. 1
( 10)


This page intentionally left blank
Applied Radiological Anatomy for Medical Students
Applied Radiological Anatomy for Medical Students is the de¬nitive atlas of
human anatomy, utilizing the complete range of imaging modalities
to describe normal anatomy and radiological ¬ndings.
Initial chapters describe all imaging techniques and introduce the
principles of image interpretation. These are followed by
comprehensive sections on each antomical region.
Hundreds of high-quality radiographs, MRI, CT and ultrasound
images are included, complemented by concise, focused text. Many
images are accompanied by detailed, fully labeled, line illustrations to
aid interpretation.
Written by leading experts and experienced teachers in imaging
and anatomy, Applied Radiological Anatomy for Medical Students is an
invaluable resource for all students of anatomy and radiology.

pau l b u t l e r is a Consultant Neuroradiologist at The Royal London
Hospital, London.

a da m w. m . m i t c h e l l is a Consultant Radiologist at Charing Cross
Hospital, London.

h a r o l d e l l i s is a Clinical Anatomist at the University of London.
Applied Radiological
Anatomy for Medical Students
The Royal London Hospital

Edited by Charing Cross Hospital . M I T C H E L L

University of London
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
Information on this title: www.cambridge.org/9780521819398

© Paul Butler, Adam W. M. Mitchell and Harold Ellis 2007

This publication is in copyright. Subject to statutory exception and to the provision of
relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2007

eBook (EBL)
ISBN-13 978-0-511-36614-7
ISBN-10 0-511-36614-0 eBook (EBL)

ISBN-13 978-0-521-81939-8
ISBN-10 0-521-81939-3

Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
List of contributors vii
Acknowledgments ix

Section 1 The basics

1 An introduction to the technology of imaging 1
thomas h. bryant and adam d. waldman

2 How to interpret an image 17
adam w. m. mitchell

Section 2 The thorax

3 The chest wall and ribs 23
jonathan d. berry and sujal r. desai

4 The breast 31
stella comitis

Section 3 The abdomen and pelvis
Contents 5 The abdomen 36
dominic blunt

6 The renal tract, retroperitoneum and pelvis 47
andrea g. rockall and sarah j. vinnicombe

Section 4 The head, neck, and vertebral column

7 The skull and brain 64
paul butler

8 The eye 81
claudia kirsch

9 The ear 86
claudia kirsch

10 The extracranial head and neck 91
jureerat thammaroj and joti bhattacharya

11 The vertebral column and spinal cord 105
claudia kirsch

Section 5 The limbs

12 The upper limb 113
alex m. barnacle and adam w. m. mitchell

13 The lower limb 129
a. newman sanders

Section 6 Developmental anatomy

14 Obstetric imaging 146
ian suchet and ruth williamson

15 Pediatric imaging 153
ruth williamson

Index 159

Alex M. Barnacle
Department of Radiology, Charing Cross Hospital, London, UK

Jonathan D. Berry
Department of Radiology, King™s College Hospital, London, UK

Joti Bhattacharya mrcp frcr
Institute of Neurological Sciences, Southern General Hospital, Glasgow, UK

Dominic Blunt mrcp frcr
Department of Radiology, Charing Cross Hospital, London, UK

Thomas H. Bryant mbchb mmedsci frcr
Department of Imaging, Hammersmith Hospitals NHS Trust, London, UK

Paul Butler mrcp frcr
The Royal London Hospital, Department of Neuroradiology, London, UK

Contributors Stella Comitis mbbch frcr
Department of Radiology, Charing Cross Hospital, London, UK

Sujal R. Desai md mrcp frcr
Department of Radiology, King™s College Hospital, London, UK

Claudia Kirsch ba md frcr
Diagnostic Neuroradiology and Head and Neck Radiology, David Geffen School
of Medicine at UCLA, Los Angeles CA, USA

Adam W. M. Mitchell mbbs frcs frcr
Department of Radiology, Charing Cross Hospital, London, UK

A. Newman Sanders mbbs mrcp frcr
Department of Diagnostic Imaging, Mayday University Hospital NHS Trust,
Thornton Heath, Surrey, UK

Ian Suchet
Department of Radiology, Charing Cross Hospital, London, UK

Andrea G. Rockall bsc mbbs mrcp frcr
Department of Radiology, Barts and the London NHS Trust, Barts and The
London School of Medicine, Department of Nuclear Medicine, London, UK

Jureerat Thammaroj md msc
Srinagarind Hospital, Khon Kaen University, Thailand

Sarah J. Vinnicombe bsc (hons) mrcp frcr
Department of Radiology, Barts and the London NHS Trust, Barts and The
London School of Medicine, Department of Nuclear Medicine, London, UK

Ruth Williamson
Department of Radiology, Charing Cross Hospital, London, UK

Adam D. Waldman phd mrcp frcr
Department of Radiology, Charing Cross Hospital, London, UK

Acknowledgments The editors and publisher wish to acknowledge with thanks the excel-
lent work of Mr. Jack Barber, Dr. Jo Bhattacharya and Dr. Ivan Moseley
in the preparation of the line drawings, which illustrate the radiology
images. Some of these line drawings have been redrawn and adapted
from originals which appeared in Grant™s Atlas of Anatomy© Williams &
Wilkins and from Langman™s Medical Embryology© Williams & Wilkins
and we are grateful for the permission of the publisher to allow their
adaptation in this work.

Section 1 The basics

Chapter 1 An introduction to the technology
of imaging

and A DA M D . WA L D M A N

Introduction wife™s hand showing the bones and her wedding ring, requiring an
exposure time of about 30 minutes. Within a month of this discovery,
Imaging techniques available to the radiologist are changing rapidly,
X-rays were being deliberately generated in a number of countries,
due largely to advances in imaging and computer technology. Three
and were being used for imaging patients by early 1896. A modern
of the ¬ve imaging modalities described in this chapter did not exist
X-ray machine is shown in Fig. 1.1.
in recognizable form 30 years ago. This chapter is a brief overview of
the major medical imaging techniques in current use with reference
X-ray generation
to the underlying principles, equipment, the type of information that
The basics of the X-ray tube have remained unchanged since
they yield, and their advantages and limitations.
Roentgen™s time, although the details have changed. X-rays are made
up of photons and are a type of electromagnetic radiation like light or
X-rays radio-waves, although they have higher energy.
The basic X-ray tube is a vacuum tube (Fig. 1.2). A high voltage is
X-rays were discovered by a physicist named Wilhelm Roentgen in
passed through a wire, heating it and allowing electrons to be freed
November 1895, using a type of cathode ray tube invented in 1877 by
and leave the wire at its surface (the cathode). The electrons are accel-
Crooke. With this “new kind of ray,” he produced a photograph of his
erated towards a second electrode with a positive charge (the anode)
causing a current to ¬‚ow between the cathode and anode. If the anode

Tungsten filament

Tungsten target


Fig. 1.1. An example of Glass
vacuum tube
a ¬‚uoroscopy machine
that uses X-radiation
to produce images of
patients. The tube can
be rotated around the
patient to provide
views from different
projections. Moving
images can be viewed
using the image
intensi¬er or static
Fig. 1.2. The essentials of a simple, ¬xed anode X-ray generation set.
images can be obtained.

Applied Radiological Anatomy for Medical Students. Paul Butler, Adam Mitchell, and Harold Ellis (eds.) Published by Cambridge University Press. © P. Butler,
A. Mitchell, and H. Ellis 2007.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

Fig. 1.3. Diagrams of the (b) Fig. 1.3. (b) Characteristic
production of X-rays. radiation. When a free
(a) Bremsstrahlung or electron knocks one of
Braking radiation the “cloud” of orbital
occurs when the free shell electrons out of an
electron is de¬‚ected by atoms, an electron from
e“ e“
the electric ¬eld around a higher energy (outer)
the nucleus of a target shell moves to ¬ll the
Nucleus atom, shedding energy gap, shedding the
in the form of a photon excess energy in the
as the free electron is form of an electromag-
slowed. netic photon which will
be an X-ray photon if
the energies are high
enough. These X-rays
is made of the correct material and the electrons are accelerated
have an energy spe-
enough (by at least 1000 volts), X-rays will be produced. Typical mate- ci¬c to the transition
rials used for the anode include tungsten and molybdenum, which between the shells,
have high atomic numbers, and high melting points (the X-ray tube and the pattern of
production is therefore
gets very hot). Over 90% of the energy supplied is lost as heat.
characteristic of the
X-ray photons are produced at the anode when a free electron trav-
anode material.
elling at high speed interacts with a target atom. Two main interac-
tions occur in the diagnostic X-ray energy range, Bremsstrahlung and
characteristic radiation (Fig. 1.3).
The X-rays then leave the tube through a ¬lter (usually made of
Detection of X-rays
copper or molybdenum), which removes X-ray photons with undesir-
Following irradiation of the patient, some of the X-rays are absorbed,
able energies, leaving those in the diagnostic range.
some are scattered (de¬‚ected) and some pass through the patient.
Finally, the X-rays pass through a collimator. X-rays produced at the
These effects depend on the nature and thickness of the tissues in
anode travel in all directions, although some features of the design
their path.
cause them to mainly be directed towards the patient. The collimator
X-ray photons are invisible. There are a number of mechanisms
is an aperture (usually made of lead) that can be opened and closed so
to detect X-ray photons and convert them to a visible image
that only the part of the patient to be imaged is exposed to the X-ray
(Fig. 1.5).

How X-rays produce an image Film
Production of a radiograph, an X-ray image, is the result of the interac- Although photographic ¬lm is sensitive to X-rays by itself, ¬‚uores-
tion of X-ray photons with the patient and detection of the remaining cent screens are used inside X-ray cassettes that convert X-ray
photons. photons to visible light, decreasing the number of X-ray photons
required to make an image and therefore the radiation dose to the
X-ray interactions patient. The light produced then exposes the photographic ¬lm by
There are two main types of interaction that are important in the converting crystals of silver halide into elemental silver. These
diagnostic X-ray range (Fig. 1.4). Photoelectric absorption is more initial specks of silver are grown during processing, and appear
important at low energy (low kV) X-ray photon energies and is seen black on the ¬lm.
more with elements with high atomic numbers “ such as calcium in
bones. Compton (incoherent) scattering becomes more important for
biological tissues as X-ray photon energies increase (high kV) and is
proportional to tissue density.
(b) Fig. 1.4. (b) Compton
(incoherent) scattering
occurs when the X-ray
photon interacts with
an atomic electron,
Fig. 1.4. A representation
resulting in de¬‚ection
of the two important

of the photon with a
types of X-ray (and -ray)
transfer of kinetic
interaction with
energy to the electron.
biological tissue.
This is known as
(a) Photoelectric
scattering as the X-ray
absorption occurs
photon continues in a
when an X-ray photon
Carbon atom
different direction
with suf¬cient energy
(which can even be the
is absorbed, breaking
reverse of the original
the bond of an atomic
direction, in the case of
electron and knocking it
a head on collision).
out of the electron shell.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

(a) (b)

Fig. 1.5. A radiograph (“plain ¬lm”) of the chest. This has been acquired on a CR system using an X-ray generation set and europium-activated barium ¬‚uorohalide
plate read by a laser. Both PA (postero-anterior) and lateral views are shown. The views are named from the direction the X-rays pass through the patient and the
location of the detector: in the case of the PA ¬lm the X-ray tube is behind the patient and the detector plate in front so the X-rays pass from posterior to anterior.

Computed radiology (CR) by X-ray photons. These are then converted to electrons, focused using
an electron lens and accelerated towards an anode where they strike
Special plates are made from europium-activated barium ¬‚uoro-
an output phosphor producing light, that is then viewed by a video
halides. These plates absorb the X-ray photons emerging from the
camera and transmitted to viewing screen or ¬lm exposure system.
patient, storing them as a latent image. The plates are then scanned
Fluoroscopy allows real-time visualization of moving anatomic struc-
with a laser, causing emission of light that can be read by a light
tures and monitoring of radiological procedures such as barium
detecting photo-multiplier tube connected to a computer on which
studies and angiography.
the image can be viewed.

Advantages and limitations of plain X-ray
Digital radiology (DR)
Plain radiography is readily available in the hospital setting and
A number of devices for direct digital acquisition of images exist.
is frequently the ¬rst line of imaging investigation. It has a higher
CCD (charged coupled device) technology such as is found in modern
spatial resolution than all other imaging modalities. It is most useful
digital cameras can be adapted to detect X-rays by coating the device
for structures with high-density contrasts between tissue types, partic-
with a visible light producing substance such as cesium iodide or by
ularly those tissues in which ¬ne detail is important, such as in
using a ¬‚uorescent screen. TFT (thin ¬lm transistor) detectors consist of
viewing bone, and in the chest. Plain radiography is relatively poor
arrays of semiconductor detectors, and another method uses a detector
for examining soft tissues, due to its limited contrast resolution.
such as amorphous selenium or cesium iodide to capture the photons
It is possible to distinguish only four natural densities in diagnostic
with amorphous silicon plates to amplify the signal produced.
radiography: calcium (bone), water (soft tissue), fat, and air. Plain
Digital and computed radiology techniques are being used increas-
¬lm radiography provides a two-dimensional representation of three-
ingly in clinical departments, with a consequent reduction in the use
dimensional structures; all structures projected in a direct line
of photographic ¬lm.
between the X-ray tube and the image receptor will overlap. This
Fluoroscopy “ image intensi¬er can be partially overcome by obtaining views from different angles,
or by turning the patient or the X-ray tube and image intensi¬er in
Image intensi¬ers use a ¬‚uoroscopic tube to form an image. The input
screen is covered with a material that emits light photons when hit

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

Conventional tomography
Simultaneously moving both the X-ray tube and the ¬lm about a pivot
point causes blurring of structures above and below the focal plane.
Objects within the focal plane show increased detail because of the
blurring of surrounding structures, providing an image of a slice of
the patient (Fig. 1.6). Movements of the X-ray tube and ¬lm can be
linear, elliptical, spiral, or hypocycloidal. With the advent of cross-
sectional imaging techniques such as CT and MRI, most imaging
departments now only use linear tomography, as part of an intra-
venous urogram (see below).

Contrast enhancing agents
To allow visualization of speci¬c structures using X-rays, a number
of contrast agents have been used. A good contrast agent should
increase contrast resolution of organs under examination without poi-
soning or otherwise damaging the patient. The best contrast agents
for use with X-rays have a high atomic weight as these have a high
proportion of photoelectric absorption in the diagnostic X-ray range.
Unfortunately, most molecules that contain these atoms are very
toxic. Iodine (atomic weight 127) is the only element that has proved
satisfactory for general intravascular use; extensive research and
development has resulted in complex iodinated molecules that are
non-toxic, hypoallergenic and do not carry too great osmotic load. The
normal physiological turnover of iodine in the body is 0.0001 g per
Fig. 1.7. Barium enema. Barium sulphate has been introduced into the large
day, while for typical imaging applications 15 g to 150 g or 150 000“1
bowel by a tube placed in the rectum and carbon dioxide gas is then used to
500 000 times as much may be required. Barium sulphate (atomic
expand the bowel, leaving a thin coating of barium on its inside surface. X-ray
weight 137), and iodinated compounds are the only agents in regular
images are used to examine the lining of the bowel for abnormal growths and
use as extravascular agents. other abnormalities.

Barium studies
Barium is only used in a modern X-ray department for studies of the Intravenous urography
gastrointestinal tract. These are usually based on a ¬‚uoroscopic The kidneys rapidly excrete Iodinated contrast agents. Plain radi-
image intensi¬er on which a moving image can be seen. Studies can ographs taken from just a few seconds after a contrast injection into
be performed of the swallowing mechanism and esophagus (barium a peripheral vein show the passage of contrast through the kidney,
swallow), the stomach and duodenum (barium meal), the small bowel into the ureters and to the bladder (Fig. 1.8).
(small bowel follow through or small bowel enema) and the colon
(barium enema). Studies of the stomach and large bowel are usually Angiography
“double contrast” which allows better visualization of surface detail. A specially shaped, thin catheter (tube) can be introduced into the
Air or carbon dioxide can be introduced into the large bowel and arterial or venous system and manipulated using ¬‚uoroscopy to
gas-forming granules (usually a combination of calcium carbonate almost any blood vessel large enough to have been named. Contrast
and citric acid) can be swallowed for imaging the stomach, resulting introduced through these catheters by hand or mechanical injection
in a thin barium coating of the bowel mucosa (Fig. 1.7). will be carried in the bloodstream and allows very detailed imaging
of the vascular system. The arterial system is usually accessed via
puncture of the femoral artery in the groin, although arteries of the
upper limb may occasionally be used. Digital subtraction angiography
(DSA) is most commonly performed “ an initial (“mask”) image is
X-ray tube
taken before the contrast agent is administered and is “subtracted”
from later images. This removes the image of the tissues, leaving
the contrast-¬lled structures. Any movement after the mask image
is taken destroys the subtracted image. Because angiography is
potentially hazardous, the balance between the potential bene¬t and
the risk of the procedure (damage to vessels and other structures,
Focal plane
bleeding) must be evaluated with particular care before undertaking
the procedure (Fig. 1.9).
X-ray table

Radiation dose
All ionizing radiation exposure is associated with a small risk. A small
Fig. 1.6. Conventional tomography. The X-ray tube and ¬lm move simultaneously
proportion of the genetic mutations and cancers occurring in the pop-
about a pivot point at the level of the focal plane, blurring structures outside
ulation can be attributed to natural background radiation. Diagnostic
the focal plane, and emphasizing the structure of interest.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

(a) (b)

Fig. 1.8. Intravenous urogram showing (a) standard view of the kidneys and upper part of the urinary collecting system and (b) linear tomogram of the intrarenal
collecting system. This blurs out the overlying structures, giving a clearer image of the collecting system and renal outline. An injection of 50 ml of iodine-
based contrast medium has been given and these radiographs have been obtained 10“15 minutes later after it has passed through the kidneys and into the
renal collecting system.

(a) (b)

Fig. 1.9. Renal angiogram. (a) A catheter has been inserted through the right femoral artery into the aorta, (b) iodinated contrast medium has been injected through it,
and a rapid sequence of radiographs taken. Digital subtraction of the background shows the passage of contrast medium through the arteries supplying both kidneys.

medical exposures (using X-rays or -rays, see Nuclear Medicine below) considered carefully, and the clinician directing the procedure (usually
are the largest source of man-made radiation exposure to the general the radiologist) is accountable in law for any radiation exposure.
population and add about one-sixth to the population dose from back-
ground radiation. The dose is calculated as “effective dose,” which is
a weighted ¬gure depending on the sensitivity of the body tissues
General principles
involved to radiation induced cancer or genetic effects. Typical doses
are given in Fig. 1.10. Children and the developing fetus are particu- Ultrasound is sound of very high frequency. In most diagnostic appli-
larly susceptible to radiation damage. As with all medical investiga- cations frequencies between two million and twenty million cycles
tions and procedures, the relative risks and potential bene¬ts must be per second are used, 100“1000 times higher than audible sound.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

ultrasound wave, so are also used as the receiver. A modern ultra-
sound probe contains an array of several hundred tiny piezoelectric
Procedure Typical effective Equivalent Equivalent period of
dose (mSv) number natural background crystals with metal electrodes on their two surfaces, the sound lenses
of chest X-rays radiation and matching layers required to form the beam shape and electronics.
Piezoelectric crystals can also be found in the speakers inside in-ear
Limbs and joints 1.5 days
0.01 0.5
headsets, quartz watches, and camera auto-focus mechanisms.
Chest 3 days
0.02 1
Lumbar spine 7 months
1.3 65
Image formation
Pelvis 4 months
0.7 35
Abdomen 6 months Ultrasound travels at near constant speed in soft tissues and this
1.0 50
IVU 14 months
2.5 125 allows the depth of re¬‚ectors to be calculated by measuring the delay
Barium enema 3.2 years
7 350
between transmission of the pulse and return of the echoes.
CT head 1 year
2.3 115
CT chest 3.6 years
8 400
CT abdomen 4.5 years
10 500
The tissues absorb ultrasound when the orderly vibration of the sound
or pelvis
Bone scan 1.8 years wave becomes disordered in the presence of large molecules. When
4 200
PET head (FDG) 2.3 years
5 250 this happens, sound energy is converted to heat energy. Absorption
depends on the molecular size, which correlates with viscosity of the
tissue, and with the frequency. Higher frequencies are more strongly
Fig. 1.10. Typical effective doses for some of the commonly performed Imaging
absorbed, so less depth of scanning comes with the improvement in
investigations. The typical United Kingdom background radiation dose is
2.2 mSv/year (ranges from 1.5 to 7.5 mSv/year depending on geographical resolution that higher frequencies allow. Ultrasound energy is also
location). It has been estimated that the additional lifetime risk of a fatal cancer lost to the transducer if it is re¬‚ected or refracted away.
from an abdominal CT scan could be as much as 1 in 2000 (although the overall
lifetime risk of cancer for the whole population is 1 in 3).
Some of the ultrasound beam is re¬‚ected whenever it crosses an inter-
face where the transmission properties change. This is directly related
to the physical structure of the tissues on either side of the interface.

Tissue harmonics
Ultrasound is generally considered to be conducted in a linear fashion
with no change in the waveform of the pulse as it travels through the
tissues. In fact, the wave originating from the transducer becomes
distorted as the speed of sound conduction changes with the density
of the conducting materials allowing some parts of the wave to travel
faster than others. The wave comes to contain higher frequency
components, called harmonics, which are much weaker in the parts of
the sound beam away from the central echoes. Scanners can transmit
at one frequency, receive at a higher frequency and use ¬lters to select
out the harmonics in the returning echoes, improving the image
resolution and increasing the contrast.

Image display
Gray-scale or B-Mode (B for brightness) is a two-dimensional real
time image formed by sweeping the beam through the tissue. The
echogenicity of the re¬‚ectors is displayed as shades of gray and is the
main mode used for ultrasound imaging (Fig. 1.12). Modern ultrasound
machines operate at a suf¬cient speed to produce real-time images
of moving patient tissue such as the heart in echocardiography and
Fig. 1.11. A diagnostic ultrasound machine.
the moving fetus.

Doppler ultrasound
Higher frequencies have shorter wavelengths, allowing greater spatial
resolution of structures being studied. An example of an ultrasound If a sound wave re¬‚ects from a moving target, there is a change in the
machine is shown in Fig. 1.11. frequency of the returning sound wave proportional to the velocity
of the re¬‚ecting target. This is known as the Doppler effect and the
Ultrasound transducers changes in frequency can be used to calculate the velocity of the
Ultrasound is generated by piezoelectric materials, such as lead zir- moving target usually ¬‚owing blood. The Doppler signal is within
conate titanate (PZT). These have the property of changing in thick- the audible range, so can be heard by sending the signal to a loud-
ness when a voltage is applied across them. When an electrical pulse speaker. Most commonly used in clinical practice is color ¬‚ow imaging
is applied, the piezoelectric crystal produces sound at its resonant (color Doppler) where ¬‚ow information is shown as an overlay on the
frequency. These crystals also generate a voltage when struck by an gray-scale image with the color and shading indicating the direction

An introduction to the technology of imaging thomas h. bryant and adam d. waldman

Mirror image artifact
A strong re¬‚ector can cause duplication of echoes, giving the appear-
ance of duplication of structures above and below the re¬‚ector.

“Ring down” artifact
A pattern of tapering bright echoes trailing from small bright
re¬‚ectors such as air bubbles.

Advantages and limitations of ultrasound
Ultrasound provides images in real time so can be used to image
movement of structures such as heart valves and patterns of blood
¬‚ow within vessels. As far as is known, ultrasound used at diagnostic
intensities does not cause tissue damage and can be used to image
sensitive structures such as the developing fetus. Patients usually ¬nd
ultrasound examination easy to tolerate, as it requires minimal prepa-
ration and only light pressure on the skin. Portable ultrasound
systems suitable for use at the bedside are widely available.
The main limitation of the technique is that parts of the body acces-
sible to ultrasound examination are limited. Ultrasound does not
Fig. 1.12. A stone within the gall bladder shows as a bright echo with black easily cross a tissue“gas or tissue“bone interface, so can only be used
“acoustic shadow” behind it, the result of almost complete re¬‚ection of the
for imaging tissues around such structures with any tissues deep to
ultrasound hitting it. The ¬‚uid in the gall bladder appears black as the contents
gas or bone obscured. It is not generally useful in the lungs and head,
of the gall bladder are homogeneous and there are no internal structures to
except in neonates where the open fontanelles provide an acoustic
cause echoes or changes in attenuation; the adjacent liver is more complex in
window. Ultrasound is also heavily operator dependent, particularly
structure and causes more re¬‚ection of sound, so appears gray.
in overcoming barriers due to the bony skeleton and bowel gas, and
in interpreting artifacts, which are common.
and velocity of ¬‚ow. Spectral Doppler is a graphical display with time
on the horizontal axis, frequency on the vertical axis and brightness
Computed tomography
of the tracing indicating the number of echoes at each speci¬c fre-
quency (and therefore blood cell velocity). A combined gray-scale and Computed tomography (CT) was invented in the 1970s, earning its
spectral Doppler display is known as a duplex scan. Power Doppler chief inventor, Sir Godfrey Houns¬eld, the Nobel Prize for medicine
imaging discards the direction and velocity information but is about in 1979. CT was the ¬rst fully digital imaging technique that provided
10 more sensitive to ¬‚ow than normal color Doppler. cross-sectional images of any anatomical structure.
Doppler ultrasound is used to image blood vessels and to examine
Basic principles
tissues for vascularity (¬g. 1.13 “ see color plate section).
Current generation CT scanners use the same basic technology as
Ultrasound contrast agents the ¬rst clinical EMI machine in 1972. In conventional CT, the X-ray
Contrast agents have been developed for ultrasound consisting of tiny tube and detector rotate around the patient with the table stationary.
“microbubbles” of gas small enough to cross the capillary bed of the The X-ray beam is attenuated by absorption and scatter as it passes
lungs. These are safe for injection into the bloodstream and are very through the patient with the detector measuring transmission
highly re¬‚ective; they can be used to improve the imaging of blood (¬g. 1.14). Multiple measurements are taken from different directions
vessels and to examine the ¬lling patterns of liver lesions. as the tube and detector rotate. A computer reconstructs the image
for this single “slice.” The patient and table are then moved to the
Ultrasound artifacts next slice position and the next image is obtained.
Acoustic shadowing
Produced by near complete absorption or re¬‚ection of the ultrasound
Fig. 1.14. Diagram of a
beam, obscuring deeper structures. Acoustic shadows are produced by
typical CT scanner. The
bone, calci¬ed structures (such as gall bladder and kidney stones), gas
patient is placed on
in bowel, and metallic structures.
the couch and the X-ray
X-ray tube
tube rotates 360° around
Acoustic enhancement the patient, producing
Structures that transmit sound well such as ¬‚uid-¬lled structures pulses of radiation that
pass through the patient.
(bladder, cysts) cause an increased intensity of echoes deep to the
The detectors rotate
with the tube, on the
other side of the patient
Reverberation artifact Detector
detect the attenuated
Repeated, bouncing echoes between strong acoustic re¬‚ectors cause X-ray pulse. This data is
multiple echoes from the same structure, shown as repeating bands sent to a computer for
of echoes at regularly spaced intervals. reconstruction.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

The use of intravenous contrast agents can increase the contrast res-
In spiral (helical) CT the X-ray tube rotates continuously while the
olution in soft tissues as different tissues show differences in enhance-
patient and table move through the scanner. Instead of obtaining data
ment patterns. Oral contrast can outline the lumen of bowel and
as individual slices, a block of data in the form of a helix is obtained.
allow differentiation of bowel contents and soft tissues within the
Scans can be performed during a single breath hold, which reduces
abdomen. Usually iodinated contrast agents are used for CT, although
misregistration artifacts, such as occur when a patient has a different
a dilute barium solution can be used as bowel contrast.
depth of inspiration between conventional scans. A typical CT scanner
is shown in Fig. 1.15.
Window and level
Image reconstruction The human eye cannot appreciate anywhere near the 4000 or so gray
To convert the vast amount of raw data obtained during scanning to scale values obtained in a single CT slice. If the full range of recon-
the image requires mathematical transformation. Depending on the structed values were all displayed so as to cover all perceived
parameters used (known as “kernels”), it is possible to get either a brightness values uniformly, a great deal of information would be lost
high spatial resolution (at the expense of higher noise levels) used for as the viewer would not be able to distinguish the tiny differences
lung and bone imaging, or a high signal to noise ratio (at the expense between differing HU values. By restricting the range of gray scale
of lower resolution) used for soft tissues. information displayed, more subtle variations in intensity can be
The CT image consists of a matrix of image elements (pixels) usually shown. This is done by varying the range (“window width”) and
256 256 or 512 512 pixels. Each of these displays a gray scale inten- centre (“window level”) (Fig. 1.16).
sity value representing the X-ray attenuation of the corresponding
block of tissue, known as a voxel (a three-dimensional “volume
Spiral CT and pitch
In conventional, incremental CT the parameters describing the proce-
CT scanners operate at relatively high diagnostic X-ray energies, in
dure are slice width and table increment (the movement of the table
the order of 100 kV. At these energies, the majority of X-ray-tissue
between slices). With spiral CT, the patient, lying on the couch, moves
interactions are by Compton scatter, so the attenuation of the X-ray
into the scanner as the tube and detectors rotate in a continuous
beam is directly proportional to the density of the tissues. The inten-
movement, rather than the couch remaining still while each “slice” is
sity value is scored in Houns¬eld units (HU). By de¬nition, water is
acquired. The information during spiral CT is obtained as a continu-
0 HU and air 1000 HU and the values are assigned proportionately.
ous stream and is reconstructed into slices.
These values can be used to differentiate between tissue types. Air
The parameters for spiral CT are slice collimation (the width of the
( 1000 HU) and fat ( 100 HU) have negative values, most soft tissues
X-ray beam and therefore the amount of the patient covered per rota-
have values just higher than water (0 HU), e.g., muscle (30 HU),
tion), table feed per rotation, and the reconstruction increment.
liver (60 HU), while bone and calci¬ed structures have values of
A spiral CT covers the whole volume even if the table feed is greater
200“900 HU. The contrast resolution of CT depends on the differences
than the collimation “ it is possible to scan with a table feed up to
between these values, the larger the better. Although better than plain
twice the collimation without major loss of image quality. Often,
X-ray in differentiating soft tissue types, CT is not a good as magnetic
scans are described by their pitch where pitch table feed/collima-
resonance imaging (MRI). For applications in the lungs and bone
tion. Typical values for collimation (slice thickness) are 1“10 mm with
(where the differences in attenuation values are large), CT is generally
rotation times of 0.5“3 seconds.
better than MRI.
To reconstruct from the helical volume, it is necessary to interpolate
the projections of one scanner rotation. It is not necessary to recon-
struct as consecutive slices “ slices with any amount of overlap can be

Multi-detector CT
CT scanners are now available with multiple rows of detectors (at
the time of writing, commonly 64) allowing acquisition of multiple
slices in one spiral acquisition. In conjunction with fast rotation
speeds, the volume coverage and speed performance are improved
allowing, for instance, an abdomen and pelvis to be scanned with an
acquisition slice thickness of 1.25 mm in about quarter the time
(approximately 10 seconds) that a 10 mm collimation CT scanner
could cover the same volume, with the same or lesser radiation dose.
The main problem with this type of scanning is the number of
images acquired; 300“400 in the example above instead of about 40
with single slice techniques.

Advanced image reconstructions
From the spiral dataset, further reconstructions can be performed.
Multiplanar reformats (MPR) can be performed in any selected plane,
although usually in the coronal and sagittal planes (Fig. 1.17). Three-
Fig. 1.15. A multi-slice CT
dimensional reconstructions can also be obtained using techniques

An introduction to the technology of imaging thomas h. bryant and adam d. waldman

(a) 1500 (b) 1500


(c) 1500 (d) 1500


Fig. 1.16. The effect of changing window levels and reconstruction algorithm on a single axial image through the chest. The dark bar indicated the range of values
displayed, the light bar the range of values available. (a) “Soft tissue” window with window level of 350 and centre 50; (b) “bone window” with window level 1500
and centre 500; (c) lung window with window level 1500 and centre 500; and (d) an HRCT (high resolution CT image) “ this is a thin slice image reconstructed
using an edge enhancement (bone or lung) algorithm, which shows better detail in the lung but increases “noise” levels, window 1500, centre 500.

such as surface-shaded display and volume rendering (Fig. 1.18 “ see often discarded. Virtual endoscopy uses a 3-D “central” projection to
color plate section). While the 3-D techniques provide attractive give the effect of viewing a hollow viscus interiorly (as is seen in
images and are useful in giving an overview of complex anatomical endoscopic examination) and is of particular use in patients too frail
structures, a lot of information from the original axial data set is or ill to have invasive endoscopy.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

(a) (b)

Fig. 1.17. (a) Sagittal and (b) coronal reformats of a helical scan through the abdomen and pelvis. The data from the axial slices is rearranged to give different

HRCT Streak artifact
High resolution CT or HRCT is used to image the lungs. Thin slices The reconstruction algorithms cannot deal with the differences in
are acquired “ usually 1 to 2 mm thick at 10“20 mm intervals. These are X-ray attenuation between very high-density objects such as metal
reconstructed using edge enhancement (bone or lung) algorithms clips or ¬llings in the teeth and the adjacent tissues and produce high
showing better detail in the lung but increasing “noise” levels (Fig. 1.16). attenuation streaks running from the dense object (Fig. 1.19).
This allows ¬ne details of lung anatomy to be seen. The whole lung
Advantages and limitations of CT
volume is not scanned, as there are gaps between the slices.
CT provides a rapid, non-invasive method of assessing patients.
A whole body scan can be performed in a few seconds on a modern
CT artifacts
multislice scanner with very good anatomical detail. CT is particu-
Volume averaging
larly suited to high X-ray contrast structures such as the bones and
A single CT slice of 10 mm thickness can contain more than one tissue
the lungs, and remains the cross-sectional imaging modality of
type within each voxel (for example, bone and lung). The CT number
choice for assessing these. It has less contrast resolution than MRI
for that voxel will be an average of the different sorts of tissue within
for soft tissue structures particularly for intracranial imaging,
it, so very small structures can be “averaged out” or if a structure with
spinal imaging, and musculoskeletal imaging. CT has no major
low CT number is adjacent to one with a high CT number, the appar-
contraindications (although the use of contrast might have), provid-
ent tissue density will be somewhere in between. This is known as
ing the patient can tolerate the scan. The major disadvantage is in
a “partial volume effect.”
the signi¬cant radiation doses required for CT. An abdominal or
pelvic CT involves 3“12 mSv of radiation, compared with a chest
Beam hardening artifact
X-ray™s 0.02 mSv or background radiation in the UK averaging
This results from greater attenuation of low-energy photons than
2.5 mSv per year.
high-energy photons as the beam passes through the tissue. The
average energy of the X-ray beam increases so there is less attenuation
Magnetic resonance imaging (MRI)
at the end of the beam than at the beginning, giving streaks of low
density extending from areas of high density such as bones.
Nuclear magnetic resonance was ¬rst described in 1946 as a tool for
determining molecular structure. The ability to produce an image
Motion artifact based on the distribution of hydrogen nuclei within a sample, the
This occurs when there is movement of structures during image basis of the modern MRI scanner, was ¬rst described in 1973 and the
acquisition and shows up as blurred or duplicated images, or as ¬rst commercial body scanner was launched in 1978. A modern MRI
streaking. scanner is shown in Fig. 1.20.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging


Fig. 1.19. (a) Movement artifact in a CT head scan. There is blurring and streaking following movement of the head. (b) Streak artifact from screws and rods used
to immobilize the lumbar spine.

“net magnetic moment,” such as phosphorus 31, can also be used. As
most protons in biological tissues are in water, clinical MRI is mainly
about imaging water.
The protons in the patient™s tissues can be thought of as containing
tiny bar magnets, which are normally randomly oriented in space.
The patient is placed within a strong magnetic ¬eld, which causes a
small proportion (about two per million) of the atomic nuclei to align
in the direction of the ¬eld and spin (precess) at a speci¬c frequency.
Current magnets typically use a 1.5 tesla ¬eld, about 30 000 times the
earth™s natural magnetic ¬eld. When radio waves (radio frequency, RF)
are applied at the speci¬c (resonance) frequency, energy is absorbed
by the nuclei, causing them all to precess together, and causing some
to ¬‚ip their orientation. When the transmitter is turned off, these ¬‚ip
back to their equilibrium position, stop precessing together and emit
radiowaves, which are detectable by an aerial and ampli¬ed electroni-
cally. The frequency of resonance is proportional to the magnetic ¬eld
that the proton experiences.
The signal is localized in the patient by the use of smaller magnetic
¬eld gradients across and along the patient (in all three planes). These
Fig. 1.20. A magnetic resonance (MR) scanner. cause a predictable variation in the magnetic ¬eld strength and in
the resonant frequency in different parts of the patient. By varying
the times at which the gradient ¬elds are switched on in relation to
Basic principles applying radio frequency pulses, and by analysis of the frequency and
Detailed explanation of the complicated physics of MRI is beyond the phase information of the emitted radio signal, a computer is able to
scope of this chapter. More detailed descriptions of MRI, using a rela- construct a three-dimensional image of the patient.
tively accessible and non-mathematical approach, may be found in the The proton relaxes to a lower energy state by two main processes,
recommended texts for further reading below. called longitudinal recovery (which has a recovery time, T1) and trans-
MRI involves the use of magnetic ¬elds and radio waves to produce verse relaxation (with a relaxation time, T2), and re-emits its energy
tomographic images. Normal clinical applications involve the imaging as radiowaves. The relative proportions of T1 and T2 vary between
of hydrogen nuclei (protons) only, although other atoms possessing a different tissues.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging



short T1 time of about 230 ms. T2 relaxation times largely depends
on tiny local variations in magnetic ¬eld due to the presence of
neighbouring nuclei. In pure water, T2 times are long (similar to T1
times); in solid structures there is very much more effect from the
neighbouring nuclei and T2 times can be only a few milliseconds.
By altering the pulse sequence and scanning parameters, one or
other process can be emphasized, hence T1 weighted (T1W) scans
where signal intensity is most sensitive to changes in T1, and T2
weighted (T2W) scans where signal intensity is most sensitive to
changes in T2. This allows signal contrast between different normal
tissue types to be optimized, such as gray and white matter and cere-
brospinal ¬‚uid in the brain, and pathological foci to be accentuated.
There are a number of ways in which the magnetic ¬eld gradients
and RF pulses can be used to generate different types of MR images

T1 and T2 weighting and proton density
Standard spin echo sequences produce standard T1 weighted (T1W), T2
weighted (T2W) and proton density (PD) scans. T1W scans traditionally
provide the best anatomic detail. T2W scans usually provide the most
sensitive detection of pathology. Proton density-weighted images
make T1 and T2 relaxation times less important and instead provide
information about the density of protons within the tissue.
In the brain, cerebrospinal ¬‚uid (mainly water) is dark on T1W scans
and bright on T2W scans (Fig. 1.21).
Fig. 1.21. (a) Coronal T1W, (b) sagittal T2W and (c) axial FLAIR slices through
the brain. Cerebrospinal ¬‚uid is low signal (black) on the T1W and FLAIR images
Inversion recovery (IR) sequences
but high signal (white) on the T2W image.
These sequences emphasize differences in T1 relaxation times of
tissues. The MR operator selects a delay time, called the inversion
T1 times are long in water and shorten when larger molecules are time, which is added to the TR and TE settings. Short tau (T1) inver-
present so cerebrospinal ¬‚uid (which is largely water) has a T1 time sion time (STIR) sequences are the most commonly used and suppress
of about 1500 milliseconds, while muscle (which has water bound to the signal from fat while emphasizing tissues with high water content
proteins) has a T1 time of 500 milliseconds and fat (which has its own as high signal, including most areas of pathology. Fluid attenuated
protons, much more tightly bound than those in water) has a very inversion-recovery (FLAIR) sequences have a longer inversion time and

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

(a) (b)

Fig. 1.22. MR images of the upper part of the thorax showing the brachial plexus, demonstrating the effects of fat suppression. On the T1W sequence (a), the fat is
high signal (white) and on the STIR sequence (b) the signal from fat is reduced.

are used to image the brain as they null the signal from cerebrospinal
¬‚uid, improving conspicuity of pathology in adjacent structures. FLAIR
images are mostly T2 weighted but CSF looks darker (Fig. 1.21).

Turbo (fast) spin echo and echo-planar imaging
These are faster MR techniques that produce multiple slices in shorter
times. There is an image quality penalty to be paid for faster acquisi-
tions and artifacts may manifest differently.

Gradient recalled echo or gradient echo sequences
Gradient echo (GE or GRE) sequences use gradient ¬eld changes
rather than RF pulse sequences. Gradient echo sequences can be T1W
or T2W, although the T2W images are actually T2* (“T2 star”), which is a
less “pure” form of T2 weighting than in spin echo. Artifacts tend to be
more prominent in gradient techniques, particularly those due to local
disturbances of the magnetic ¬eld because of the presence of tissue
interfaces and metal (including iron in blood degradation products).

Fat suppression
Fat-containing tissues have high signal on both T1W and T2W scans.
This can overwhelm the signal from adjacent structures of more
interest, so MR sequences have been developed to reduce the signal
from fat. The STIR sequence described above is one of these. Fat
saturation is another technique that can be used in which a presatura-
tion RF pulse tuned to the resonant frequency of fat protons is applied
Fig. 1.23. A single MIP (maximum intensity projection) view from an MR
to the tissues before the main pulse sequence, causing a nulling of the
angiogram showing the large vessels of the intracerebral circulation. This
signal from the fatty tissues (Fig. 1.22).
angiogram has been created from a time-of-¬‚ight (TOF) scanning sequence.

Diffusion-weighted imaging (DWI)
or use MR contrast agents. In these, ¬‚owing blood in vessels is of high
Changes in the diffusion of tissue water can be visualized using this
signal. A MR angiogram is usually viewed as a maximum intensity
technique, which relies on small random movements of the molecules
projection or MIP (Fig. 1.23). To create an MIP, only the high signal
changing the distribution of phases. This technique is used to image
structures are shown and all the MR slices are compressed together
pathology within the brain, particularly early ischemic strokes.
(or projected) to give a single view as if looking at the subject from
MR angiography a particular angle. Usually, projections from multiple angles are
used. Other methods relying on phase contrast or injected intravas-
MR angiograms often use a “time of ¬‚ight” sequence where the
cular contrast media may also be used.
in¬‚owing blood is saturated with a preliminary RF pulse sequence,

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging


Fig. 1.24. (a) Sagittal T1W and b) coronal T2W images from an MR examination of the spine in a patient who has had surgery with metal screws and rods along the
lower spine. There is marked loss of signal and distortion of the surrounding structures over most of the scan.

Magnetic resonance cholangiopancreaticogram
MRCP or magnetic resonance cholangiopancreaticography images are
used to image the biliary system non-invasively, and are created as a
MIP of a sequence in which ¬‚uid is of high signal.

MR artifacts
Ferromagnetic artifact
All ferromagnetic objects, such as orthopedic implants, surgical clips
and wire, dental ¬llings, and metallic foreign bodies cause major
distortions in the main magnetic ¬eld, giving areas of signal void and
distortion (Fig. 1.24). Even tattoos and mascara can contain enough
ferromagnetic pigments to cause a signi¬cant reduction in image

Susceptibility artifact
This is due to local changes in the ¬eld from to the differing
magnetisation of tissue types, rather like a less pronounced form
of ferromagnetic artifact. Susceptibility artifacts usually occur at inter-
faces between other tissue types and bone or air-¬lled structures.

Motion artifact
The acquisition time for MR is relatively lengthy and image degrada-
tion due to movement artifacts is common. General movement,
including breathing, causes blurring of the image. Pulsation from
blood vessels causes ghosts of the moving structures (Fig. 1.25)
Fig. 1.25. Axial T2W image of the brain in a patient unable to lie suf¬ciently still.

Chemical shift artifact
Aliasing (wraparound) artifact
This occurs at interfaces between fat and water. Protons in fat have a
This can occur when part of the anatomy outside the ¬eld of view of
slightly different resonance frequency compared with those in water,
the scan is incorrectly placed within the image, on the opposite side.
which can lead to a misregistration of their location. This gives a high
This occurs in the phase encoding direction and can be removed by
signal“low signal line on either side of the interface.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

increasing the ¬eld of view (although at the expense of either resolu- thallium, and strontium are all in regular use. Radiopharmaceuticals
tion or time). It is common in echo planar imaging. are normally administered by injection into the venous system but are
also administered orally, directly into body cavities, and by injection
MRI safety into soft tissues.
MR is contraindicated in patients with electrically, magnetically, or
The gamma camera
mechanically activated implants including cardiac pacemakers,
cochlear implants, neurostimulators and insulin, and other implantable Standard nuclear medicine images are acquired using a gamma
drug infusion pumps. Ferromagnetic implants such as cerebral camera (Fig. 1.26). The basic detector in the gamma camera consists of
aneurysm clips and surgical staples, and bullets, shrapnel, and metal a sodium iodide crystal that emits light photons when struck by a
fragments can move. Patients with a history of metallic foreign bodies -ray, with photo-multiplier tubes to detect the light photons emitted.
in the eye should be screened with radiographs of the orbits. The photo-multiplier tube produces an electrical voltage that is con-
A number of implants have been shown to be safe for MR including verted by the electronic and computer circuitry to a “dot” on the ¬nal
non-ferrous surgical clips and orthopedic devices made from non- image. The build-up of dots gives the ¬nal image (Fig. 1.27). Between
ferrous metals. Contemporary devices are largely MRI compatible, the patient and the detector is a collimator which consists of a large
although older ones may not be. lead block with holes in it that select only photons travelling at right
MR magnetic ¬elds can induce electrical currents in conductors, angles to the detector. Those passing at an angle do not contribute to
such as in cables for monitoring equipment attached to the patient the image.
(e.g., ECG leads), with a risk of electric shock to the patient. Any
Single photon emission computed tomography (SPECT)
monitor leads must be carefully designed and tested for MR compati-
bility to avoid this possibility. Computed tomography (CT, described above) allows the reconstruc-
There is no evidence that MR harms the developing fetus. Pregnant tion of a three dimensional image from multiple projections of an
patients can be scanned, although as a precaution MR is not usually external X-ray beam. A similar effect can be obtained in nuclear medi-
performed in the ¬rst 3 months of pregnancy. cine with reconstruction of emissions of radionuclide within the
patient from different projections. This is usually achieved by rotating
Advantages of MR the gamma camera head around the patient.
MR allows outstanding soft tissue contrast resolution and allows SPECT has the advantage of improving image contrast by minimiz-
images to be created in any plane. No ionizing radiation is involved. ing the image activity present from overlying structures in a two-
It gives limited detail in structures such as cortical bone and dimensional acquisition and allows improved three-dimensional
calci¬cation, which return negligible signal. MR has long scanning localization of radiopharmaceuticals.
times in relation to other techniques and requires patients to be sta-
Positron emission tomography (PET)
tionary while the scan is performed. Because of long imaging times
and complexity of the equipment, MR is relatively expensive. The PET deals with the detection and imaging of positron emitting
space within the magnet is restricted (a long tunnel) and some radionuclides. A positron is a negative electron, a tiny particle of
patients experience claustrophobia and are unable to tolerate the antimatter. Positrons are emitted from the decay of proton rich
scan. Access to medically unstable patients is hindered and special, radionuclides such as carbon-11, nitrogen-13, oxygen-15 and ¬‚uorine-
MR compatible, monitoring equipment is required. 18. When a positron is emitted, it travels a short distance (a few mm)
before encountering an electron; the electron and positron are

Nuclear medicine
Nuclear medicine involves the imaging of Gamma rays ( -rays), a type
of electromagnetic radiation. The difference between -rays and X-rays
is that -rays are produced from within the nucleus of the atom when
unstable nuclei undergo transition (decay) to a more stable state,
while X-rays are produced by bombarding the atom with electrons.
Nuclear medicine imaging therefore is emission imaging “ the -rays
are produced within the patient and the photons are emitted from the
subject and then detected.

The -ray emitter must ¬rst be administered to the patient “ the sub-
stance given is known as a radiopharmaceutical. These consist of
either radioactive isotopes by themselves, or more commonly
radioisotopes (usually called radionuclides) attached to some other
molecule. Radionuclides can be created in nuclear reactors, in
cyclotrons and from generators. The most commonly used
radionuclide is Technetium 99 m (Tc-99 m), which is produced from a
generator containing Molybdenum-99 that is ¬rst created in a nuclear
reactor as a product of Uranium-235 ¬ssion. Isotopes of iodine,
krypton, phosphorus, gallium, indium, chromium, cobalt, ¬‚uorine, Fig. 1.26. A gamma camera.

thomas h. bryant and adam d. waldman
An introduction to the technology of imaging

Fig. 1.28. Coronal presentation of data from an FDG PET scan in a patient with
lymphoma. A previously unrecognized site of disease within a right common
iliac lymph node takes up the FDG and appears a an area of high uptake (black).
Other normal, physiological sites of uptake include heart muscle, the liver and
spleen, and the bones. Excretion is via the renal system, so the bladder also
appears of high activity. (FDG ¬‚uoro-deoxy-glucose; the glucose labelled with

Manufacturers have now combined PET and CT in a single scanner in
which the PET image is coregistered with CT. This improves the
anatomical accuracy of PET and is valuable in localizing disseminated
disease, notably cancer.
PET CT is particularly helpful in recurrent cancers of the head
Fig. 1.27. A bone scan. Tc-99 m MDP, which is taken up by osteoblasts within
and neck where post surgical change and scarring can mask new
bone, has been intravenously injected and an image acquired 3 hours later
using a gamma camera. Uptake of the radionuclide can be seen within
the bones, and also within the kidneys (faintly) and bladder “ this radiophar-
maceutical is excreted by the renal system. Advantages of nuclear medicine
Isotope scans provide excellent physiological and functional infor-
annihilated, releasing energy as two 511 keV -rays, which are emitted mation. They can often indicate the site of disease before there has
in opposite directions. The detectors in the PET scanner are set up in been suf¬cient disruption of anatomy for it to be visible on other
pairs and wait for a “coincidence” detection of two 511 keV -rays. imaging techniques. Scans can be repeated over time to show the
A line drawn between the two detectors is then used in the computed movement or uptake of radionuclide tracers. However, nuclear
tomography reconstruction (as in CT). medicine studies sacri¬ce the high resolution of other imaging
Most PET isotopes are made in cyclotrons and have very short half- techniques. Isotope studies involve ionizing radiation, and for
lives (usually only a few minutes to hours). A commonly used PET some longer half-life radioisotopes, patients can continue to emit
chemical is FDG or ¬‚uoro-deoxy-glucose “ glucose labelled with low levels of ionizing radiation for several days. Some isotopes, par-
¬‚uorine-18. Tissues that are actively metabolizing glucose take this up. ticularly those used in PET scanning, are relatively expensive, and
PET has been particularly successful in imaging brain, heart, and some isotopes for PET scanning are so short lived that an on-site
oncological metabolism. PET scans generally have a higher resolution cyclotron is required.
than SPECT scans (Fig. 1.28).

Section 1 The basics

Chapter 2 How to interpret an image

A DA M W. M . M I T C H E L L

In order to attempt to interpret a radiographic image, it is essential
that you ¬rst identify the type of examination and understand some-
thing of the principles behind it. Before examining any image, the
name of the patient and the date of the study should be checked. The
¬lm should also be hung correctly and right and left sides ascertained.

Plain radiography
Plain radiographs are the most commonly encountered of all imaging
studies. The following chapters explain the radiological anatomy
involved, but it is equally important to appreciate how the ¬lm was
Staff in the radiology department can offer advice on any additional
projections but it is very important from the outset to provide as
much information as possible in the request for an examination, so
that the correct views and exposures are used.
In general, over-exposed (dark), radiographs are more useful than
those that are under-exposed, since the former retain the information.
Rather than request another ¬lm and expose the patient to more ion-
izing radiation, the dark ¬lm should be examined with a bright light
in the ¬rst instance. Trachea
Digital radiographs can be interrogated by “windowing” (see below), Apical artery right
and although the original exposure must be correct, the resulting Apical vein right
Chest wall (rib cage,
image can be manipulated to highlight bone or soft tissue detail as Superior vena cava pleural line)

required. Azygos knob (6mm) Aortic arch

Main pulmonary
Ascending aorta
Right main bronchus
The chest radiograph Left main bronchus
Right pulmonary artery
Left pulmonary artery
The frontal chest radiograph is the most commonly requested plain Right pulmonary veins
¬lm. The image is taken either as a “PA” (posteroanterior) or as an Left pulmonary vein
Right interlobar artery
“AP” (anteroposterior), depending on the direction of the X-ray beam. Left auricular
Right intermediate appendage
The projection is usually marked on the ¬lm. bronchus
Region of contact
A PA projection is the better quality ¬lm and allows the size and Right middle lobe
of oesophagus and
arteries and bronchi
shape of the heart and mediastinum to be assessed accurately. A PA left atrium
Right atrium
¬lm is taken with the patient erect and is performed in the radiology Apex of left ventricle

department. This, of course, requires the patient to be reasonably Right hemidiaphragm Left hemidiaphragm
mobile (¬g. 2.1).
For the less mobile or bed-bound patient, portable ¬lms are taken.
These are all AP and can be taken with the patient supine or erect. Fig. 2.1. Normal PA chest radiograph.

Applied Radiological Anatomy for Medical Students. Paul Butler, Adam Mitchell, and Harold Ellis (eds.) Published by Cambridge University Press. © P. Butler,
A. Mitchell, and H. Ellis 2007.
How to interpret an image adam w. m. mitchell

Gas in


fat stripe


Gas in

Gas in
Fig. 2.2. AP chest radiograph. There has been a poor respiratory effort and there and
is a false impression of cardiac enlargement.

Fig. 2.3. Plain abdominal radiograph.

Because the divergent X-ray beam causes magni¬cation, AP ¬lms can
give a false impression of cardiac enlargement and mediastinal
widening (¬g. 2.2).
Plain ¬lms of the musculoskeletal system
Once the patient™s identity has been checked and the ¬lm hung
Interpretation of these images is often more straightforward and it
properly, it is important to check for any rotation. This can change
is usual, in trauma, to take two views, at right angles to each other.
the shape of the heart and the appearance of the lungs, creating
Fractures may be missed on a single view (¬g. 2.4).
a spurious difference in radiolucency between the two sides. In a
It is also the case that the soft tissue patterns on a plain ¬lm can
properly centered ¬lm, the medial ends of the clavicles should be
provide clues to the diagnosis.
a similar distance from the spinous processes of the thoracic
Remember to look at the periphery of any ¬lm as well as its centre.
Contrast studies of the gastrointestinal tract
In the case of the chest ¬lm, the cervical soft tissues and the upper
High density contrast medium is often used in the investigation of
abdomen should be examined.
the gastrointestinal (GI) tract. Clinical staff (and medical students)
If the ¬lm appears rather dark, the bones will be well demon-
will often be confronted with these studies in clinico-radiological
strated, but it will be worth using a bright light to examine the lungs,
meetings, in the outpatients™ clinic and perhaps under examination
to avoid missing a small pneumothorax.
Barium is the commonest contrast medium used and is generally
The abdominal radiograph
very safe. It is contraindicated in suspected rupture of the GI tract
because the presence of barium in the mediastinum or the
The plain abdominal ¬lm is also a commonly requested investiga-
peritoneum has a very high morbidity rate. In these situations
tion. Its particular importance in everyday practice is in the
a water-soluble contrast medium, such as gastrogra¬n,

. 1
( 10)