fetal inactivity and in the absence of
situations where there is oligohydramnios, and fetal breathing; duplex and colour Doppler ultra-
particularly in twin pregnancies with a ‘stuck sound are of value in enabling the waveform to
twin’, sampling flow from the umbilical arteries be sampled rapidly and accurately. Measurement
within the fetal abdomen as they course alongside of three consecutive cardiac cycles reduces the
the bladder may be the only option to achieve coefficient of variation of measurements to less
satisfactory Doppler signals (Fig. 14.5). than 5%.29
Doppler waveforms within the umbilical artery All Doppler devices have inherent filtration
change with gestational age in a similar fashion which removes low-frequency noise and vessel wall
to those of the uterine artery (Fig. 14.6). Resist- movement artefact. On most current machines the
ance and pulsatility indices demonstrate a gradual high-pass filter can be varied, but it is important
reduction with increase in gestational age. not to set this higher than 100 Hz and preferably
Normal values of the pulsatility index are shown lower than 50 Hz, if a false impression of absent
in Figure 14.7. diastolic flow is not to be created.
Problems and pitfalls Pathology and applications

As end-diastolic values fall, the A/B ratio (peak (Tables 14.2–14.4)
systole/end diastole) and resistance index tend Intrauterine growth retardation
towards a value of 1. For clinical purposes resistance Normal waveforms from the umbilical artery are
indices are therefore not sensitive to the quan- unidirectional and demonstrate forward flow
tification of reversed flow, and the pulsatility throughout the cardiac cycle. Decreased end-
Fig. 14.6 Umbilical artery

waveform at 26 weeks gestation.
Note the PI of 1.39 which is just
above the mean value for this
gestation.
320
14

(a) 0.9 (b) 1.8
Umbilical artery pulsatility index

Uterine artery resistance index 1.6
0.8
1.4
0.7
1.2
0.6
1.0
0.5
0.8
0.4 0.6
0.3 0.4
16 20 24 28 32 36 40 16 20 24 28 32 36 40
Gestation (weeks) Gestation (weeks)
(c) 5.5 (d) 2.50

5.0
Fetal CCA pulsatility index
2.25
MCA pulsatility index

4.5 95th centile
4.0 2.00
3.5 1.75
3.0 Mean
2.5 1.50
2.0 1.25
1.5 5th centile
1.00
1.0
0.5 0.75
20 22 24 26 28 30 32 34 36 38 40 16 20 24 28 32 36 40
Gestational age (weeks) Gestation (weeks)
(e) 2.8 (f) 4.0

95th centile
Fetal RA pulsatility index
Thoracic aorta pulsatility index
2.6 3.5
2.4 3.0
Mean
2.2 2.5
2.0 2.0 5th centile
1.8 1.5
1.6 1.0
1.4 0.5
16 20 24 28 32 36 40 20 22 24 26 28 30 32 34 36 38 40
Gestation (weeks) Gestational age (weeks)
Fig. 14.7 Normal values during pregnancy for (a) uterine artery resistance index (RI), and (b) umbilical artery
pulsatility index (PI). Fetal common carotid artery (CCA) PI (c) falls steeply after 32 weeks gestation, mirrored by
the middle cerebral artery (MCA) PI (d). Normal values are also shown for the fetal descending thoracic aorta PI
(e) and renal artery (RA) PI (f). Reproduced with permission from Pearce41.
321
14
Table 14.2 Indications for Doppler examination in pregnancy
Uterine artery Umbilical artery Middle cerebral artery

? Screening Intrauterine growth retardation Intrauterine growth retardation
Pre-existing disease, e.g. renal disease, Post dates
systemic lupus erythematosus
Fetal abnormality
Pre-existing uterine disease
Trophoblastic disease
Table 14.3 Features of abnormality

Raised indices Raised indices Decreased indices
Notch Absent end-diastolic flow
Reversed end-diastolic flow Ratio of middle cerebral artery to umbilical artery ≤1
Table 14.4 Factors contributing to abnormal flow

Pre-existing disease Fetal activity Direct cerebral compression
Exercise Breathing
Uterine activity Cord compression
?Antihypertensives ?Cigarettes
?Cigarettes
Fig. 14.8 In intrauterine growth retardation there is Fig. 14.9 Severe intrauterine growth retardation with
322 reduction of end-diastolic flow. absent end-diastolic flow in the umbilical artery.
Fig. 14.10 Reversed diastolic flow in the umbilical

14

artery. This implies a fetus at risk of significant
morbidity/mortality.
a Fig. 14.11 (a) Normal pattern of

flow in the uterine artery on the
placental side. (b) Low diastolic
flow and persistence of the
diastolic notch in the non-placental
uterine artery in a patient with
previous intrauterine growth
retardation.
323
14
diastolic flow (Fig. 14.8) and consequently raised artery waveforms appear to carry no greater risk
Doppler indices are considered abnormal, and of fetal morbidity or fetal loss than a normal
are thought to reflect increased placental resistance pregnancy.The corrected perinatal mortality rate
caused by damage to placental tertiary villi.30 In is reduced when the results of umbilical artery
more extreme cases, end-diastolic flow may be Doppler are made available to clinicians, who
absent (Fig. 14.9) or even reversed (Fig. 14.10). are then able to act with more appropriate and
Reduction in end-diastolic flow velocities within timely intervention.33 The finding of absent or
the umbilical artery waveform is thought to occur reversed end-diastolic flow in a pregnancy with
as a result of reduced tertiary villi formation and established IUGR is an indication for the serious
therefore could be an indicator of placental consideration of early delivery,34,35 although there
dysfunction, IUGR and fetal distress. In spite of are as yet no precise data indicating the temporal
the attraction of the hypothesis that Doppler relationship between these end-diastolic flow
would provide an early indicator of failure of the changes and a subsequent poor outcome. There
tertiary villus, it has not proved of value as a is some evidence that in pregnancies with previous
primary screening tool in low risk pregnancies IUGR, assessment of uterine artery flow will allow
for IUGR. However, in high-risk pregnancies, the identification of a group of women at high
particularly those in which there is pregnancy- risk of recurrence of third trimester complications.
induced hypertension, or which have other clinical Uterine artery impedance remains high in this
or sonographic evidence of fetal growth retar- group, usually on the non-placental side and there
dation, the umbilical artery waveform is a good is often an early diastolic notch. Although of poten-
indicator of fetal compromise. Progressive reduc- tial value in this group there is no established
tion in the diastolic component of umbilical artery role for uterine Doppler in population screening
flow mirrors the risk and severity of potential (Fig. 14.11).
fetal compromise.31,32 Furthermore not only may
fetuses at increased risk be identified and managed Post dates pregnancy
more intensively, but high-risk pregnancies, such The management of a post dates pregnancy is
as those with established IUGR or pregnancy- extremely difficult. No single parameter has been
induced hypertension, with normal umbilical established that will confidently predict outcome
Fig. 14.12 Umbilical Doppler in

twin pregnancy.
324
Fig. 14.13 (a) Colour flow Doppler of the circle of

14

a
Willis in the fetus. The middle cerebral arteries are
demonstrated. (b) Spectral flow within the normal
middle cerebral artery. Pulsatility index of 1.6 and
a peak velocity of 37 cm s-1. (c) Fetal anaemia
increase in peak systolic flow velocity to 72 cm s-1
is demonstrated.
325
14
and, although there has been some enthusiasm more fetus where preterm delivery of that fetus
for a role for Doppler in the assessment of post is indicated, yet can compromise well grown co-
dates pregnancy, there is not universal agreement fetuses, who can be iatrogenically damaged by
about its value.36 In these authors’ view, there- preterm delivery. Use of intra-abdominal or
fore, umbilical artery Doppler should be looked abdominal wall insertion umbilical artery Doppler
upon as one of the factors that may contribute to studies is sometimes necessary to be confident
the monitoring of post dates pregnancies. Absence of appropriate cord sampling.
or reversal of diastolic flow is an indication for
immediate delivery, and in this situation an opera- Diabetes
tive delivery may need to be considered. Some Both pre-existing vascular disease and hypertensive
authors have advocated the use of the ratio between disorders of pregnancy are common in mothers
the Doppler shifts (A/B ratio) in the uterine artery with diabetes. In these cases, the value of an
and the fetal middle cerebral artery (UA/MCA abnormal umbilical artery Doppler signal has
ratio) in post-term pregnancy, suggesting that values the same significance in identification of utero-
of <1.05 yield sensitivities of 80% and specificities placental insufficiency as in the non-diabetic
of 90% in predicting an adverse outcome but population. However, diabetic pregnancies are also
this has not been independently confirmed.37 at risk of metabolic complications and Doppler
flow patterns will not detect these complications,
Twin and higher-order pregnancy and it is therefore vital that a normal umbilical
In uncomplicated multiple pregnancy, there is artery flow does not give either the clinician or
the same progressive decrease in placental resist- the mother false reassurance.22
ance as is seen in singleton pregnancies, with a
consequent fall in indices with increasing gesta- Fetal anaemia
tional age. Doppler appears to be of no value in An exciting more recent application of middle
predicting adverse outcomes in unselected preg- cerebral artery Doppler studies is in the field of
nancies.38 However, in pregnancies with discordant suspected fetal anaemia due to maternal alloim-
growth patterns, Doppler analysis of the umbilical munisation. The pregnant woman develops an
arteries appears to be a useful adjunct to serial immune response to paternally derived red-cell
growth measurements, allowing recognition of antigen that is foreign to the mother and inherited
high-risk twin pregnancies which require more by the fetus. Antibodies may cross the placenta,
intense surveillance39 (Fig. 14.12). bind to the fetal red blood cells, causing haemolysis,
In twin-to-twin transfusion syndrome, the shared leading to fetal anaemia.The physiological response
circulation has the result that Doppler wave- to chronic anaemia is the development of a
forms in the umbilical arteries of both twins are hyperdynamic fetal circulation. Initial studies
similar. Doppler studies of these vessels appear investigating a potential ultrasound predictor of
to be of no predictive or management value in this fetal anaemia looked at many Doppler40–42 and
condition, although colour and power Doppler other parameters such as fetal spleen size.43 These
may allow documentation of placental anastomoses early studies failed to identify good correlation
and consequent guidance of laser ablation. between ultrasound findings and the presence of
In triplet and higher order pregnancies Doppler fetal anaemia. However, a link between peak
studies of the umbilical vessels are valuable for velocities of the middle cerebral artery and fetal
the serial surveillance of fetal well-being. Discordant anaemia was demonstrated44 (Fig. 14.13). Subse-
growth with early placental failure becomes more quent studies have shown this technique to be
common (in one or more of the fetuses) as the total reproducible,45 and now it has, to a large degree,
number of fetuses of a given pregnancy increases. replaced invasive fetal assessment by amniotic
Clinical challenges arise with the management fluid spectroscopy or cordocentesis, both of which
326 of severe preterm fetal compromise in one or carried risks including further sensitisation leading
14
The fetal circulation

to exacerbation of the underlying pathology, and the presence of early onset growth retardation,
pregnancy loss. Use of this technique has accu- (both symmetrical and asymmetrical) which has
rately predicted anaemic fetuses allowing the up to 20% association with abnormal karyotype.
timing of any invasive testing (if appropriate) to In the presence of a raised alpha-fetoprotein
be at the point when fetal transfusions are clinically but structurally normal fetus, the appearance of
required. Middle cerebral artery Doppler peak an abnormal uteroplacental Doppler waveform
velocities increase with gestational age.45 at 20 weeks is associated with a significantly
raised perinatal mortality,47 with the commonest
Fetal abnormality contributor to perinatal mortality being placental
Fetuses with autosomal trisomy may also have abruption. Similarly it has been suggested that
abnormal placentation with reduced tertiary villi fetal and placental morphological features should
formation.46 An abnormal umbilical artery wave- be combined with Doppler to improve prediction
form should therefore prompt a detailed study of outcome.48
of fetal structural anatomy and an assessment of Colour flow Doppler is of value in demon-
amniotic fluid volume; karyotyping should also strating abnormalities of the cord including true
be considered.34 This is particularly important in knots and a single umbilical artery (Fig.14.14).
Similarly the use of colour Doppler will aid in the
diagnosis of abdominal wall abnormalities such
a as omphaloceles and ectopia vesicae (Fig.14.15).
THE FETAL CIRCULATION

The advent of colour and power Doppler has meant
that it is now possible to visualise flow within
many of the fetal vessels.These include the aorta,
inferior vena cava, carotid, intracerebral and renal
arteries. Furthermore, it is possible to study
cardiac and cardiopulmonary haemodynamics.
Fig. 14.14 (a) Colour flow Doppler through the pelvis

of a fetus with a single umbilical artery. (b) Single Fig. 14.15 Omphalocele colour flow Doppler
umbilical artery in the pelvis flanked by two femoral demonstrates abnormal vessels within the anterior
arteries. abdominal wall defect. 327
14
The detailed study of congenital cardiac abnor- ment. The long axis of the vessel is identified
malities and the role of Doppler techniques are, and the probe is angled to improve the angle of
however, beyond the scope of this chapter. insonation. The sample volume is placed above
There is a redistribution of fetal blood flow in the vessel, close to the root of the neck, in order
response to hypoxia with a selective increase in to get a waveform from the common carotid artery
blood flow to the brain, heart and adrenal glands and to avoid the external carotid artery.
at the expense of the viscera. This redistribution The middle cerebral artery is the most commonly
reflects the morphological findings in IUGR, interrogated of the intracerebral vessels.The skull
with the head continuing to grow at the expense is scanned as if to perform a biparietal diameter
of a relatively smaller abdomen. Similarly there measurement. A colour Doppler examination is
is also reduced renal perfusion leading to reduced then performed in a plane slightly closer to the
fetal urine production and hence reduction in base of the skull, where the middle cerebral artery
amniotic fluid volume in the presence of chronic will be identified as a vessel coursing towards the
fetal compromise. probe from the circle of Willis in the Sylvian fissure
Fetal outcome is thought to relate both to the (Fig. 14.13a). Later in gestation the location of
severity and duration of hypoxia, as well as the the middle cerebral artery can be identified without
gestational age at delivery. It is assumed that hypoxia colour Doppler simply by demonstrating pulsa-
will tend to increase with increasing gestational tions within the brain substance. Placement of
age in fetuses with growth retardation and that this the sample volume either within these pulsations
will lead to a greater redistribution of blood flow. or within the colour signals will allow recording
of the Doppler waveforms.
The descending thoracic aorta The common carotid artery pulsatility index
The descending thoracic aorta was the subject remains constant until 32 weeks of gestation and
of significant early research interest since it was then falls steeply (Fig. 14.7c)50–52 and this is
thought to afford comparison between the visceral mirrored by changes seen in the middle cerebral
blood supply and cerebral perfusion.49 Although artery (Fig. 14.7d).53
the waveform has characteristic features and The mean pulsatility index of the middle cere-
pulsatility and resistance indices are constant bral artery is 2 (SD 0.48) at 20 weeks falling to
throughout gestation50–52, the haemodynamics 1.15 (SD 0.18) at 37 weeks. Fetal haemodynamics
of the fetal aorta are complex. Flow within the suggest a brain sparing effect in IUGR and
aorta will reflect flow and impedance not only evidence of cerebral redistribution by a decrease
the visceral vessels and the lower limbs but more in the pulsatility index in the presence of estab-
especially the umbilical arteries.Thus, while fetal lished growth retardation is commonly considered
hypoxia may produce alterations in the Doppler an indicator for delivery.54,55 The presence of
waveform of the thoracic aorta, much of the abnormal Doppler studies can also aid in the
information derived from flow redistribution postnatal management of growth restricted neonates
can also be obtained from examining the who are at greater risk of ischaemic compli-
umbilical artery waveform. This is generally cations such as necrotising enterocolitis due to
easier to obtain and, consequently in current their intrauterine adaptations to a suboptimal
routine clinical practice, aortic Doppler studies environment.56,57 Clinicians vary as to at what
are very rarely used. stage in the progress of the growth retardation at
any given gestation the fetus should be delivered,
The fetal cerebral circulation given the absence of a clear temporal relationship
The common carotid artery can usually be identi- between an abnormal pulsatility index reduction
fied on colour Doppler within the neck. It is a and incidence of ischaemic complications.58
relatively straight vessel with a long course and Data suggesting decompensation of the cerebral
328 is therefore amenable to pulsed Doppler assess- circulation and a consequent increase in the
Fig. 14.16 Middle cerebral artery

14

flow in severe fetal compromise.
There is marked reduction of diastolic
flow as evidence of cerebral
redistribution. The features imply
ischaemic complication.
pulsatility index are scanty but this may represent large adrenal vessels may be misidentified as kidney
evidence of irreversible ischaemic change (Fig. (Fig. 14.17).
14.16). In spite of variation in clinical practice,
immediate delivery versus delay does not appear The ductus venosus
to impact on neonatal or long-term outcomes. The ductus venosus is an embryological channel
It is suggested that the variation in practice which connects the fetal umbilical vein with the
simply reflects appropriate clinical response to a inferior vena cava and hence the right heart. In
wide variety of clinical and laboratory data of the fetus it carries most of the blood from the
which middle cerebral artery Doppler is only umbilical vein to the right atrium.
one. The ductus venosus is identified within the
liver by following the umbilical vein in a sagittal
The renal arteries plane into the fetal liver using colour Doppler.
Oligohydramnios consequent upon poor renal The ductus venosus arises at the junction of the
perfusion is cardinal sign of IUGR. Although umbilical vein and the right branch of the portal
Doppler evaluation of the renal arteries is tech- vein angling posteriorly and cephalad towards the
nically feasible and Doppler spectra can be recorded inferior vena cava with which it makes an angle
(Fig. 14.7f)59 the fetal kidneys only receive 3%
of the fetal cardiac output and the renal artery
velocities are low. As a result, demonstration of
variation of flow and impedance indices within
the renal arteries is an insensitive and unreliable
indicator of fetal compromise. It is not clear
whether increased sensitivity of ultrasound
equipment, particularly to low flow states, will
alter this.
Colour Doppler of the renal vessels may also
be used to assess congenital renal abnormalities,
such as renal agenesis, where the artery is also
absent, although care should be taken as hypoplastic Fig. 14.17 Coronal colour flow Doppler image of the
arteries may be difficult to demonstrate on ultra- fetus. There are two renal arteries, one of which arises
sound, and enlarged adrenal glands supplied by from the iliac artery. 329
14
of 45 to 60°. Colour flow will demonstrate a high
velocity pattern with aliasing at low pulse repetition
frequency (Fig. 14.18).
Normal Doppler blood flow studies of the
ductus venosus show a triphasic forward flowing
cycle reflecting ventricular systole, ventricular
diastole and atrial systole (Fig. 14.19). Spectral
analysis can therefore be used to reflect fetal
ventricular function and cardiac afterload (Fig.
14.20). In the presence of end-stage hypoxaemia
with hypoxic cardiomyopathy, there is a fall in
cardiac function, and hence rise in central venous
pressure. This can be identified by reverse flow
in the ductus venosus. Presence of reverse flow
is associated with decreased fetal maternal perfu-
sion and fetal death.60,61 In severely growth restricted
fetuses where the gestational age is close to
viability, use of this technique can help guide the
clinician in the exact moment at which delivery Fig. 14.19 Normal spectral flow within the ductus
is required. venosus.
In the presence of a thoracic space occupying
lesion, ductus venosus studies can aid in assessing
the degree of cardiac restriction. Reduction in of reported incidences of low-lying placenta in
flow velocity and reversal of flow may occur. excess of 40%62 when scanning is performed before
the last trimester; routine screening for placenta
Placental abnormalities praevia in the second trimester is therefore not
The diagnosis of placenta praevia is largely now widely used. By contrast, when there is
performed by ultrasound imaging.Transabdominal antepartum haemorrhage, ultrasound is used to
ultrasound has fallen into some disrepute because establish the relationship of the placenta to the
internal os.This is more commonly now performed
using transvaginal ultrasound. The addition of
colour Doppler allows the demonstration of the
rare but potentially serious vasa praevia. In this
situation a vessel, or vessels, can be seen coursing
from the margin of the placenta in close proximity
to, or covering, the internal os. Similarly, other
placental abnormalities can be documented: vela-
mentous insertion of the cord can be identified
with colour Doppler and chorioangioma can be
differentiated from a placental haematoma by
the demonstration of a rich vascular supply.63
Fig. 14.18 Oblique scan through the fetal abdomen Postpartum

with colour flow Doppler demonstrating the aliased A small amount of postpartum haemorrhage is a
flow within the ductus venosus. The Doppler sample normal feature but where haemorrhage is
volume is placed within the area of maximum prolonged, or heavy, the possibility of retained
330 aliasing. products of conception must be considered.
Fig. 14.20 Intermittent flow

14

reversal within the ductus venosus
in a cardiac dysrhythmia.
Transvaginal ultrasound will frequently demon- There is a difference in the blood flow charac-
strate abnormalities within the uterine cavity. As teristics of the uterine artery when there are
a general rule brightly reflective structures are retained products of conception. Patients with
taken to represent retained placental fragments residual trophoblast exhibit a resistance index in
and/or membrane; while echo-poor contents are the myometrial vessels of 0.35 ± 0.1 whereas those
thought to represent fresh or clotted blood. without residual trophoblast have resistance
However, it is frequently difficult to differentiate indices in the myometrial vessels of 0.54 ± 0.15.
between retained products of conception and However, the combination of low-impedance
uncomplicated blood clot. flow and intrauterine material is common after
Fig. 14.21 Spectral Doppler in

retained products of conception.
The resistance index of 0.48 implies
residual trophoblast.
331
14
abortion and does not necessarily imply retained is no doubt that volume flow is increased in preg-
products of conception. It may simply reflect nancy but this appears to be mediated largely by
physiological involution; the temporal course of vasodilatation of the large supply vessels, although
return to non-pregnant flow characteristics in the absolute changes in flow velocity have not
the uterine vessels has not been documented64 been investigated.There is no change in pulsatility
(Fig. 14.21). or resistance indices in patients with essential
hypertension.70 There has been no documented
change in these indices in pregnancy induced
TROPHOBLASTIC DISEASE
hypertension. However there is significant increase
Transvaginal ultrasound will demonstrate uterine in acceleration time within the intrarenal vessels.
abnormalities in persistent gestational trophoblastic This is suggestive of proximal vasospasm as being
tumour. However, the sensitivity of ultrasound at least partly responsible for the pathogenesis of
imaging is only 70% but abnormal uterine artery pregnancy induced hypertension.71,72
waveforms are seen in 90% of cases with persistent There is no change in Doppler indices in patients
gestational trophoblastic tumour, and raised presenting with progressive physiological dilatation
uterine artery impedance may predict resistance of the collecting system of the kidneys during
to chemotherapy.65,66 However, magnetic reso- pregnancy. In the past this has been attributed to
nance imaging (MRI) appears to be more sensi- a combination of a hormonal effect, together with
tive in the detection of trophoblastic tumour, a degree of mechanical obstruction of the ureter
being more accurate at identifying myometrial by the enlarging uterus. There is no correlation
invasion and therefore establishing a diagnosis between the degree of dilatation of the collecting
of choriocarcinoma. system and any change in the resistance index in
uncomplicated cases, but it may be clinically useful
in cases of suspected obstruction where Doppler
MATERNAL HAEMODYNAMICS
ultrasound may show a significant difference
Pregnancy is associated with marked changes in between the affected and unaffected kidney. The
maternal haemodynamics. There is an increase obstructed kidney will usually show a resistance
in circulating blood volume with a corresponding index differing by 0.04 or more than that in the
increase in the cardiac output. This might be unobstructed kidney. In these authors’ department,
expected to have effects on blood flow to a number resistance indices are used as a discriminator in
of the intra-abdominal organs, in addition to the deciding which patients with loin pain in asso-
effect that it has on uterine blood flow. Further- ciation with pregnancy should proceed to an
more, it might be hoped that certain conditions, intravenous urogram73 (see Ch. 8).
particularly associated with pregnancy and
thought to be mediated through an abnormality Haemodynamic changes in other
in maternal physiology, might be reflected in vessels
changes in blood flow characteristics detectable There is naturally a decrease in the resistance index
by Doppler. in the aorta and common iliac arteries which
reflects the increased flow to the uterus that occurs
Doppler ultrasound of the maternal during pregnancy. However, there are also changes
kidney in pregnancy in the flow patterns in other vessels consequent
Doppler blood flow characteristics to the maternal upon changes in volume flow.There is, for example,
kidney appear to alter little during pregnancy. an increase in volume flow through the portal
Although certain authors report a slight reduction vein and consequently in the hepatic veins. In the
in the mean resistance index in renal artery exami- hepatic veins this affects not only the volume of
nations, this is not statistically significant. There flow but also the Doppler waveform where the
332
14
Conclusions
normal pulsatile flow is lost in most patients. the use of screening Doppler affect perinatal
This occurs as early as the first trimester in some morbidity and outcome.77 Doppler ultrasound
patients and therefore does not appear to be related cannot be recommended as a screening proce-
to pressure effects from the enlarging uterus. dure for the identification of pregnancy compli-
This is of limited diagnostic significance except cations in low-risk pregnancies. Doppler exami-
that some authors have suggested that changes in nation of the uterine and umbilical arteries between
hepatic venous flow might be useful in the assess- 18 and 26 weeks of gestation in high-risk preg-
ment of diffuse liver disease, including pregnancy- nancies may be predictive of significant pregnancy
related liver disease. However, there is no corre- complications, but at present there is insufficient
lation between liver disease in pregnancy and evidence to suggest that the level of surveillance
changes in hepatic venous flow, whatever the in the Doppler-negative group can be relaxed. If
severity.74 a relatively safe, cheap and effective treatment
for early-onset maternal placental disease becomes
established, such as antioxidant therapy, then
CONCLUSIONS
uteroplacental waveform screening may prove
The role of Doppler in obstetrics is becoming a useful in deciding upon early treatment.
mature technology.The advent of colour Doppler Doppler has most potential in the manage-
has enabled more precise examination of the ment of high-risk pregnancies such as those with
uteroplacental and fetoplacental circulations, pre-existing maternal disease, previously identified
together with more intricate examination of the placental disease and those at high risk of fetal
fetal vasculature. However, this is an expensive IUGR or where IUGR is already established.
modality and its benefits beyond a research tool Umbilical artery studies help identify fetuses at
have yet to be established. Combined analysis of risk of hypoxia and acidaemia; when the results
the uteroplacental and fetoplacental (umbilical) are applied clinically they contribute to reduced
circulation may increase diagnostic accuracy.75 perinatal mortality rates. Doppler ultrasound of
However, fetomaternal haemodynamics are the fetal circulation may identify chronic hypoxia
complex, showing variation in their response to and there is evidence that documentation of
many maternal and fetal physiological and redistribution of flow is important in the iden-
pathological states. It is possible, even likely, that tification of fetal compromise (Tables 14.3 and
other factors will influence fetal maternal blood 14.4). Doppler ultrasound has significantly
flow. Currently Doppler methods appear capable impacted in the care of maternal alloimmunisation
only of detecting gross changes in placentation pregnancies at risk of severe fetal anaemia.
and gross changes of fetal well-being. For example, Middle cerebral artery peak velocity measurements
umbilical artery waveforms may be normal in a have now superceded routine invasive monitoring,
morphologically normal but small placenta. which is highly likely to lead to a reduction in
Similarly, chronic hypoxia induces placental changes fetal loss in this condition.
which are presumably responsible for the altered It is likely that the future lies in comparative
waveform in umbilical and fetal vessels. However, investigation of the uteroplacental umbilical and
in experimentally induced acute hypoxia there is one or more fetal vessels and the development of
no alteration in fetomaternal flow as detected by more sophisticated parameters of flow dynamics.
Doppler techniques.76 It is important, however, to stress that feto-
In the same way, the role of uterine artery maternal haemodynamics are not the only indi-
Doppler in the management of pregnancy is not cator of maternal or placental disease, or of fetal
clear cut. While apparently of value in patients well-being. Doppler ultrasound may contribute
with pre-existing maternal disease, its findings to, but not replace, other methods of fetal and
are not sufficiently sensitive or specific, nor does maternal surveillance.
333
14
REFERENCES
1. Pijnenborg R, Dixon G, Robertson WB, et al. 15. Hastie SJ, Howie CA, Whittle MJ, et al. Daily
Trophoblastic invasion of human decidua from 8 to variability of umbilical and lateral uterine wall artery
18 weeks of pregnancy. Placenta 1980; 1:3–19. blood velocity waveform measurements. Br J
2. Bewley S, Campbell S, Cooper D. Uteroplacental Obstet Gynaecol 1988; 95:571–574.
Doppler flow velocity waveforms in the second- 16. Duggan PM, McCowan LM, Stewart AW.
trimester. A complex circulation. Br J Obstet Antihypertensive drug effects on placental flow
Gynaecol 1989; 96:1040–1046. velocity wave forms in pregnant women with
3. Schulman H, Winter D, Farmakides G, et al. Doppler severe hypertension. Aust N Z J Obstet Gynaecol
examinations of the umbilical and uterine arteries 1992; 32:335–338.
during pregnancy. Clin Obstet Gynecol 1989; 17. Hirose S, Yamada A, Kasugai T, et al. The effect of
32:738–745. nifedipine and dipyridamole on the Doppler blood
4. Bower S, Vyas S, Campbell S, et al. Colour Doppler flow waveforms of umbilical and uterine arteries in
imaging of the uterine artery in pregnancy: normal hypertensive pregnant women. Asia Oceania J
ranges of impedance to blood flow, mean velocity Obstet Gynaecol 1992; 18:187–193.
and volume flow. Ultrasound Obstet Gynecol 1992; 18. Castro LC, Allen R, Ogunyemi D, et al. Cigarette
2:261–265. smoking during pregnancy: acute effects on uterine
5. Schulman H, Fleischer A, Farmakides G, et al. flow velocity waveform. Obstet Gynecol 1993;
Development of uterine artery compliance in 81:551–555.
pregnancy as detected by Doppler ultrasound. 19. Morrow RJ, Ritchie JW, Bull SB. Maternal cigarette
Am J Obstet Gynecol 1986; 155:1031–1036. smoking: the effects on umbilical and uterine blood
6. Fleischer A, Schulman H, Farmakides G, et al. flow velocity. Am J Obstet Gynecol 1998;
Uterine artery Doppler velocimetry in pregnant 159:1069–1071.
women with hypertension. Am J Obstet Gynecol 20. North RA, Ferrier C, Long D, et al. Uterine artery
1986; 154:806–813. flow velocity waveforms in the second trimester
7. Bower S, Schuchter K, Campbell S. Doppler for the prediction of pre-eclampsia and fetal growth
ultrasound screening as part of routine antenatal retardation. Obstet Gynecol 1994; 83:378–386.
screening: prediction of pre-eclampsia and growth 21. Ferrier C, North RA, Becker G, et al. Uterine artery
retardation. Br J Obstet Gynaecol 1993; waveform as a predictor of pregnancy outcome in
100:989–994. women with underlying renal disease. Clin Nephrol
8. Voigt HJ, Becker V. Doppler flow measurements 1994; 42:362–368.
and histomorphology of the placental bed in 22. Johnstone FD, Steel JM, Haddad NG, et al.
uteroplacental insufficiency. J Perinat Med 1992; Doppler umbilical artery flow velocity waveforms
20:139–147. in diabetic pregnancy. Br J Obstet Gynaecol 1992;
9. Kofinas AD, Espeland M, Swain M, et al. Correcting 99:135–140.
umbilical artery flow velocity waveforms for fetal 23. Ben-Ami M, Battino S, Geslevich Y, et al. A random
heart rate is unnecessary. Am J Obstet Gynecol single Doppler study of the umbilical artery in the
1989; 160:704–707. evaluation of pregnancies complicated by diabetes.
10. Rightmire DA, Campbell S. Fetal and maternal Am J Perinatol 1995; 12:437–438.
Doppler blood flow parameters in postterm 24. Kofinas AD, Penry M, Simon NV, et al.
pregnancies. Obstet Gynecol 1987; 69:891–894. Interrelationship and clinical significance of
11. Bewley S, Cooper D, Campbell S. Doppler increased resistance in the uterine arteries in
investigation of uteroplacental blood flow patients with hypertension or pre-eclampsia or
resistance in the second trimester: a screening both. Am J Obstet Gynecol 1992; 166:601–606.
study for pre-eclampsia and intrauterine growth 25. Fitzgerald DE, Drumm JE. Non-invasive measurement
retardation. Br J Obstet Gynaecol 1991; of human fetal circulation using ultrasound: a new
98:871–879. method. Br Med J 1977; 2:1450–1451.
12. Qu LR, Kan A, Masahiro N. Fetal circulation in 26. Giles WB, Trudinger BJ, Baird PJ. Fetal umbilical
relation to various maternal body positions. artery flow velocity waveforms and placental
Chung Hua Fu Chan Ko Tsa Chih 1994; 29:589–591. resistance: pathological correlation. Br J Obstet
13. Mulders LG, Jongsma HW, Wijn PF, et al. Gynaecol 1985; 92:31–38.
The uterine artery blood flow velocity waveform: 27. Mires G, Dempster J, Patel NB, Crawford JW.
reproducibility and results in normal pregnancy. The effect of fetal heart rate on umbilical artery
Early Hum Dev 1988; 17:55–70. flow velocity waveforms. Br J Obstet Gynaecol
14. Morrow R, Ritchie K. Doppler ultrasound fetal 1987; 94:665–669.
velocimetry and its role in obstetrics. Clin Perinatol 28. Gill RW, Trudinger BJ, Garrett WJ, et al. Fetal
334 1989; 16:771–778. umbilical venous flow measured in utero by pulsed
Doppler and B-mode ultrasound. Am J Obstet blood cell isoimmunisation. Obstet Gynecol 1995;
14
References
Gynecol 1980; 139:720–725. 85:122–128.
29. Kofinas AD, Penry M, Swain M, et al. The effect of 43. Bahodo-Singh R, Oz U, Jones D, et al. Fetal spleen
placental laterality on uterine artery resistance and size in anaemia due to Rh-alloimmunization. Obstet
development of pre-eclampsia and intrauterine Gynecol 1998; 92:828–832.
growth retardation. Am J Obstet Gynecol 1989; 44. Vyas S, Nicolaides KH, Campbell S. Doppler
161:1536–1539. examination of the middle cerebral artery in
30. Erskine RLA, Ritchie JWK. Umbilical artery blood anaemic fetuses. Am J Obstet Gynecol 1990;
flow characteristics in normal and growth-retarded 162:1066–1068.
fetuses. Br J Obstet Gynaecol 1985; 92:605–610. 45. Mari G. Noninvasive diagnosis by Doppler
31. McParland P. Modern approach to the poorly grown ultrasonography of fetal anemia due to maternal
fetus. Ir Med J 1992; 85:88–89. red-cell alloimmunization. N Engl J Med 2000;
32. Trudinger BJ, Cook CM, Giles WB, et al. Fetal 342:9–14.
umbilical artery velocity waveforms and subsequent 46. Rochelson B, Kaplan C, Guzman E, et al.
neonatal outcome. Br J Obstet Gynaecol 1991; A quantitative analysis of placental vasculature in
98:378–384. the third-trimester fetus with autosomal trisomy.
33. Neilson JP. Doppler ultrasound in high-risk Obstet Gynecol 1990; 75:59–63.
pregnancies. In: Enkin MW, Keirse MJNC, 47. Pearce JMF. The application of uteroplacental
Renfrew MJ, et al, eds. Pregnancy and childbirth waveforms to complicated pregnancies. In: Pearce
module. Cochrane Database of Systematic JMF, ed. Doppler ultrasound in perinatal medicine.
Reviews, No. 03889. London: BMJPublishing; Oxford: Oxford University Press; 1992:173–174.
1993. 48. Jauniaux E, Ramsay B, Campbell S.
34. Poulain P, Palaric JC, Paris-Liado J, et al. Fetal Ultrasonographic investigation of placental
umbilical Doppler in a population of 541 high-risk morphologic characteristics and size during the
pregnancies: prediction of perinatal mortality and second trimester of pregnancy. Am J Obstet
morbidity. Doppler Study Group. Eur J Obstet Gynecol 1994; 170:130–137.
Gynecol Reprod Biol 1994; 54:191–196. 49. European Association of Perinatal Medicine.
35. Devoe LD, Gardner P, Dean C, et al. The Regulation for the use of Doppler technology in
significance of increasing umbilical artery systolic- perinatal medicine. In: Consensus of Barcelona.
diastolic ratios in the third-trimester pregnancy. Barcelona: Instituto Barcelona; 1989:22–26.
Obstet Gynecol 1992; 80:684–687. 50. Bilardo CM, Campbell S, Nicolaides KH. Mean
36. Zimmermann P, Alback T, Koskinen J, et al. blood velocity and flow impedence in the fetal
Doppler flow velocimetry of the umbilical artery, descending thoracic aorta and common carotid
uteroplacental arteries and fetal middle cerebral artery in normal pregnancy. Early Hum Dev 1988;
artery in prolonged pregnancy. Ultrasound Obstet 18:213–217.
Gynecol 1995; 5:189–197. 51. Lingman G, Marsal K. Fetal central blood circulation
37. Devine PA, Bracero LA, Lysiliewicz A, et al. in the third trimester of normal pregnancy – a
Middle cerebral to umbilical artery Doppler ratio in longitudinal study; aortic and umbilical blood flow.
post date pregnancies. Obstet Gynecol 1994; Early Hum Dev 1986; 13:137–150.
84:856–860. 52. Griffin D, Bilardo K, Masini L, et al. Doppler blood
38. Faber R, Viehweg B, Burkhardt U. Predictive value flow waveforms in the descending thoracic aorta of
of Doppler ultrasound findings in twin pregnancies. the human fetus. Br J Obstet Gynaecol 1984;
Zentralbl Gynäkol 1995; 117:353–357. 91:997–1002.
39. Giles WB, Trudinger BJ, Cook CM, et al. Umbilical 53. Vyas S, Nicolaides KH, Bower S, et al. Middle
artery flow velocity waveforms and twin pregnancy cerebral artery flow velocity waveforms in fetal
outcome. Obstet Gynecol 1988; 72:894–897. hypoxaemia. Br J Obstet Gynaecol 1990;
40. Nickolaides KH, Fontanarosa M, Gabbe SG, et al. 97:797–803.
Failure of ultrasonographic parameters to predict 54. Rowlands DJ, Vyas SK. Longitudinal study of fetal
the severity of fetal anaemia in rhesus middle cerebral artery flow velocity waveforms
isoimmunization. Am J Obstet Gynecol 1988; preceding fetal death. [Comment in Br J Obstet
158:920–926. Gynaecol 103(8):852.] Br J Obstet Gynaecol 1995;
41. Copel JA, Grannum PA, Green JJ, et al. Pulsed 102:888–890.
Doppler flow-velocity waveforms in the predicition 55. Laurin J, Lingman G, Marsal K, et al. Fetal blood
of fetal hematocrit of the severely isoimmunized flow in pregnancies complicated by intrauterine
pregnancy. Am J Obstet Gynecol 1989; growth retardation. Obstet Gynecol 1987;
161:341–344. 69:895–902.
42. Hercher K, Snijders R, Campbell S, et al. Fetal 56. Hackett GA, Campbell S, Gamsu H, et al.
venous, arterial, and intracardiac blood flows in red Doppler studies in the growth-retarded fetus and
335
14
prediction of neonatal necrotising enterocolitis, 68. Dobkin GR, Berkowitz RS, Goldstein DP, et al.
haemorrhage, and neonatal morbidity. Br Med J Duplex ultrasonography for persistent gestational
1987; 294:13–16. trophoblastic tumor. J Reprod Med 1991;
57. Laurin J, Marsal K, Persson PH, et al. Ultrasound 36:14–16.
measurements of fetal blood flow in predicting 69. Long MG, Boultbee JE, Langley R, et al. Doppler
fetal outcome. Br J Obstet Gynaecol 1987; assessment of the uterine circulation and the
94:940–948. clinical behaviour of gestational trophoblastic
58. The GRIT trial study group. Infant wellbeing at tumours requiring chemotherapy. Br J Cancer 1992;
2 years of age in Growth Restriction Intervention 66:883–887.
Trial (GRIT): mulitcentred randomized controlled 70. Zeeman GG, McIntire DD, Twickler DM. Maternal
trail. Lancet 2004; 364:513–520. and fetal artery Doppler findings in women with
59. Vyas S, Nicolaides KH, Campbell S. Renal artery chronic hypertension who subsequently developed
flow velocity waveforms in normal and hypoxemic superimposed pre-eclampsia. J Matern Fetal
fetuses. Am J Obstet Gynecol 1989; 161:168–172. Neonat Med 2003; 14(5):318–323.
60. Gudmunsson S, Tulzer G, Huhta JC, et al. Venous 71. Myake H, Nakai A, Kshino T, et al. Doppler
Doppler velocimetry in fetuses with absent end- velocimetry of maternal renal circulation in
diastolic blood velocity in the umbilical artery. pregnancy induced hypertension. J Clin Ultrasound
J Matern Fetal Invest 1993; 3:196. 2001; 29(8) 449–455.
61. Baschat AA, Gembruch U. Triphasic umbilical 72. Nakai A, Asakura H, Oya A, et al. Pulsed Doppler
venous blood flow with prolonged intrauterine US findings of renal interlobar arteries in pregnancy
survival in intrauterine growth retardation. induced hypertension. Radiology 1999;
Ultrasound Obstet Gynecol 1996; 8:201–205. 213(2):423–428.
62. Marsal K, Lindblad A, Lingman G, et al. Blood flow 73. Weston MJ, Dubbins PA. The diagnosis of
in the descending aorta: intrinsic factors affecting obstruction: colour Doppler ultrasonography of renal
fetal blood flow, i.e. fetal breathing movements and blood flow and ureteric jets. Curr Opin Urol 1994;
cardiac arrhythmia. Ultrasound Med Biol 1984; 4:69–74.
10:339–348. 74. Roobottom CA, Hunter JD, Weston MJ, et al.
63. Vyas S, Campbell S, Bower S, et al. Maternal Hepatic venous Doppler waveforms: changes in
abdominal pressure alters fetal cerebral blood flow. pregnancy. J Clin Ultrasound 1995; 23:477–482.
Br J Obstet Gynaecol 1990; 97:740–742. 75. Campbell S, Hernandez CJ, Cohen-Overbeek T,
64. Scherjon SA, Kok JH, Oosting H, et al. Fetal and et al. Assessment of fetoplacental and
neonatal cerebral circulation: a pulsed Doppler uteroplacental blood flow using duplex pulsed
study. J Perinatol Med 1992; 20:79–82. Doppler ultrasound in complicated pregnancies.
65. Ott W. Placenta praevia. Ultrasound Obstet Gynecol J Perinat Med 1984; 12:261–265.
1993; 139:1493–1494. 76. Trudinger BJ, Giles WB, Cook CM. Flow velocity
66. Heinonen S, Ryynanen M, Kirkinen P, et al. Perinatal waveforms in the maternal uteroplacental and fetal
diagnostic evaluation of velamentous umbilical cord umbilical placental circulations. Am J Obstet
insertion: clinical, Doppler, and ultrasonic findings. Gynecol 1985; 152:155–163.
Obstet Gynecol 1996; 87:112–117. 77. Newnham JP, O’Dea MR, Reid KP, et al. Doppler
67. Dillon EH, Case CQ, Ramos IM, et al. Endovaginal flow velocity waveform analysis in high-risk
ultrasound and Doppler findings after first-trimester pregnancies: a randomised control trial. Br J Obstet
abortion. Radiology 1993; 186:87–91. Gynaecol 1991; 98:956–963.
336
15
Microbubble ultrasound
contrast agents
Jonathan D. Berry and Paul S. Sidhu
The intense echo enhancement observed with

microbubbles is a result of their compressibility.
BACKGROUND
A microbubble undergoes volumetric oscillation
The principle of ultrasound microbubble contrast while being insonated, to a greater degree than a
agents is based on the selective augmentation of rigid sphere of similar size, and consequently
echo strength. Unlike other imaging modalities, scatters more energy. In addition there is a fortu-
the majority of which have benefited from contrast itous relationship between the size of microbubble
enhancing agents for many years, contrast that is able to pass through a capillary and that
enhancement in ultrasound is a relatively new which will resonant at frequencies typically used
concept. The phenomenon was first observed in medical imaging (3 MHz).2 As such, at the
when during echocardiography an injection of resonant frequency the returning echoes from a
indocyanine green through a catheter resulted in microbubble are maximised in such an effective
transient echo enhancement in the region of the manner that these microbubbles behave as though
catheter tip.1 This observed enhancement was they are 1014 times larger than in reality.3
the result of small air bubbles forming in the region Microbubbles are better able to resist compres-
of the catheter tip and strongly reflecting the sion than expansion, and consequently as the
ultrasound beam back towards the transducer. acoustic power of the ultrasound increases
Following this observation much research has complex non-linear movements of the bubbles
been undertaken in both establishing the physics occur. The resultant signals emitted from the
of ultrasound contrast enhancement as well as microbubbles display harmonic characteristics
developing agents for clinical use. at both fractions and multiples of the insonating
frequency. These features can be utilised in the
optimisation of contrast to tissue signal ratio.
CHARACTERISTICS OF
Ultimately as acoustic power is increased physical
MICROBUBBLE ULTRASOUND
disruption of the microbubbles begins to occur,
CONTRAST
a feature that may be used to therapeutic
Ultrasound contrast agents in current use are advantage.4
typically small (<10 μm in diameter, approxi-
mately the size of an erythrocyte) gas filled bubbles
SAFETY OF MICROBUBBLE
(Fig. 15.1). Microbubble contrast agents are
CONTRAST
physiologically inert, non-toxic and pass
through the pulmonary circulation following Microbubbles are injected intravenously where
intravenous injection.The microbubbles contain the volumes of gas involved (typically less than
either air or an inert gas, encapsulated either by 200 μL) are considered non-hazardous.The safety
a membrane of surfactant only, a soft shell or a profile of microbubble contrast agents in general
firm shell. is extremely favourable when compared to other 337
15
Microbubble ultrasound contrast agents
at the time of writing include Levovist® (Schering
AG), Optison® (Amersham Health) and SonoVue®
(Bracco International B.V.).6 Definity® (Bristol-
Myers Squibb Medical Imaging) is licensed in
North America for cardiac use only.
Levovist®
In a similar manner as bubbles form on imper-
fections of the surface of a glass containing a
carbonated drink, galactose microcrystals will
provide a nidation site upon which air bubbles
form in Levovist®, an example of a surfactant
encapsulated microbubble.9 Palmitic acid is the
added surfactant that stabilises the bubbles allowing
Fig. 15.1 Electron microscopy of a microbubble
(SonoVue®); ultrasound microbubble contrast agents
their transit across the pulmonary circulation.
in current use are typically small (<10 μm in diameter) Even so, these microbubbles are fragile and disrup-
gas filled bubbles. Courtesy of Bracco International tion of the bubbles occurs at mechanical indexes
BV. used for standard B-mode imaging. To overcome
this, intermittent imaging and low frame rates
must be employed to allow refilling of the vascu-
radiographic contrast agents.5 Limited data lature with microbubbles; for example using
suggests that microvascular rupture may occur intermittent imaging of myocardial perfusion
when gas bubbles are insonated and there is a triggered by echocardiography to limit bubble
theoretical risk of injury should structures where bursting.10 Licensed clinical indications for the
vascular damage may be critical, for example the use of Levovist® are cardiac, abdominal and
eye or brain, are examined.6. Microembolism of transcranial imaging.
particles is a potential hazard. Evidence exists that
single microbubbles do not pose a significant Optison®
embolic risk but if coalescence of particles occurs Unlike Levovist® which utilises air as the gas agent,
the evidence is less clear.7 It is likely that the both Optison® and SonoVue® use larger gas
structure of the microbubbles influences coales- molecules; perfluoropropane in the case of Optison®.
cence with, for example, albumin-based shells These have the advantage of being long lived
being unlikely to coalesce.8 Current ultrasound relative to air. Excretion of the gas is ultimately
guidelines recommend that a maximum mechanical via the pulmonary route. Albumin provides the
index (MI) of 1.9 for diagnostic imaging should soft shell of the Optison® microbubble and the
not be exceeded; whether this is relevant in the only licensed clinical indication for use at present
context of microbubble use is unknown. is in cardiac imaging.
COMMERCIALLY AVAILABLE SonoVue®

Introduced in 2001, SonoVue® consists of sulfur
CONTRAST AGENTS
hexafluoride encased by a phospholipid soft
The need to form biologically and physically shell.11 Formation of the microbubbles is
stable microbubbles in commercial quantities in achieved by hand mixing of provided carrier
a simple, cost effective fashion has led to the fluid and freeze dried powder. Once mixed the
development of a variety of microbubble tech- bubbles remain stable in the vial for several
nologies. Commercially available contrast agents hours until required for use. The stability of this
338 in Europe approved for transpulmonary transit formulation in vivo allows real time imaging to
15
Ultrasound technologies used in contrast imaging

occur for several minutes following adminis- and frequency, microbubbles will scatter the
tration. Licensed clinical indications for use of insonating beam at subharmonic frequencies,
this agent include cardiac, vascular, liver and usually half the insonating frequency.12 While
breast applications. this technique has the potential of enhancing
differentiation between tissues and microbubbles
Definity® it is currently limited by the need for a narrow
Licensed in North America for cardiac use only, band of insonating signals. Such problems are
Definity® consists of lipid-stabilised octafluoro- likely to be overcome with further research.2
propane gas. The indication for use is to opacify
the left ventricular chamber and to improve the Phase inversion imaging
delineation of the left ventricular endocardial Phase or pulse inversion imaging (PII) achieves
border. improved signal to clutter ratio similar to harmonic
A summary of the features of the above agents imaging. Unlike harmonic imaging, PII has the
is provided in Table 15.1. advantage of maintaining spatial resolution in B-
mode imaging since it is not compromised by
the formers narrow band pulse.13 PII involves
ULTRASOUND TECHNOLOGIES
transmission of an initial imaging wave following
USED IN CONTRAST IMAGING
at a suitable delay by a second wave which is an
Optimising visualisation of microbubbles has inverted copy of the original.14 The resultant echo
required the development of specific ultrasound if these two waves were scattered linearly would
imaging modes. Manufacturers attach a variety be zero. However, as microbubbles are non-
of propriety names to the techniques employed; linear scatters, the sum of the two signals after
however there are several common themes. encountering a microbubble will not be zero15
(Fig. 15.2). In addition to the ability to use a
Harmonic imaging wider bandwidth than in harmonic imaging, PII
Insonating microbubbles at a given frequency allows relatively low insonating pressures (low
will result in the emission of energy at multiples MI) therefore reducing the risk of microbubble
of the insonating frequency and by tuning recep- disruption.This technique allows for continuous
tion to only these emitted frequencies, to form real time imaging during the various vascular
an image or Doppler signal, clutter signals from phases, allowing detection of vascular patterns
surrounding tissue will be suppressed. An peculiar to different liver tumours.This technique
advancement on this technique makes use of the works best with a robust microbubble agent
fact that, given the correct conditions of pressure such as SonoVue®.
Table 15.1 Characteristics of commercially available ultrasound contrast agents
Brand name Producer Gas Stabilisation Applications

Levovist® Schering Air Galactose/ Palmitic acid Cardiac
Abdominal
Transcranial
Optison® Amersham Health Perfluoropropane Sonicated albumin Cardiac
SonoVue® Bracco Sulfur hexafluoride Phospholipids Cardiac
Vascular
Hepatic
Breast
Definity® Bristol-Myers-Squibb Octafluropropane Phospholipids Cardiac
Medical Imaging
339
15
will present as defects within this colour pattern
(Fig. 15.3).The technique appears sensitive with
SAE improving conspicuity of liver metastasis
The sum of returned signals from a linear reflector of two pulses relative to B-mode imaging.17
180° out of phase with each other
a
The sum of returned signals from a non-linear reflector (such as
a microbubble) of two pulses 180° out of phase with each other
Fig. 15.2 The concept of phase or pulse inversion

imaging involves transmission of an initial imaging
wave followed at a suitable delay by a second wave
which is an inverted copy of the original. The resultant
echo if these two waves were scattered linearly
would be zero but as microbubbles are non-linear
scatters, the sum of the two signals after encountering
a microbubble will not be zero.
Stimulated acoustic emission

Although the theories and nomenclature regarding
this mode of imaging have been the matter of
some debate,16 stimulated acoustic emission (SAE)
refers to the strong transient signals associated
with microbubble disruption when imaged using
b
two-dimensional (2D) Doppler mode.16 The prin-
ciple of this imaging technique relies on the fact
that if a microbubble lies in the beam of a train
of insonating pulses then disruption of the micro-
bubble by the first pulse in the train will result in
an apparent large Doppler shift. This will be
registered on screen as extremes in the colour
Doppler scale or even what appears to be aliasing.
In reality there has been no large movement to
account for these extreme changes; it is simply
the disappearance of a microbubble. As such,
stationary microbubbles are as equally well
visualised as ones moving within blood vessels.
SAE has found application in the study of, for
example, focal liver lesions using Levovist®.17,18
By avoiding imaging the organ of interest while
Fig. 15.3 Agent Detection Imaging (ADI®, Acuson
the microbubble contrast agent is in the blood
Siemens, Mountain View, CA), a form of late phase
pool phase (approximately 2–5 min after injection)
destructive imaging; (a) baseline image of the right
early destruction of the microbubbles is prevented. lobe of the liver with a suggestion of an abnormality
Levovist® uniquely remains in the liver (and splenic (arrow); (b) ADI® in the late phase using Levovist®,
tissue) in the late phase. Subsequent imaging demonstrates two lesions with absence of
demonstrates the normal liver parenchyma to microbubble contrast within the lesions in keeping
340 have colour signal whereas malignant tumours with metastatic disease (arrows).
15
Quantification techniques
Artefacts
Doppler studies is that of ‘blooming’ in which
Quantification of microbubbles may take several colour pixels appear to extend beyond the bounds
forms including time to arrival, wash-in/wash- of the vessel (Fig. 15.4). The phenomenon may
out characteristics, time to peak and area under prove problematic when objective computer
a curve.3 The situation is complicated by virtue of analysis of contrast enhancement is required.
the fact that there is, unlike for example computed While the problem in part arises from multiple
tomography (CT) Hounsfield units, no single re-reflections between adjacent microbubbles, it
variable unit of measurement for ultrasound contrast is also in part due to limitations of the hardware
which can be reliably and reproducibly measured. analysis of the Doppler signal, a problem which
Furthermore, colour contrast enhancement is may be overcome with technological improve-
subject to multiple artefacts including motion arte- ments. A further artefact resulting from
fact from patient or operator movement which, limitations in the ultrasound hardware is the
in long examinations, can be difficult to avoid. apparent reversal of Doppler flow following
Strategies for attaining ‘quantification analyses’ administration of contrast. This effect represents
include; measurement of the back-scattered signal overloading of the Doppler circuitry and once
and Doppler intensity, both of which are related appreciated is easily recognised.
to the microbubble concentration. Using propri- The administration of contrast may render
etary software regions of interest may be drawn high velocity signals, which precontrast were not
and analysis pre- and postcontrast achieved. detectable, now detectable.19 Whilst this may
initially appear to be a desirable effect it proves
problematic when absolute angle-corrected values
ARTEFACTS
are essential to clinical decision making, for
One of the most commonly seen artefacts when example in assessment of internal carotid artery
using microbubbles to enhance conventional narrowing. Although adding a further layer of
a b
Fig. 15.4 A commonly seen artefact when using microbubbles to enhance conventional Doppler studies is
‘blooming’ in which colour pixels appear to extend beyond the bounds of the vessel; (a) baseline colour Doppler
image at the level of the porta hepatis demonstrating absence of colour signal in the portal vein (arrow)
but colour is present in the overlying hepatic artery; (b) following the administration of SonoVue® ‘blooming’
from the hepatic artery obscures the portal vein. 341
15
complexity to the measurement, the problem
may be overcome by use of peak systolic ratios
to define limits rather than absolute velocity
values.
Sharp spikes of Doppler ‘noise’ may occa-
sionally be seen (Fig. 15.5). Such effects may arise
from collapse of the microbubbles or aggregation
of microbubbles.13
CLINICAL APPLICATIONS OF
MICROBUBBLE CONTRAST
Cardiac
Cardiology represents the single greatest area of Fig. 15.5 Sharp spikes of Doppler ‘noise’ (arrows)
use of microbubble contrast agents with several are seen in the spectral Doppler trace obtained from
established and evolving applications. a transcranial ultrasound examination following the
administration of Levovist®, arising from collapse of
the microbubbles or aggregation of microbubbles.
a Fig. 15.6 Apical four chamber

view using tissue harmonic
imaging (a) in an intensive care
unit patient with very poor
imaging windows. The patient
had been recovering slowly post
cardiac surgery and the request
was to assess left ventricular
function (LV) function. Image
quality was so poor that no
comment could be confidently
made about LV function.
342
Fig. 15.6 cont’d Before

15
Clinical applications of microbubble contrast

b
proceeding to perform a
transoesophageal echo a
microbubble contrast study was
performed using a 0.3 mL
injection of Optison® through an
existing intravenous line and
power modulation low mechanical
index contrast imaging (b).
Evaluation of the LV endocardial
border was significantly
enhanced, especially at the apex
and the lateral wall, allowing an
accurate assessment of LV
function without needing to
proceed to a transoesophageal
echo. The improvement in image
quality was even more evident
on real-time images than on
these still frames. (Copyright MJ
Monaghan).
Echocardiography be acquired during a stress echocardiography

Microbubble ultrasound contrast allows improved examination. Furthermore valvular disease may
delineation of the endocardial border and be more accurately assessed.3
detection of wall motion abnormalities.20,21 This
is particularly true in the 10–20% of routine Myocardial perfusion
echocardiograms which are technically difficult, Assessment of myocardial perfusion offers poten-
for example in lung disease or obesity, where up tial for the diagnosis of acute myocardial infarction
to 74% of otherwise non-diagnostic studies are as well as assessing myocardium which is at risk
‘rescued’ by the use of microbubble contrast22 of infarction and that which has responded to
(Fig. 15.6). Unlike isotope studies, contrast thrombolytic therapy. With use of a more robust
echocardiography is a mobile technique which microbubble formulation in addition to the
may be performed at the bedside, avoids the development of low MI techniques allows a more
need for ionising radiation and with consequent comprehensive myocardial assessment. Further-
cost benefits to be gained. Use of microbubble more, the application of intermittent high power
contrast allows additional information regarding pulses to destroy the microbubbles allows
regional and global left ventricular function to assessment of rate of refilling of the myocardium 343
15
microcirculation so providing a measure of of microbubble contrast by normal functioning
microcirculation flow speed.22 cells.
Several studies have demonstrated improved
Coronary arteries detection of liver lesions using microbubble
Localisation and quantification of degree of contrast.25,26 Much of the early data refers to the
coronary artery stenosis may be performed using use of delayed imaging of Levovist® during the
intermittent harmonic imaging techniques. late liver specific phase, using the technique of
Analogous to the assessment of microcirculation SAE. In these studies malignant focal liver lesions
described above, coronary artery flow rate may generally appear as defects in an otherwise
be calculated by intermittent destruction of enhancing liver parenchyma, whereas benign lesions
microbubbles and measurement of time taken to demonstrate microbubble contrast uptake. Data
refill.22 regarding the use of other agents, using continuous
non-linear techniques with low MI, is emerging
Liver with initial results using SonoVue® proving particu-
The histological architecture, rich vascular supply larly useful.27 Imaging of a focal liver lesion during
and superficial nature, allows the liver to be suited different vascular phases provides clues to the
to interrogation with microbubble ultrasound identity of the lesion with good correlation between
contrast agents. A variety of different contrast the findings of microbubble contrast enhanced
agents have been employed to study the liver; ultrasound and CT or magnetic resonance (MR)
consequently, much of our understanding imaging. Generally the current practice is to use
regarding microbubble contrast has been gained non-linear low MI techniques to characterise
from studies of liver pathology.23,24 lesions; the liver lesion is identified with B-mode
imaging then, with the probe held stationary over
Focal liver lesions the lesion, continuous imaging is performed for
One of the most promising areas of use for 60–90 s after injection of a bolus of microbubble
microbubble contrast agents is in the study of focal contrast followed by a saline flush via a
liver lesions. In clinical terms two important aspects peripheral vein.
of analysis of focal liver lesions are detection and Using the low MI method the characteristics
then characterisation into benign and malignant. of different types of focal liver lesions have been
In the healthy liver approximately 70% of the established. Haemangiomas, common benign liver
blood is supplied by the portal vein with the lesions, typically display progressive enhancement
remaining 30% by the hepatic artery. Venous starting from the periphery of the lesion, as also
drainage for the majority of the liver (except the demonstrated with CT and MR imaging24 (Fig.
caudate lobe which drains directly into the inferior 15.7). Focal fatty sparing, focal fatty change and
vena cava) is via the left, right and middle hepatic regenerative nodules may have alarming appear-
veins into the inferior vena cava. Analogous to ances with contrast enhanced CT and MR imaging
the phases seen with CT and iodinated contrast, but reassuringly display isoenhancement compared
microbubble ultrasound contrast displays an early to surrounding liver parenchyma following
(arterial) phase at around 20 s post intravenous microbubble contrast administration. Focal
administration followed by a portal venous nodular hyperplasia, a benign vascular anomaly,
phase at approximately 90–120 s. Depending on tends to hyperenhance with microbubble contrast
the formulation of the microbubble contrast agent with demonstration of a characteristic central
there is then a late phase, greatest with Levovist®, hypoenhancing scar. Adenomas, which on
as the agent is cleared from the liver parenchyma. occasion may be very large and associated with
This late phase differs in its pharmacokinetics haemorrhage, generally show hyperenhancement
from the clearance of iodinated radiographic in the arterial and portovenous phases becoming
344 contrast, and is likely to be attributable to uptake isoenhancing with surrounding parenchyma on
15

a abscess and the surrounding liver segment may
be hyperenhancing (Fig. 15.8).
Unfortunately there is no single feature to
differentiate a benign from a malignant lesion on
low MI imaging. Hepatocellular carcinomas (HCC)
tend to be hyperenhancing in the arterial phase,
iso- or hypoenhancing in the portovenous phase
Fig. 15.7 A haemangioma demonstrates progressive

enhancement starting from the periphery of the
lesion during the late phase; (a) baseline image of an
atypical focal liver lesion (arrows); (b) microbubble
enhanced image in the late phase, using SonoVue®
and CPS® imaging (Acuson, Siemens) demonstrates
‘creeping’ enhancement from the periphery (arrows)
typical of a haemangioma.
Fig. 15.8 Complications relating to the liver

parenchyma following transplantation that can be
delayed imaging. Simple liver abscesses are often
effectively interrogated using microbubble contrast
readily diagnosed on B-mode imaging but if there include abscesses; (a) on baseline imaging a large
is diagnostic doubt, administration of microbubble heterogeneous mass (arrows) is present in the right
contrast displays a hyperenhancing rim and a low lobe of the liver; (b) following administration of
reflective central area on arterial phase. Occa- SonoVue®, and using CPS®, a multiseptated abscess
sionally enhancing septa may be seen within the is demonstrated. 345
15
and hypoenhancing on delayed imaging. Occa- of access, the possibility of a bedside examination
sionally an HCC may show some late phase in the emergency room, lack of ionising radiation
enhancement (Fig. 15.9). Cholangiocarcinomas absence of nephrotoxicity.
generally do not enhance with microbubble
contrast although there may be early arterial rim Liver transplantation and liver
enhancement in some cases. As with other imaging vasculature
modalities the ultrasound characteristics of liver Microbubble contrast agents are a useful diag-
metastasis depend on the source of the primary. nostic tool in the practice of liver transplantation
In general metastasis are hypoenhancing however, ultrasound imaging.30 Pretransplant patient
predictably, hypervascular metastasis are hyper- assessment is essential to select candidates who
enhancing in the arterial phase of scanning. are most likely to benefit from a liver transplant.
Although not specific to hepatic malignancies, Both liver parenchyma and hepatic vasculature
one development which may in the future provide need to be assessed; in particular portal vein and
supplementary evidence of a malignant versus a superior mesenteric vein (SMV) patency must
benign tumour is the study of the disorganised be documented. Not uncommonly patients will
neovascularity which is associated with rapidly have a cirrhotic highly attenuating liver, large
growing malignant tumours.28 volume ascites or bowel gas rendering visuali-
sation of the vasculature problematic. In such
Trauma cases demonstration of portal vein and SMV
Microbubble ultrasound contrast may reveal solid patency is improved by the use of microbubble
abdominal organ injuries not evident with B-mode contrast (‘Doppler rescue’) (Fig. 15.10). Further-
imaging.29 Although at present contrast enhanced more in cases of long-term portal vein throm-
CT remains the reference standard in the context bosis the presence of cavernous transformation
of trauma, microbubble enhanced ultrasound may be seen (Fig. 15.11).31 Diagnostic confidence
has certain inherent advantages including ease of hepatic vein and inferior vena cava thrombus
b
a
Fig. 15.9 Benign lesions tend to demonstrate persistence of contrast enhancement in portal venous and late
phases whereas malignant lesions tend to show early wash out of contrast in the portal and late phases;
a notable exception is hepatocellular carcinoma which can show some late phase enhancement; (a) baseline
image of a hepatocellular carcinoma in the right lobe of the liver (arrow); (b) following administration of SonoVue®,
346 there remains some microbubble contrast within the lesion (arrow).
15

b
Fig. 15.10 Demonstration of portal vein patency status is improved by the use of microbubble contrast,
‘Doppler rescue’; (a) on baseline imaging there is no colour Doppler signal from the portal vein (arrow) but flow
is seen in the hepatic artery; (b) following administration of SonoVue®, and using CCI® (Acuson Siemens,
low mechanical imaging) contrast is seen in the hepatic artery (small arrows) but no contrast is present in the
portal vein (arrow), confirming occlusion.
a
b
Fig. 15.11 With long-term portal vein thrombosis the presence of cavernous transformation may be confirmed,
‘Doppler rescue’; (a) baseline imaging demonstrates a dilated high reflective portal vein (arrow) and serpiginous
low reflective structures (small arrows) in keeping with cavernous transformation; (b) following SonoVue®,
the cavernous transformation and the occluded portal vein (arrow) are demonstrated.
occlusion in the Budd–Chiari syndrome can be Transjugular intrahepatic portosystemic shunt

improved and the spider-web like pattern of intra- (TIPS) can provide effective palliation in cases
hepatic veins may be demonstrated following of recurrent varicocele haemorrhage or intractable
administration of microbubble ultrasound ascites. The principle long-term complication of
contrast. TIPS, occurring in up to 30% of patients in the 347
15
a
first year, is occlusion.32 As with other vascular
structures in the cirrhotic liver this can be diffi-
cult to assess using standard Doppler ultrasound
with administration of microbubbles providing
evidence of low flow within a TIPS.
In assessing the liver parenchyma pretrans-
plantation, it is important to establish the presence
of occult malignancy, in particular HCC. Initial
experience suggests microbubble contrast
enhanced ultrasound may improve conspicuity
of such lesions.
In the post-transplant liver, the principal
vascular supply is from the anastomosed hepatic
artery, with the portal vein playing a less signifi-
b
cant role.When performing ultrasound assessment
post-transplant, the single most important
structure to visualise and document patency is
the hepatic artery; compromise of this vessel may
result in bile duct ischaemia, necrosis and bile
duct leaks.33 If a hepatic artery spectral Doppler
trace is not demonstrated using standard Doppler
the patient often undergoes formal angiography;
use of microbubble contrast could negate the
need for arteriography in such cases.34 Pseudo-
aneurysms of the hepatic artery may arise secondary
to localised infection or as a consequence of liver
biopsy; rupture may result in fatal haemorrhage.
The swirling blood flow within a pseudoaneurysm
(the ‘yin-yang’ sign) may be elegantly demon-
Fig. 15.12 Following endovascular embolisation of a
strated post microbubble contrast allowing hepatic
hepatic artery pseudoaneurysm abdominal
artery reconstruction or endovascular embolisation
haemorrhage continues; (a) baseline colour imaging
to proceed with confidence (Fig. 15.12). Although
demonstrates the low reflective pseudoaneurysm
relatively uncommon post-transplant, both portal (arrow) and embolisation coils (small arrows) without
vein and hepatic vein thrombosis can be assessed any colour flow signal present; (b) using SonoVue®
with microbubble contrast.There are a variety of and CCI®, there is clear demonstration of continuing
complications relating to the liver parenchyma blood flow (arrow) and the need to perform further
and surrounding structures that can be effec- endovascular embolisation to stop the haemorrhage.
tively interrogated using microbubble contrast
including parenchymal infarction (Fig. 15.13),
recurrent malignancy, post-transplant lympho- the spleen may cause diagnostic confusion; the
proliferative disorder, localised fluid collections addition of microbubble contrast will enhance
and abscesses.35 the operators’ confidence when assessing such
lesions.36 Splenunculi occur in up to 30% of the
Spleen population, may be multiple and can be found at
The spleen is often a technically difficult organ locations remote to the spleen with the primary
to adequately visualise on B-mode ultrasound, differential for splenunculi, particularly when
348 where several conditions within and surrounding multiple, that of lymph node enlargement.
15

a and greater than 2 cm in diameter. When large,
microbubble contrast demonstrates a centripetal
enhancement pattern, analogous to haemangiomas
found in the liver. Other benign solid lesions are
rare but include lymphangioma and hamartoma
which both show varying degrees of enhancement
post microbubble administration.Whilst primary
malignant tumours of the spleen such as lym-
phoma and angiosarcoma do occur they are both
Fig. 15.13 Complications relating to the liver

parenchyma following transplantation that can be
effectively interrogated using microbubble contrast
include parenchymal infarction; (a) low reflective
areas (arrows) are present within the left lobe of the
liver and there is a subhepatic collection noted
(arrowhead); (b) following the administration of
SonoVue® and using CPS®, areas of infarction are
demonstrated (arrows) in this patient with a hepatic
artery occlusion.
Microbubble contrast enhanced ultrasound accu-

Fig. 15.14 Microbubble contrast enhanced
rately demonstrates that the parenchymal enhance-
ultrasound demonstrates the parenchymal
ment of the splenunculi matches the main spleen
enhancement of a splenunculi that matches the
with a clear vascular hilum (Fig. 15.14).
spleen; (a) baseline image demonstrates a well
circumscribed iso reflective (to spleen) area (arrow)
Focal splenic lesions adjacent to the splenic hilum; (b) following the
Cavernous haemangiomas are the most common administration of Levovist® using ADI®, there is
of the benign solid splenic lesions, usually small microbubble contrast uptake in the late phase
and solitary but they may occasionally be multiple confirming the presence of splenic tissue (arrow). 349
15
a
rare.36 More common, although still unusual,
are metastatic deposits to the spleen.The majority
of these deposits are serosal (cause scalloping of
the spleen surface) as opposed to parenchymal
and arise from a primary neoplasm in, for example,
the ovary.36 Simple splenic cysts, either congenital
or acquired, rarely cause diagnostic uncertainty.
Spleen abscesses may have ultrasound appearances
ranging from cystic to solid and generally appear
avascular on both Doppler ultrasound and micro-
bubble contrast enhanced ultrasound. Splenic
infarction typically occurs in patients with infec-
tive endocarditis or myeloproliferative disorders,
often difficult to visualise on initial B-mode ultra-
sound but may show lack of vascular perfusion b
on colour Doppler ultrasound.37 Conspicuity of
infarcts is improved by the use of microbubble
ultrasound contrast media (Fig. 15.15).
Trauma
The spleen is prone to injury following blunt
abdominal trauma. Prompt diagnosis is central
to effective management of this potentially fatal
condition.While CT remains the reference standard
for diagnosis of splenic injury, microbubble contrast
may prove to be a useful adjunct. Following
administration of contrast, a non-enhancing linear
branching laceration not evident on B-mode Fig. 15.15 Splenic infarction may occur in patients
imaging may be shown (Fig. 15.16). with infective endocarditis; (a) on baseline imaging
there is a heterogeneous area in the spleen (arrow)
Renal without particular characteristics; (b) conspicuity of
Complex renal cysts this area is improved by the use of microbubble
On finding a complex renal cyst most institu- contrast (SonoVue® and CCI®) to reveal an area of
infarction (arrows).
tions currently undertake contrast enhanced CT
to further assess the kidney, looking for a
potential malignancy. Microbubble contrast may
ultimately prevent the need for so many CT Renal vasculature
examinations but at present experience is still Whilst Doppler ultrasound of the renal arteries
being accumulated.The vascularity of a soft tissue is an accepted screening examination for renal
mass within a renal cyst may be readily demon- artery stenosis, this technique is technically chal-
strated with microbubble contrast (Fig. 15.17). lenging. Problems arise in visualising the entire
Microbubble contrast may help delineate ‘pseudo- renal artery, identifying accessory arteries and
tumours’ from real renal masses as in the case of obtaining accurate Doppler traces due to patient
a prominent column of Bertin. With a true renal factors and being unable to access a suitable angle
mass, microbubble contrast may illustrate the on the artery. Microbubble contrast enhancement
abnormal distorted vasculature and confirm or may overcome some of these problems with the
350 refute the patency of the renal vein (Fig. 15.18). number of successful examinations increasing
15

a b
Fig. 15.16 In blunt abdominal injury microbubble contrast may be a useful adjunct; (a) on the baseline image
there is a high reflective heterogeneous pattern observed in the spleen (arrows); (b) following administration of
SonoVue®, and imaging with CCI®, a non-enhancing laceration is demonstrated (arrow) as well as other areas of
non-perfusion in keeping with significant splenic trauma.
a
b
Fig. 15.17 With a complex renal cyst the vascularity of a soft tissue mass within the cyst may be readily
demonstrated with microbubble contrast; (a) on baseline imaging a soft tissue component (arrow) makes this
renal cyst suspicious for a malignant lesion; (b) following the administration of SonoVue® and using CPS®,
the soft tissue component does not enhance (arrow) demonstrating that there is no vascularity to this area,
and it is likely benign.
by as much as 20%.38 Microbubble ultrasound microbubble contrast can aid in the diagnosis of
contrast is employed in the visualisation of the renal artery stenosis or renal vein thrombosis.
vasculature post renal transplantation. In such Aside from vascular applications in the renal
cases viability of the transplant is dependent on tract, microbubble contrast is used to investigate
the patency of the renal artery and vein and ureteric reflux via direct installation of micro- 351
15
a pelvic organs. The use of hystero-contrast salp-
ingography (HyCoSy) avoids these risks. Instillation
of microbubble contrast in to the uterine cavity
is performed and using transvaginal ultrasound,
the passage of contrast through the Fallopian
tubes and spillage into the pelvis may be followed
in real time.40 While this serves as a good
screening test for tubal patency, it cannot at
present provide detailed anatomical delineation
of tubal pathology. As such in cases where tubal
occlusion is demonstrated the patients may
progress to a formal HSG prior to therapeutic
intervention. Both HSG and HyCoSy can result
in significant patient discomfort, with some
b studies suggesting pain is more common with
HyCoSy.41
Neurological
Transcranial ultrasound
Accurate visualisation of the intracerebral vascu-
lature allows diagnosis and monitoring of arterial
spasm after stoke or subarachnoid haemorrhage
and vessel occlusion in sickle cell disease. Assess-
ment of these vessels using standard Doppler
ultrasound is limited by the severe attenuation
of the signal due to the skull vault. Use of micro-
bubble contrast agents can overcome this difficulty
and allow accurate visualisation of the vessels
Fig. 15.18 In a true renal mass, microbubble contrast
will demonstrate abnormal distorted vasculature;
(Fig. 15.19).42 Indeed in one study the use of
(a) on the baseline image a complex mass is present microbubble contrast technically improved the
at the upper aspect of the left kidney (arrow); images in 77% of patients.43
(b) following the administration of SonoVue®,
and using CPS®, abnormal vessels are demonstrated Carotid artery
within the tumour (arrow). Microbubble contrast use in the assessment of
carotid arteries is focused on two areas. First is
to accurately visualise flow within the carotid
bubble contrast into the bladder and observing artery, a task that is frequently made complex by
reflux on ultrasound. Although this is appealing overlying calcified plaque and vessel tortuosity.
in that ionising radiation could be avoided in In one study, on baseline colour Doppler imaging
young patients, practical difficulties remain in 21% of vessel stenosis was not identified in com-
the implementation of the method.39 parison to only 6% post microbubble contrast.44
Second, is to differentiate between vessel occlu-
Hystero-contrast salpingography sion and high grade stenosis, a distinction which
In the investigation of subfertility, Fallopian tube can be difficult to make with standard colour
patency has traditionally been established by Doppler. Microbubble contrast improves this,
hysterosalpingography (HSG), a technique requiring decreasing the false-negative rate from 30% to
352 both iodinated contrast and irradiation of the 17%.45
15

a
Fig. 15.20 Inflammatory change in the proximal

inter-phalangeal joint of a patient with rheumatoid
disease and synovitis is demonstrated following the
administration of SonoVue® as a microbubble
contrast agent.
monitoring disease progression and response to

therapy.
Treatment
The currently available microbubbles consist of
gas contained within a shell. Potential exists to
replace this gas with a ‘cargo’ and effectively use
the microbubble as a vehicle to transport a thera-
peutic agent to a specified site. Ultimately this
may prove to be an application of microbubbles
more important than any of those discussed above.
Perhaps the most interesting use would be in the
Fig. 15.19 The application of microbubble contrast in area of gene therapy where the initial hopes of
transcranial ultrasound is useful; (a) on the baseline effective treatments have often been limited by
image a single vessel is visualised on colour Doppler lack of suitable delivery mechanism. Sonoporation,
ultrasound (arrow); (b) following the administration of
the mechanism by which cell membranes are
Levovist® and using conventional colour Doppler
rendered porous to large molecules by ultrasound,
ultrasound, both the circle of Willis and the draining
is enhanced in the presence of microbubbles.
veins of the brain are demonstrated (arrows).
The ability of this process to deliver gene therapy
has been shown.48 Aside from gene therapy the
Musculoskeletal possibility for delivery of other therapeutic agents
Currently principally a research tool, microbubble also exists. These include chemotherapeutic
contrast is finding application in the study of agents whose site of delivery could be dictated
articular inflammatory disease. Several studies by selective microbubble disruption at the tumour
have shown synovial enhancement in patients with site. Alternatively, drug delivery could be directed
synovial inflammation confirmed with MR imaging by use of ligands within the microbubble mem-
(Fig. 15.20).46,47 Such findings may offer oppor- brane which bind to specific cell receptors at
tunity not only in terms of diagnosis but also target tissue.
353
15
REFERENCES
1. Gramiak R, Shah PM. Echocardiography of the 18. Forsberg F, Liu JB, Burns PN, et al. Artifact in
aortic root. Invest Radiol 1968; 3:356–366. ultrasonic contrast agents studies. J Ultrasound
2. Stride E, Saffari N. Microbubble ultrasound contrast Med 1994; 13:357–365.
agents: a review. Proc Inst Mech Eng [H] 2003; 19. Petrick J, Zomack M, Schlief R. An investigation of
217:429–447. the relationship between ultrasound echo
3. Harvey CJ, Blomley MJK, Eckersley RJ, et al. enhancement and Doppler frequency shift using a
Developments in ultrasound contrast media. pulsatile arterial flow phantom. Invest Radiol 1997;
Eur Radiol 2001; 11:675–689. 32:225–235.
4. Porter TR, Xie F, Li S, et al. Increased ultrasound 20. Nihoyannopoulos P. Contrast echocardiography.
contrast and decreased microbubble destruction Clin Radiol 1996; 51:28–30.
rates with triggered ultrasound imaging. J Am Soc 21. Kaul S. New developments in ultrasound
Echocardiogr 1996; 9:599–605. systems for contrast. Clin Cardiol 1997;
5. Nanda NC, Carstensen C. Echo-enhancing agents: 20:127–130.
safety. In: Nanda NC, Goldberg BB, eds. Advances 22. Blomley MJK, Cooke JC, Unger EC, et al.
in echo imaging using contrast enhancers. Microbubble contrast agents; a new era in
Dordrecht: Kluwer; 1997:115–131. ultrasound. Br Med J 2001; 322:1222–1225.
6. Albrecht T, Blomley MJK, Bolondi L, et al. 23. Leen E, McArdle CS. Ultrasound contrast agents in
Guidelines for the use of contrast agents in liver imaging. Clin Radiol 1996; 51:35–39.
ultrasound. Ultraschall Med 2004; 25:249–256. 24. Leen E. Ultrasound contrast harmonic imaging of
7. Nyborg WL. Biological effects of ultrasound: abdominal organs. Semin Ultrasound CT MRI 2001;
development of safety guidelines. Part II; general 22:11–24.
review. Ultrasound Med Biol 2001; 27:301–333. 25. Bryant TH, Blomley MJK, Albrecht T, et al. Liver
8. Gromley G, Wu J. Observation of acoustic phase uptake of a liver specific microbubble
streaming near Albunex spheres. J Acoust Soc Am improves characterization of liver lesions: a
1998; 104:3115–3118. prospective multi-center study. Radiology 2004;
9. Schlief R. Developments in echo-enhancing agents. 232:392–399.
Clin Radiol 1996; 51:5–7. 26. Blomley MJK, Sidhu PS, Cosgrove DO, et al.
10. Wei K, Kaul S. Recent advances in myocardial Do different types of liver lesions differ in their
contrast echocardiography. Curr Opin Cardiol 1997; uptake of the microbubble contrast agent SH U
12:539–546. 508A in the late liver phase? Early experience.
11. Bokor D. Diagnostic efficacy of SonoVue. Am J Radiology 2001; 220:661–667.
Cardiol 2001; 86:19G–24G. 27. Albrecht T, Oldenburg A, Hohmann J, et al. Imaging
12. Neppiras EA. Acoustic cavitation. Phys Rep 1980; of liver metastases with contrast-specific low-MI
61:159–251. real-time ultrasound and SonoVue. Eur Radiol 2003;
13. Cosgrove DO. Ultrasound contrast agents. 13:N79–N86.
In: Meire H, Cosgrove DO, Dewbury K, et al, 28. Cosgrove DO. Microbubble enhancement of
eds. Clinical Ultrasound: a comprehensive text. tumour neovascularity. Eur Radiol 1999;
London: Churchill Livingstone; 2000:67–79. 9:S413–S414.
14. Simpson DH, Chin CT, Burns PN. Pulse inversion 29. Kraemer N, Cosgrove DO, Blomley MJK.
Doppler: a new method for detecting nonlinear Microbubble ultrasound demonstrates liver trauma.
echoes from microbubble contrast agents. J Trauma 2004; 56:913–914.
IEEE Trans Biomed Eng 1999; 46:372–382. 30. Berry JD, Sidhu PS. Microbubble contrast-enhanced
15. Albrecht T, Hoffmann CW, Schettler S, et al. B-mode ultrasound in liver transplantation. Eur Radiol 2004;
enhancement at phase-inversion US with air-based 14:P96–P103.
microbubble contrast agent: initial experience in 31. Marshall MM, Beese RC, Muiesan P, et al.
humans. Radiology 2000; 216:273–278. Assessment of portal venous patency in the liver
16. Tiemann K, Becher H, Bimmel D, et al. Simulated transplant candidate: a prospective study
acoustic emission nonbackscatter contrast effect of comparing ultrasound, microbubble-enhanced
microbubbles seen with harmonic power Doppler colour Doppler ultrasound with arteriography and
imaging. Echocardiography 2000; 14:65–70. surgery. Clin Radiol 2002; 57:377–383.
17. Blomley MJK, Albrecht T, Cosgrove DO, et al. 32. Shaw AS, Ryan SM, Beese RC, et al. Liver
Improved imaging of liver metastases with transplantation. Imaging 2002; 14:314–328.
stimulated acoustic emission in the late phase of 33. Zajko AB, Campbell WL, Logsdon G.
enhancement with the US contrast agent SH U Cholangiographic findings in hepatic artery
508A: early experience. Radiology 1999; occlusion after liver transplantation. AJR Am J
354 210:409–416. Roentgenol 1987; 149:485–489.
34. Sidhu PS, Shaw AS, Ellis SM, et al. Microbubble
42. Bauer A, Becker G, Krone A, et al. Transcranial

15
References
ultrasound contrast in the assessment of hepatic duplex sonography using ultrasound contrast
artery patency following liver transplantation: role in enhancers. Clin Radiol 1996; 51:19–23.
reducing frequency of hepatic artery arteriography. 43. Otis SM, Rush M, Boyajian R. Contrast-enhanced
Eur Radiol 2004; 14:21–30. transcranial imaging. Results of an American phase-
35. Shaw AS, Ryan SM, Beese RC, et al. Ultrasound of two study. Stroke 1995; 26:203–209.
non-vascular complications in the post-transplant 44. Sitzer M, Furst G, Siebler M, et al. Usefulness of an
patient. Clin Radiol 2003; 58:672–680. intravenous contrast medium in the characterization
36. Peddu P, Shah M, Sidhu PS. Splenic abnormalities: of high-grade internal carotid stenosis with color
a comparative review of ultrasound, microbubble Doppler-assisted duplex imaging. Stroke 1994;
enhanced ultrasound and computed tomography. 25:385–389.
Clin Radiol 2004; 59:777–792. 45. Furst G, Saleh A, Wenserski F, et al. Reliability and
37. Goerg C, Schwerk WB. Color Doppler imaging of validity of noninvasive imaging of internal carotid
focal splenic masses. Eur J Radiol 1994; artery pseudo-occlusion. Stroke 1999;
18:214–219. 30:1444–1449.
38. Claudon M, Plouin PF, Baxter GM, et al. Renal 46. Carotti M, Salaffi F, Manganelli P, et al. Power
arteries in patients at risk of renal arterial stenosis: Doppler sonography in the assessment of synovial
multicentre evaluation of the Echo-enhancer SH U tissue of the knee joint in rheumatoid arthritis:
508A at color and spectral Doppler US. Radiology a preliminary experience. Ann Rheum Dis 2002;
2000; 214:737–746. 61:877–882.
39. Mentzel HJ, Vogt S, Patzer L, et al. Contrast- 47. Margarelli N, Guglielmi G, Matteo L, et al.
enhanced sonography of vesicoureterorenal reflux Diagnostic utility of an echo-contrast agent in
in children: preliminary results. AJR Am J patients with synovitis using power Doppler
Roentgenol 1999; 173:737–740. ultrasound: a preliminary study with comparison
40. Ayida G, Harris P, Kennedy S, et al. Hysterosalpingo- to contrast-enhanced MRI. Eur Radiol 2001;
contrast sonography (HyCoSy) using Echovist-200 11:1039–1046.
in the outpatient investigation of infertility patients. 48. Price RJ, Skyba DM, Kaul S, et al. Delivery of
Br J Radiol 1996; 69:910–913. colloidal particles and red blood cells to tissue
41. Stacey C, Brown C, Manhire A, et al. HyCoSy – through microvessel ruptures created by targeted
as good as claimed? Br J Radiol 2000; microbubble destruction with ultrasound.
73:133–136. Circulation 1998; 98:1264–1267.
355
Appendix: System controls
and their uses
In addition to the basic choice of transducer type penetration for imaging is normally chosen. The
and frequency for the examination in hand, there Doppler frequency used by any transducer is
are many other factors on a Doppler ultrasound often 1–2 MHz below the imaging frequency,
system which need to be adjusted. Despite the although modern equipment has a wide range of
efforts of the manufacturers to automate and receive frequencies such as 5–14 MHz. In addition
simplify things, it is still necessary to adjust to the basic imaging requirements, it should be
continually many of the scan and Doppler remembered that the deeper a vessel lies, the
parameters during the course of an examination. longer it takes a sound pulse to travel there and
Is the vessel superficial or deep? Is flow fast or back, so that the Nyquist limit becomes very
slow, high volume or low volume? Most systems relevant (Ch. 1) and limits the Doppler frequencies
now come with a variety of preset programs for that can be used and therefore velocities that can
different situations: peripheral arteries, veins, be recorded accurately.
cerebrovascular, etc. Many also have the facility
to allow users to save their own program prefer- B-mode image
ences.Whilst these presets allow the basic appro- For colour Doppler examinations this should be
priate settings to be entered, fine adjustments set up with relatively low overall gain, so that the
will still be required during the course of the image is a little on the dark side as the software
examination to make the most of the available tends to allocate colour to darker areas, rather
information. Familiarity with the ultrasound system than to areas which contain echoes. See ‘colour
and experience enable skilled operators to set up write priority’ below.
their system appropriately for the examination
being performed. Different manufacturers Transmit power
sometimes use different names for the same The transmit power of the system should be set
controls or functions; it is not possible to give an at the lowest level consistent with an adequate
exhaustive list of all the possible options but the examination, especially during obstetric and
following notes describe the basic controls, or gynaecological examinations. It is better to start
parameters on most systems that can be adjusted at a medium level and increase the power only
during the course of an examination to improve after other measures to improve system sensitivity
and maximise the information that is obtained have been tried, such as adjusting colour gate
from the examination. size, removing filters, adjusting the scale/pulse
repetition rate. For low mechanical index (MI)
contrast studies (see Ch. 15) the transmit power
GENERAL PRINCIPLES
should be set as low as possible and the MI
Transducer frequency reading ideally should be less than 0.4; modern
The highest frequency which will achieve the systems may have contrast agent-specific
highest resolution consistent with adequate programs already installed, or the relevant 357
Appendix: System controls and their uses
settings can be obtained from the supplier of the as much information about aliasing as those which
contrast agent. have significantly different colours, such as pale
green and pale orange. Variance maps were said
Update/duplex/triplex to register the amount of spectral broadening in
In duplex ultrasound there is the ability to the colour map, although this is not easy to
acquire and display both imaging and Doppler appreciate and may, in fact, not be true as this
information either simultaneously, or alternately. ‘variance’ seems to reflect the velocity, rather than
Simultaneous display results in degradation of spectral broadening. Variance maps are most
both the image and the spectral display as the frequently used in cardiological examinations.
computer has to process data from both sources. In addition, colour tags can be put into the scale
The update facility allows the operator to set the so that velocities above or below the chosen
system to handle either imaging data, or Doppler range can be identified.
data. This results in a higher quality of display
for the selected mode and it is usually used for Colour gain
acquisition of the best-quality spectral display The colour gain is set to the optimal level. This
for analysis. Duplex scanning is of some value in can be identified by turning up the gain until
the initial stages of an examination in order to noise/colour speckle is seen in the colour box
position the sample volume in the area of interest. and then backing off slightly on the overall
Triplex mode refers to the simultaneous acqui- colour gain. It should also be adjusted to remove
sition, processing and display of colour Doppler, any spill of the colour map over the boundaries
spectral display and imaging information. As with of the vessel seen on the B-mode images; this is
duplex scanning, this requires significant division sometimes called colour bleeding.
of processing power and consequent compro-
mise in the quality of the display. Newer systems Colour scale
with more powerful computers are less prone to The colour scale should be set to levels appro-
these problems but there is still a division of priate for the range of velocities under investi-
processing power and experienced operators will gation. It should be remembered that colour
still tend to switch between imaging and spectral Doppler only gives information on the mean
Doppler for optimal results. Doppler shift in a pixel; the mean velocity in a
pixel is only calculated when an angle correction
Automatic image optimisation for that pixel is performed. The scale is closely
Many systems now have automatic image opti- related to the pulse repetition frequency, or
misation controls which adjust factors such as sampling rate; this has to be above the Nyquist
time gain compensation, receiver gain and colour limit for the frequency shift being measured. If it
Doppler gain to improve the image. In some cases is set too low then aliasing will occur; this can be
this facility can also be applied to the spectral removed by increasing the colour scale range,
display. These can be useful in getting a also known as the pulse repetition frequency. If
reasonable display but should not be relied upon it is set too high then there will be poor colour
entirely and the operator needs to make the final sensitivity for slower velocities, resulting in
adjustments for the best images. inadequate colour fill-in across the vessel.
COLOUR DOPPLER CONTROLS Colour inversion

This changes the colour map display for mapping
Colour map the direction of blood flow in relation to the
Choose one that has good contrast for the iden- transducer, e.g. red for flow towards the trans-
tification of aliasing. Maps which grade to white ducer and blue for away from the transducer, or
358 at each end, or to very pale colours, do not give vice versa.
Colour gate Colour persistence
Power Doppler controls

This relates to the size of the colour pixels, a This reflects the amount of frame-averaging which
smaller size gives better spatial resolution, whereas occurs. A low level of persistence results in an
a larger size provides better sensitivity at the image which has a fast temporal response but
expense of spatial and sometimes temporal may have a poor signal to noise ratio. A high
resolution. level of persistence improves the signal to noise
ratio by summating data from several frames, but
Colour baseline this results in an impaired temporal response, so
The colour baseline can be adjusted to provide that pulsatile flow information is dampened.
a wider range of shifts on one side or the other,
depending on the characteristics of the blood flow
POWER DOPPLER CONTROLS
in the vessel being examined and particularly if
the flow is in only one direction. Although the power mode is intrinsically simpler
than velocity mode, there are still several
Colour filters controls and options which influence the power
These help remove low-frequency noise and image.
clutter from the image. It is best to set filters at
the lowest level compatible with an acceptable Power maps
image as they also remove low-frequency shift Most systems offer a variety of colour maps
information from the image. from which the operator can select according to
personal preference, usually a yellow- or magenta-
Colour write priority based scale is used for display of power Doppler
A pixel in the image can display either B-mode information. The threshold or sensitivity can be
imaging information, or colour Doppler infor- varied; this results in changes in the trans-
mation. The colour write priority facility allows parency of the power box. When the power map
the relative priorities for these two types of background is opaque the sensitivity is maximal.
information to be defined. High colour priority Various degrees of increasing translucency are
results in colour information being displayed in available so that sensitivity can be traded for
areas which might contain low-intensity imaging imaging and anatomical information.
information, for instance at the margins of
vessels. Alternatively, high imaging priority Power gain
results in grey-scale information displacing This is similar to colour gain and amplifies the
colour information, such as might occur with received Doppler signal. Excessive gain will result
reverberation artefacts appearing within the in unnecessary noise and artefact in the image.
vessel lumen amongst the colour Doppler
information. Scale
The energy scale affects the sensitivity of the
Colour box system for signal intensities of varying strengths:
This defines the volume of tissue from which lower scales are more sensitive to lower-intensity
colour Doppler information will be gathered. It signals; on some systems the scales are linked to
is better to keep the box as small and as super- the filters.
ficial as practical, because larger areas take more
processing power and time; deeper boxes need Dynamic range
slower pulse repetition frequencies. Most systems As in spectral Doppler, this affects the range of
now allow the direction of the box to be steered signal intensities over which the available colour
but this also reduces sensitivity for Doppler scale is spread and the appearance of the Doppler
information. information on the screen: increasing the dynamic 359
Spectral scale
range will tend to increase the amount of colour
on the screen. Altering the scale affects the pulse repetition
frequency and thus the range of shifts which can
Power box steering angle be registered without aliasing. In practice the
Although the power display is much less dependent scale is adjusted so that the Doppler waveform is
on angle than the velocity display, there may still displayed without wrap-around, which indicates
be loss of the power signal at 90° because these aliasing. The scale may be displayed with either
low Doppler frequencies fall below the motion a KHz scale for frequency shift, or m s-1 for
discrimination filter cut-off level. In most situa- velocity. The use of m s-1 allows some com-
tions, steering the power Doppler box does not parison of different examinations, performed
have much effect on the display; however, the with different transducer frequencies and with
loss of signal at 90° may be overcome by having different angles of insonation.
a degree of angulation of the transducer.
Spectral inversion
Filters This allows the operator to change the orien-
Motion discrimination filters are used to filter tation of the display. Many operators prefer to
out excessive signal noise from structures, other display arterial and venous waveforms above the
than blood, which are moving in the Doppler baseline, even if flow is away from the trans-
box. Low filter settings provide more sensitivity ducer. However, care is therefore required when
but are more prone to flash artefacts, whereas assessing the vertebral arteries for reverse flow,
higher filter settings reduce flash artefacts but or the leg veins for reflux, as errors may occur if
will also filter out some blood flow information. spectral inversion is not recognised. Some
centres do not allow spectral inversion because
of the potential for misinterpretation, particu-
SPECTRAL CONTROLS
larly in relation to examinations for venous
Spectral gain insufficiency.
This affects the receiver gain for the spectral
display. As in B-mode imaging, the level should Spectral sweep speed
be adjusted in order to give a balanced distri- A medium speed is adequate for most arterial
bution of grey shades across the displayed spectra. work, with a slower speed for venous flow. A fast
Excessive gain will produce erroneous estimates sweep speed is useful for acceleration time
of Doppler shift/velocity. measurement and waveform analysis, particularly
if there is tachycardia.
Spectral dynamic range
This can be varied to optimise the display of Angle correction
particular frequency shifts. A narrow dynamic The measurement of the angle of insonation
range results in the loss of low-intensity shifts relative to the direction of flow is required in
above and below the main shift frequencies. spectral Doppler in order to convert frequency
Conversely a wide dynamic range, particularly if shift information into velocity information and
associated with a high level of spectral receiver also in colour Doppler to convert mean pixel shifts
gain, can result in artefactual broadening of the into mean velocities. The main direction of flow
spectra displayed. Further manipulation of the may not necessarily be parallel to the vessel wall
spectral display can be performed by altering and colour Doppler is useful in precise posi-
the postprocessing algorithms in order to empha- tioning of the angle-correction cursor along the
sise, or suppress, particular frequency shifts. In line of the jet. The angle of insonation should be
normal practice a simple linear allocation is less than 60° or the errors in velocity calculation
360 most convenient. become significant.
Gate size Filters
Spectral controls
This defines the range of depths from which Removal of low-frequency noise and clutter arising
Doppler data are collected. The gate should from the vessel wall and surrounding tissues
be positioned across the lumen of the vessel contributes to a cleaner signal, but filters should
but clear of the walls in order to reduce wall be set as low as is practical, otherwise low-
thump. The position of the gate which frequency shifts from slow blood flow will be
corresponds to the maximum Doppler shift is filtered out, which could result in the mistaken
located with the use of the colour map and impression of absent diastolic flow in arteries, or
the operator’s ears, which are the most sensitive occlusion in veins. In practice it is best to keep
and efficient spectral analyser available; with the filters set at the lowest setting and only increase
experience, the ears will register when the peak filtration as required during an examination.
frequency shift is obtained. If an assessment of Modern systems allow the operator to preset
volume flow is being made, the gate should many of these parameters and create different
be wide enough to encompass the entire vessel profiles for different types of examination such
width so that all the flow contributes to the as veins, carotids, renal, etc. However, it should
signal and the time-averaged mean velocity be remembered that ultrasound is a dynamic
will be most representative. Smaller gates examination and the best results will be obtained
give a cleaner signal, especially with laminar if the system controls are adjusted to optimum
flow, but at the expense of a reduction in settings for the task in hand, rather than relying
sensitivity. on the preset profiles alone.
361

Clinical Doppler Ultrasound PDF

Uploaded by

Copyright:

Available Formats

Clinical Doppler Ultrasound 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

Clinical Doppler Ultrasound PDF

Uploaded by

Copyright:

Available Formats

Commissioning Editor: Meghan McAteer

Development Editors: Hilary Hewitt and Louise Cook

Paul Allan BSc, DMRD, FRCR, FRCPE

Paul A. Dubbins BSc, MBBS, FRCR

W. Norman McDicken PhD, FIPEM

© 2006, Elsevier Limited. All rights reserved.

First published 2006

First edition 2000

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in

British Library Cataloguing in Publication Data

Library of Congress Cataloging in Publication Data

Paul Allan BSc, DMRD, FRCR, FRCPE Fred T. Lee, Jr, MD

Jonathan D. Berry BSc, MBBS, FRCR W. Norman McDicken PhD, FIPEM

W. Norman McDicken and Peter R. Hoskins

Continuous and pulsed wave Doppler instruments

Tear-drop sample volume

Consecutive sample volumes

Imaging and Doppler

Fig. 1.2 Duplex system combining

Fig. 1.3 (a) Colour flow image

Imaging and Doppler

Ultrasonic microbubble contrast agents

Table 1.1 Properties of some common intravenous, lung-crossing contrast agents

Information from Doppler signals

equals the length of the portion of Doppler

Spectrograms (Sonograms) and indices

High resistance in the distal vessels produces

Pulsatility index (PI)

maximum velocity excursion

Fig. 1.8 Waveform indices,

Spectrogram (Sonograms) and indices

A/ ratio (carotid arteries)

PI = Maximum velocity excursion

Artefacts in Doppler techniques

Harmonic generation by large signal

High or low sensitivity

Artefacts in Doppler techniques

Fig. 1.11 Variation of gain in a Doppler image.

Effect of beam angle to flow

Artefacts in Doppler techniques

Effect of velocity scale

Interference from neighbouring

Power Doppler images are relatively insensitive Vessel compression

Patient or vessel movement

Beam position within the vessel

Fig. 1.16 Variation of sonogram

Safety and prudent use of Doppler instruments

Three-dimensional Doppler flow

Fig. 1.19 Velocity components along beam directions.

Peter R. Hoskins, W. Norman McDicken

Principles of blood flow

where V is the measured velocity in m s-1, and P

Fig. 2.3 Velocity profiles during

Fig. 2.4 Velocity profiles from a

Principles of blood flow

Entrance Outside wall

Axial velocity profile

Secondary flow motions

Fig. 2.8 A simple model of flow