Clinical Cardiac MRI - 2nd Edition

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

https://medidownload.com/product/clinical-cardiac-mri-2nd-edition/

Click Download Now

Jan Bogaert • Steven Dymarkowski
Andrew M. Taylor • Vivek Muthurangu
Editors

Clinical Cardiac MRI

Foreword by
Maximilian Reiser

123
Prof. Dr. Jan Bogaert Prof. Andrew M. Taylor
Department of Radiology Cardio-respiratory Unit
Katholieke Universiteit Leuven Hospital for Children
University Hospital Leuven Great Ormond Street
Herestraat 49 London WC1N 3JH
3000 Leuven UK
Belgium
Dr. Vivek Muthurangu
Prof. Dr. Steven Dymarkowski Cardio-respiratory Unit
Department of Radiology Hospital for Children
Katholieke Universiteit Leuven Great Ormond Street
University Hospital Leuven London WC1N 3JH
Herestraat 49 UK
3000 Leuven
Belgium

Additional material to this book can be downloaded from http://extras.springer.com/

ISSN 0942-5373
ISBN 978-3-642-23034-9 e-ISBN 978-3-642-23035-6
DOI 10.1007/978-3-642-23035-6
Springer Heidelberg New York Dordrecht London

Library of Congress Control Number: 2012930015

Ó Springer-Verlag Berlin Heidelberg 2012

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9,
1965, in its current version, and permission for use must always be obtained from Springer. Violations are
liable to prosecution under the German Copyright Law.
The use of general descriptive names, registered names, trademarks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
Product liability: The publishers cannot guarantee the accuracy of any information about dosage and
application contained in this book. In every individual case the user must check such information by
consulting the relevant literature.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Foreword

For this second edition of the highly successful reference book on Clinical
Cardiac MRI the editorial team has been enlarged and several chapters have
been added or rewritten in order to take the developments of the last 7 years
into account. MRI has only recently been established as diagnostic as well as
prognostic method in cardiovascular imaging and is now also used for car-
diovascular intervention.
Cardiovascular diseases are the leading cause of death, counting for about
30% percent of global deaths. The value of an up to date, thoroughly resear-
ched and comprehensive textbook on cardiac imaging written by leading
international experts in the field can therefore not be overestimated.
Clinical Cardiac MRI includes chapters on physics, anatomy, cardiac func-
tions as well as MRI imaging techniques, contrast agents, guidelines for
imaging interpretation and—where applicable-interventions for all common
cardiac pathologies. Additionally 100 life cases can be found in the online
material for the book. These also include less frequent cardiac diseases.
I would like to sincerely thank the editors as well as the authors of this
textbook for their time and expertise and am very confident that this edition
will, as its predecessor, be a very useful tool for everyone involved in cardiac
MRI imaging.

Maximilian Reiser

v
Preface

By the time a book preface is written, usually most of the work has been
accomplished, chapter proofs have been forwarded for correction to the
authors, while the book index is still waiting to be finished. It is also the
moment the editors get a first glimpse whether the book will match their
expectations. About 7 years after the first edition, and almost two years after we
agreed with Springer to edit a second edition of our textbook on ‘Clinical
Cardiac MRI’, we are pleased to present you with a new, completely updated
textbook. The decision to write a second version was largely driven by the huge
success of the first edition, with almost exclusively positive comments not only
by reviewers but by the many readers of our book throughout the world,
readers that appreciated our book for being a highly useful guide for daily use,
for the high-quality of the images and the addition of a CD ROM with 50 real-
life cases. Their enthusiasm has been the strongest drive to edit a new version,
while their comments have been most helpful to prepare an improved second
edition.
For the new edition, we welcome Dr. Vivek Muthurangu, from Great
Ormond Street Hospital for Children, London as the fourth member of the
editorial board. Dr. Muthurangu has great expertise in the field of cardiac MR
physics, pulmonary hypertension and cardiac modeling.
At the end of 2004, when the first edition of ‘Clinical Cardiac MRI’ was
released, cardiac MRI had been through five truly exciting years that had
caused a paradigm shift in cardiovascular imaging. Balanced steady-state free
precession bright imaging had rapidly become the reference technique to assess
cardiac function, and moreover yielded promise for other applications such as
coronary artery imaging. Non-invasive comprehensive cardiac tissue charac-
terization was no longer a far off dream. For instance, T2-weighted imaging
offered the possibility of in-vivo imaging of reversible myocardial injury, while
the nature of the underlying disease could often be deduced by the pattern of
myocardial enhancement using (inversion-recovery) contrast-enhanced imag-
ing, thus obviating the need for other, more invasive procedures. Besides its
diagnostic role, cardiac MRI was beginning to show promise as a prognostic
tool that could provide predictive information about future cardiac events.
Ever since MRI was proposed to have a role in the assessment of cardio-
vascular disease, cardiac MRI has experienced some resistance from the
broader cardiology community with regard to its clinical value and the daily use
of this ‘exotic’ technique. Fortunately, things have moved in the right direction.
Cardiac MRI has now become the technique of choice when it comes to the

vii
viii Preface

depiction of therapeutic effects (e.g. regenerative cell therapy), and for an

increasing number of clinical indications a cardiac MRI study is becoming a
crucial investigation that guides patients care. This is due in great extent to an
increased visibility and awareness of cardiac MRI at congress meetings and in
scientific journals, and the integration of this technique into appropriateness
criteria and guidelines. Also the availability of dedicated textbooks has helped
toward a broader recognition of cardiac MRI.
For this edition, a new chapter on cardiac modeling has been added; the
chapter on heart failure, pulmonary hypertension and heart transplantation has
been split in two separate chapters, yielding a total of twenty chapters. Some of
the chapters have been extensively rewritten and also extended, aiming to
appropriately highlight the rapidly evolving role of cardiac MRI. In particular,
this was the case for ischemic heart disease and heart muscle diseases. For other
chapters, such as the chapter on congenital heart disease, the emphasis is now
on daily clinical applications to investigate simple and more complex cardiac
malformations. Throughout the textbook, practical schemes are provided
indicating how to apply cardiac MRI for a wide variety of cardiac diseases. And
last, but by no mean least, a series on 100 new clinical cases is available as
online material. These cases cover a wide spectrum of cardiac diseases,
including some less frequent cardiac abnormalities, which have been selected to
underscore the added value of cardiac MRI. The online material has the
advantage of bringing the dynamic features of cardiac MRI (e.g., functional or
stress imaging).
We sincerely hope that readers will receive this edition with the same
enthusiasm as our first effort.

Jan Bogaert
Steven Dymarkowski
Andrew M. Taylor
Vivek Muthurangu
Contents

Cardiac MRI Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Vivek Muthurangu and Steven Dymarkowski

MR Contrast Agents for Cardiac Imaging . . . . . . . . . . . . . . . . . . . . . . 31

Yicheng Ni

Practical Set-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
S. Dymarkowski

Cardiac Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
J. Bogaert and A. M. Taylor

Cardiovascular MR Imaging Planes and Segmentation . . . . . . . . . . . . . 93

A. M. Taylor and J. Bogaert

Cardiac Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

J. Bogaert

Myocardial Perfusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

J. Bogaert and K. Goetschalckx

Ischemic Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

J. Bogaert and S. Dymarkowski

Heart Muscle Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

J. Bogaert and A. M. Taylor

Pulmonary Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355

Shahin Moledina and Vivek Muthurangu

Heart Failure and Heart Transplantation. . . . . . . . . . . . . . . . . . . . . . . 367

S. Dymarkowski and J. Bogaert

Pericardial Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383

J. Bogaert and A. M. Taylor

ix
x Contents

Cardiac Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411

J. Bogaert and S. Dymarkowski

Valvular Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465

Andrew M. Taylor, Steven Dymarkowski, and Jan Bogaert

Coronary Artery Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511

S. Dymarkowski, J. Bogaert, and A. M. Taylor

Congenital Heart Disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553

Marina L. Hughes, Vivek Muthurangu, and Andrew M. Taylor

Imaging of Great Vessels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611

Oliver R. Tann, Jan Bogaert, Andrew M. Taylor, and Vivek Muthurangu

MR Guided Cardiac Catheterization . . . . . . . . . . . . . . . . . . . . . . . . . . 657

Vivek Muthurangu and Andrew M. Taylor

Cardiovascular Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669

Giovanni Biglino, Silvia Schievano, Vivek Muthurangu,
and Andrew Taylor

General Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695

J. Bogaert, S. Dymarkowski, A. M. Taylor, and V. Muthurangu

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Contributors

G. Biglino Centre for Cardiovascular Imaging, UCL Institute of Cardiovas-

cular Science and Great Ormond Street Hospital for Children, Great Ormond
Street, WC1N 3JH, London, UK
J. Bogaert Department of Radiology and Medical Imaging Research Center
(MIRC), University Hospitals Leuven, Catholic University Leuven, Herestraat
49, 3000, Leuven, Belgium, e-mail: jan.bogaert@uzleuven.be
Steven Dymarkowski Department of Radiology and Medical Imaging Research
Center (MIRC), University Hospitals Leuven, Catholic University Leuven,
Herestraat 49, 3000, Leuven, Belgium, e-mail: steven.dymarkowski@
uzleuven.be
K. Goetschalckx Department of Cardiovascular Diseases, University Hospi-
tals Leuven, Catholic University Leuven, Herestraat 49, 3000, Leuven,
Belgium, e-mail: kaatje.goetschalckx@uzleuven.be
Marina L. Hughes Centre for Cardiovascular Imaging, UCL Institute of Car-
diovascular Science and Great Ormond Street Hospital for Children, Great
Ormond Street, WC1N 3JH, London, UK
Shahin Moledina, UCL Centre for Cardiovascular Imaging and Great Ormond
Street Hospital for Children, London, WC1N 3JH, UK
Vivek Muthurangu Cardio-respiratory Unit, Hospital for Children, Great
Ormond Street, London, WC1N 3JH, UK; Centre for Cardiovascular Imaging,
UCL Institute of Cardiovascular Science and Great Ormond Street Hospital
for Children, Great Ormond Street, WC1N 3JH, London, UK
Yicheng Ni Department of Radiology, University Hospitals Leuven, Catholic
University Leuven, Herestraat 49, 3000, Leuven, Belgium, e-mail: yicheng.ni@
med.kuleuven.be
Silvia Schievano Centre for Cardiovascular Imaging, UCL Institute of
Cardiovascular Science and Great Ormond Street Hospital for Children, Great
Ormond Street, WC1N 3JH, London, UK
Oliver R. Tann Consultant in Cardiovascular Imaging, Cardio-Respiratory
Unit, Great Ormond Street Hospital for Children, London, WC1N 3JH, UK

xi
xii Contributors

Andrew M. Taylor Centre for Cardiovascular Imaging, UCL Institute of

Cardiovascular Science and Great Ormond Street Hospital for Children,
London, UK, e-mail: a.taylor76@ucl.ac.uk
 Cardiac MRI Physics
Vivek Muthurangu and Steven Dymarkowski

Contents Abstract
This chapter addresses the use of MRI and to a lesser
1 Basic Physics ............................................................ 1 extent CT in the diagnosis and management of
1.1 Spin ............................................................................ 1 pulmonary hypertension. The basics of pulmonary
1.2 Resonance .................................................................. 2
1.3 The MR Signal .......................................................... 2
hypertension will be addressed, including epidemi-
1.4 Relaxation .................................................................. 3 ology and treatment strategies. Then different MRI
techniques will be discussed in the context of their
2 Magnetization Preparation Pulses......................... 4
2.1 Inversion Recovery.................................................... 4 relevance to pulmonary hypertension. Finally the
2.2 Saturation Recovery .................................................. 7 role of CT in pulmonary hypertension will be
2.3 T2 Preparation ........................................................... 8 discussed. By the end of the chapter the reader
3 Spatial Encoding and Image Construction........... 8 should have a better understanding of how to use
3.1 k-Space....................................................................... 9 cross-sectional imaging in pulmonary hypertension.
3.2 k-Space Filling Strategies.......................................... 12
3.3 Parallel Imaging......................................................... 15
4 Motion Compensation ............................................. 16
4.1 Cardiac Gating........................................................... 16 1 Basic Physics
4.2 Multi-Phase Acquisitions .......................................... 17
4.3 Respiratory Gating..................................................... 18
4.4 Single Shot and Real-Time Acquisitions ................. 20 The basic principles of magnetic resonance imaging
(MRI) are the same irrespective of the part of the body
5 Cardiac MRI Sequences ......................................... 20
5.1 Spin Echo Sequences ................................................ 20 that is being imaged. However, there are specific areas
5.2 Spoiled Gradient Echo Sequences ............................ 22 of MRI physics that are particularly important for
5.3 Balanced Steady-State Free Precession .................... 25 cardiac MRI specialists to understand. Thus, in
6 Conclusion ................................................................ 28 this chapter we will review both basic MRI physics
(i.e. generation of the MR signal and spatial encoding),
7 Key Points................................................................. 29
as well as more cardiac-specific topics (i.e. motion
References.......................................................................... 29 compensation and cardiac relevant MRI sequences).
The purpose of this chapter is to enable the reader to
V. Muthurangu (&) better understand and optimize their MR imaging.
Cardio-Respiratory Unit, Great Ormond Street,
Hospital for Children, Great Ormond Street, 1.1 Spin
London, WC1N 3JH, UK
e-mail: v.muthurangu@ucl.ac.uk
Nuclei with unpaired protons or neutrons (i.e. an odd
S. Dymarkowski
proton or neutron numbers) possess a property called
Department of Radiology, University Hospital Leuven,
Katholieke Universiteit Leuven, Herestraat 49, quantum spin, which makes them ‘MR active’. The
3000 Leuven, Belgium most common of these ‘MR active’ nuclei is 1H, but

J. Bogaert et al. (eds.), Clinical Cardiac MRI, Medical Radiology. Diagnostic Imaging, 1
DOI: 10.1007/174_2011_412,  Springer-Verlag Berlin Heidelberg 2012
2 V. Muthurangu and S. Dymarkowski

given by the Larmor equation: x = c B0, where c is the

gyromagentic constant, a nuclei specific constant.
Hydrogen exposed to a 1.5T field precess around the B0
axis at approximately 64 MHz. However, as they are
out of phase with each other, the NMV does not precess
and only has a component in the direction of the B0
field. It is in this state that radiofrequency (RF) energy
can be inputted into the system causing the NMV to
move toward a plane perpendicular to the B0 field.

Fig. 1 a Proton spinning around its own axis while precessing 1.2 Resonance
around the z-axis (i.e. the direction of the static field). b RF
excitation causing flipping of z magnetization into the x–y plane
RF energy is transmitted as an electromagnetic wave
and its magnetic component (the B1 field) can interact
other nuclei are used in MRI (e.g. 19F, 13C and 23Na).
with the magnetic moments of spinning protons. If the
In the rest of this chapter only the 1H nucleus
B0 field is assumed to be in the z direction (along the
(essentially a single proton) will be considered. In
bore of the MR scanner), then a perpendicular RF pulse
Newtonian terms, nuclei with spin can be thought of
is in the x–y plane. Unlike the B0 field, the B1 field
as spheres spinning on their own axis (much like the
oscillates and it is this fact that forms the basis of res-
earth spinning around the polar axis). As these nuclei
onance. Resonance only occurs if the frequency of the
have a net positive charge (due to their proton com-
RF pulse equals the precessional frequency of the
ponent) they generate a magnetic field as they spin,
hydrogen nucleus at the given field strength. On
giving rise to their popular analogy as bar magnets. At
transmission of a resonant RF pulse, protons, which
rest, the protons are randomly arranged in the body.
were previously precessing around the z-axis will line
However, in the presence of an external magnetic
up and start precessing around the axis of the B1 field.
field (B0) protons will become aligned. In quantum
This leads to two important changes in the NMV (M0).
terms, nuclei align either parallel or antiparallel to the
Firstly, because the protons have aligned with the B1
B0 field due to the fact that protons can occupy
field they precess around the z-axis in phase. This is
multiple energy states. Low-energy protons line up
important, as now M0 possesses coherent x-y magne-
parallel to B0 while high-energy protons line up anti-
tization. Secondly, the precession of protons around
parallel. At room temperature there is always a small
both the z and B1 axis causes the M0 to nutate or spiral
excess of parallel protons and thus the net magnetic
into the x–y plane. The spiral motion during nutation is
vector (NMV) is in the direction of the B0 field. The
difficult to visualize and therefore resonance is usually
exact excess of parallel protons, and thus the magni-
described in the rotating frame of reference (i.e. the
tude of the NMV, is governed by the Boltzmann
observer is rotating around the z-axis at the same
distribution. This states that as field strength increa-
frequency as the protons). In the rotating frame of ref-
ses, and temperature decreases, the magnitude of
erence, nutation becomes a simple flip into the x–y plane
NMV increases. This explains the greater signal at
(Fig. 1b). The flip angle is dependent on the strength and
higher field strengths. Although MR is a quantum
duration of RF pulse, with a 90o flip placing all the
phenomenon from this point forward it is easier to
longitudinal magnetization into the transverse plane.
think of the magnetic moments in purely Newtonian
The flipped magnetization vector now has a transverse
terms. This is because it simplifies the explanation of
component, which forms the basis of the MR signal.
precession, resonance and spatial encoding.
In the presence of a B0 field the protons do not simply
line up, they actually precess or ‘wobble’ around the B0 1.3 The MR Signal
axis (Fig. 1a). This is analogous to the motion of a
spinning top, which spins around its own axis, while Faraday’s law of electromagnetic induction states
also precessing around its surface point of contact. The voltage will be induced in a conductor exposed to a
precessional frequency (x) of a MR active nucleus is changing magnetic field. Longitudinal magnetization
Cardiac MRI Physics 3

the surrounding lattice (spin-lattice relaxation). This

causes the NMV to flip back into the z direction; during
this process longitudinal magnetization recovers
exponentially (Fig. 2). The rate of longitudinal recov-
ery is dependant on the rate constant T1. As T1 depends
on the atomic structure of the tissue, it is a tissue-spe-
cific constant. In tissues with a short T1 (such as fat)
longitudinal magnetization will be recovered more
quickly than in tissue with a longer T1 (such as muscle).
This is important in the generation of T1-weighted
contrast, which will be discussed later in this chapter.
The nature of the exponential recovery curve means
that when time equals T1, 63% of z magnetization will
Fig. 2 T1 relaxation curve—note that at time = T1 the
z magnetization has relaxed back to 0.63 times its original value have recovered. Recently T1 mapping has become a
great interest in cardiac MRI. In T1 mapping, multiple
images are acquired at different times after an excita-
does not change and therefore it cannot induce a tion pulse (or more usually after an inversion pulse
voltage. Transverse magnetization on the other hand which will be discussed in more detail later in this
rotates in the x–y plane and therefore it will induce a chapter). This allows reconstruction of the T1 recovery
voltage in a conductor. This is an important point to curve and calculation of the tissue T1. The reason that
note: only the transverse component of M0 induces T1 mapping has become of great interest is that there is
voltage. As the transverse magnetization rotates at the evidence to suggest that after contrast administration
Larmor frequency, the induced voltage will also the tissue T1 correlates with the amount of myocardial
oscillate at the same frequency. However it is not in fibrosis. This will be addressed in more detail in
this form that the data is ultimately used. The sinu- ‘‘Heart Muscle Diseases’’.
soidally varying voltage undergoes a process called The other relaxation process is transverse relaxa-
complex demodulation, which essentially converts the tion and is due to dephasing of the individual spins
data into the rotating frame of reference. Thus, the leading to a reduction in coherent transverse magne-
resultant MR signal has a magnitude (the amplitude of tization. This is due to the interaction between the
the varying voltage) and a phase, which after RF magnetic fields of adjacent protons (spin–spin inter-
excitation is zero. It can easily be represented as a actions) and results in different protons precessing at
hand on a clock face, whose size is equal to the different rates. In the rotating frame of reference, this
magnitude and whose position is equal to the phase. It variation in frequency is seen as dephasing. Thus, the
is within this signal that spatial information must be coherent magnetization vector in the x–y plane starts
encoded. However this signal does not stay the same to fan out resulting in a reduction in the net transverse
indefinitely, but rather relaxes back to its resting state. magnetization. Transverse relaxation results in expo-
It is this relaxation that forms the basis of MRI nential decay of coherent transverse magnetization at
contrast. a rate governed by T2 (Fig. 3). Thus, when time
equals T2, transverse magnetization will have
decayed to 37% of its original value. Much like T1,
1.4 Relaxation T2 also depends on the atomic structure of the tissue,
and is therefore an independent tissue-specific con-
Relaxation is the process by which magnetization stant. In tissues with a long T2 (such as tissue with a
returns to its resting state after RF excitation. There are high water content) transverse magnetization will
two processes involved, both of which are dependent on persist longer than tissue in tissue with a shorter T2
the atomic arrangement within tissues. Thus, the rate of (such as fat). This is important in the generation of
relaxation is tissue specific and can be used to develop T2-weighted contrast, which will be discussed later in
tissue contrast. Longitudinal relaxation (or recovery) is this chapter. However, there is a second process that
due to transfer of energy from high-energy protons to results in loss of transverse magnetization. This is B0
4 V. Muthurangu and S. Dymarkowski

trying to quantify myocardial edema, while T2* is

useful when assessing iron overload (iron causes local
field inhomogeneity). Mapping T2 or T2* is done
by acquiring multiple images at different times after
the excitation pulse. This allows reconstruction of the
T2/T2* decay curve.
With prior knowledge of tissue T1 and T2, timing
parameters (i.e. TR and TE) can be altered to provide
specific tissue contrasts. Other ways to change con-
trast are to add exogenous contrast agents or to pre-
pare magnetization prior to imaging. The next section
will discuss in detail the use of magnetization prep-
aration to change MR contrast.
Fig. 3 T2 and T2* relaxation curves—note that the transverse
magnetization has fallen to 0.37 times its original value at
time = T2/T2*
2 Magnetization Preparation Pulses

Magnetization preparation is the process by which the

magnetic vector is manipulated prior to imaging in
order to produce specific tissue contrast. This tech-
nique is used heavily in cardiac MRI and the most
common techniques are described below.

2.1 Inversion Recovery

The most commonly used form of magnetization

preparation is inversion recovery (IR). IR depends on
the fact that different tissues have different T1 char-
acteristics. In IR sequences, an 180o RF pulse (or
Fig. 4 Inversion recovery curve—note that z-axis magnetiza- inversion pulse) is used to flip the magnetization into
tion passes through 0 at time = 0.693 times the T1 of the tissue the opposite direction along the z-axis. From this
position the magnetization relaxes back to its original
field inhomogeneity, which also results in dephasing. state following the T1 curve of the tissue (Fig. 4). At
This accelerated dephasing is encapsulated in the time a time of approximately T1 * Ln2 (0.693) the longi-
constant T2*. The T2* value is dependant on the tudinal magnetization will pass through zero (i.e. the
underlying T2 and any field inhomogeneity and is magnetization will be completely in the x–y plane).
therefore not purely a tissue constant. One way to As different tissues have different T1 characteristics,
improve field homogeneity is to shim. Shimming is a each tissue will pass through zero (or the null point) at
process by which either metal is used to distort the different times. During RF excitation (which is
magnetic field (passive shimming) or shim coils are applied some time after the IR pulse) only tissues with
used to generate a corrective magnetic field (active non-zero longitudinal magnetization will produce an
shimming). These techniques can be used together MR signal. Therefore if the time between inversion
and active shimming is vital for some newer cardiac and imaging (TI) is chosen carefully, signal from a
MR sequence. In the same way that one can measure given tissue can be completely abolished. All IR
the T1 of myocardium, one can also measure myo- sequences work on this principle, and that different
cardial T2 or T2*. Quantification of T2 is useful when tissues can be nulled by choosing specific TI’s.
Cardiac MRI Physics 5

Fig. 5 a Short axis view

through the atria with no fat
saturation. b STIR sequence
in the same image plane—
note that the anterior and
pericardial fat are nulled
because of the inversion pulse
(TI = 160 ms)

Fig. 6 a SPIR dark blood

sequence—note the
inhomogeneous nulling of
the fat when using spectrally
selective inversion pulses.
b Non-fat saturated dark
blood image in the same
image plane

2.1.1 Short Tau Inversion Recovery sequences (Kaldoudi et al. 1993). Spectral selective
Fat suppression can be an important requirement in pulses rely on the fact that water and fat precess at
cardiac MRI. A robust method of fat suppression is slightly different frequencies (approximately 220 Hz
STIR (Simonetti et al. 1996), which relies on the short difference at 1.5T). Therefore a special RF pulse can
T1 of fat compared to other tissues. Therefore, the fat be used that only excites fat. In SPIR a spectrally
magnetization will pass through null point of an IR selective 180o pulse is used to invert only the fat
sequence before the tissue of interest. If imaging is magnetization. The water magnetization is unchanged
performed at the null point of fat, the signal from the by the spectrally selective 180o pulse. The fat mag-
fat will be suppressed. As the T1 of fat is around netization is then allowed to recover and a TI is
230 ms, a TI of between 150 and 170 ms can be used chosen that coincides with the null point of fat. Unlike
to robustly suppress fat. Of course the magnetization STIR, at the onset of imaging all of the water mag-
from other tissue (such as muscle) will also be netization is in the longitudinal axis and therefore
recovering and thus the signal produced will be lower there is no loss in SNR (Fig. 6). However, SPIR
than if no inversion had been performed. This is techniques are very susceptible to magnetic field
particularly true for tissue with short T1’s. Never- inhomogeneity and shimming is important. In real-
theless STIR is frequently used in cardiac MRI due to world applications of SPIR an inversion pulse of
its robustness and the fact that it can be combined between 90 o and 180o is used.
with most imaging sequences (Fig. 5).
2.1.3 Contrast-Enhanced Inversion Recovery
2.1.2 Spectral Inversion Recovery Contrast-enhanced inversion recovery is an extremely
The problem with STIR is the loss of signal to noise important technique in cardiac MRI (Kim et al. 2000).
ratio (SNR); this can be overcome by the use of SPIR It relies on the fact that tissue containing gadolinium