Since the initial clinical experience with computed tomography
(CT) and magnetic resonance imaging (MRI), the heart and coronary
tree have loomed in the technologic distance like the Holy Grail,
at once inviting, tempting, and daring. Imaging of the heart has
always posed a technical challenge because of its continuous
motion. Yet, proponents of CT and MRI resolutely accepted the
challenge and have crept toward the goal in small steps
interspersed with a few large bounds. Now, both CT and MR appear to
be on the verge of culminating the quest with the prize in
Unlike other organs, the heart must be imaged during a rapid and
complex cyclical motion that has substantial beat-to-beat
variation, making high temporal resolution essential. Moreover, the
coronary arteries' small diameter, complex anatomy, and tortuosity
make high spatial resolution a prerequisite for any cardiac
CT and MRI have both shown potential for noninvasive cardiac
imaging. CT allows visualization of the coronary arteries,
including calcifications, stenoses, and atherosclerotic plaque. MRI
permits assessment of cardiac morphology and function, perfusion,
and myocardial viability. Recent technologic developments in both
imaging modalities have positioned CT and MRI for expanded
application in cardiac imaging.
Cardiac CT Imaging
For CT, the journey has been a particularly long one.
Cross-sectional CT imaging of the coronary arteries was first
performed >20 years ago, using electron-beam tomography (EBT).
Since EBT does not require movement of an X-ray tube around the
patient, it provides fast exposure times and image acquisition with
prospective electrocardiographic (ECG) triggering. Two-detector
arrays allow simultaneous acquisition of 2 contiguous 1.5-mm
images. However, low spatial resolution and high image noise
detract from EBT image quality.
In the late 1990s, CT systems with simultaneous acquisition by
4-detector rows and a minimum rotation time of 500 msec were
Subsequently, "the development and the introduction of 16-slice CT
represented a major advance over 4-slice scanners," noted Dr.
Richard D. White, Chair of Radiology at the University of Florida
and Shands-Jacksonville. The recent development of the 64-slice CT
scanner resulted in another significant advance.
The introduction of 64-slice CT provided the capability for
"true isotropic resolution, equal resolution in all 3
orientations," according to Dr. White, who believes that 64-slice
CT "finally made CT angiography possible. Even though the jump from
4 to 16 slices made it possible to capture the heart and
coronaries, it took the next incremental change to make that really
The 64-slice scanners offer a combination of spatial resolution
and speed that was previously unavailable for noninvasive cardiac
imaging. With ECG synchronization, 2 gating techniques are
available: Prospective and retrospective ECG gating. Prospective
gating permits selection of a single point in the cardiac cycle for
imaging. The technique reduces radiation exposure because the
radiation beam emits only a short burst. Retrospective ECG gating
requires slow table motion and simultaneous recording of the ECG
tracing. This technique allows for greater flexibility in image
reconstruction and offers the ability to reconstruct images at any
phase of the cardiac cycle.
Increased spatial and temporal resolution will hone the accuracy
of CT for evaluation of the coronary arteries. A goal of detecting
10% to 20% stenoses should be within reach. The combination of
ECG-synchronized scanning, new reconstruction techniques, fast
volume coverage, and high spatial and temporal resolution positions
CT to assume a major role in the assessment of coronary heart
disease and cardiac function. By offering a "one-stop shop" for
anatomic and physiologic data, cardiac CT has the potential to
replace several cardiac examinations, making it a cost-effective
With 64-slice CT, coronary CT angiography has become very robust
(Figure 1). Dr. Ella Kazerooni, Director of Thoracic Radiology at
the University of Michigan Medical Center in Ann Arbor, MI noted
that the robustness "makes 64-slice CT an attractive alternative to
stress testing, nuclear testing, and stress echocardiography as a
means of evaluating patients with suspected cardiac disease."
However, little data have been published in support of that cardiac
application of CT.
"If a patient has a stress test, stress echocardiography, or
stress nuclear test, we know from years of accumulated data what
the probability is that a patient will have a coronary event in 3
or 5 years," said Dr. Kazerooni. "We know the probability that a
patient will have a critical lesion at catheterization. While
64-slice CT is clearly an accurate test for looking at the coronary
arteries when compared with cardiac catheterization, there is no
published literature on how it does in comparison with other
diagnostic tests in the evaluation of patients."
The next-generation CT scanner, a 256-slice machine, is already
on the near horizon. The scanner has been introduced in Japan,
where it is undergoing initial clinical evaluation.
At the American College of Cardiology meeting earlier this year,
Japanese investigators reported findings from the evaluation of
patients with prior myocardial infarction. The 256-slice scanner
covers 12.8 cm along the z-axis, scanning the entire heart in a
single rotation, which the investigators described as "one-beat
whole-heart imaging." "The ability to scan the entire heart in a
single heartbeat reduces radiation exposure," said Dr. Akira
Kurata, a Cardiologist at Ehime University School of Medicine in
The coronary arteries could be assessed without artifacts.
Abnormal left ventricular functional parameters were identified and
evaluated, and images of the coronary tree and left ventricular
function were fused successfully. The results indicate that the 2nd
Spec 256-slice scanner promises next-generation CT for cardiac
Dr. Kurata pointed out that the machine can complete a whole-heart
scan in a single heart beat, even in patients who have rhythm
disturbances. He also speculated that the 256-slice scanner might
have a role in volumetric cine imaging of the circulatory
Cardiac CT applications
Enhancements to CT technology have translated into new imaging
capabilities and improvement of existing clinical applications. The
advent of 64-slice CT, for example, spares some patients from a
trip to the cardiac catheterization laboratory. "A patient with
atypical chest pain might undergo a stress test that produces
results that are discordant with the physicians' suspicions about
the patient's condition," said Dr. Kazerooni. Cardiac CT offers a
"tie-breaker" that also saves the patient from an invasive
procedure (Figure 2). When used in that tiebreaker capacity,
cardiac CT complements catheterization but does not replace it.
"If you have a 65-year-old man with any kind of twinge, CT
probably should not be considered as first line because there is a
high likelihood of stenotic disease," said Dr. White. "But, if you
have a 65-year-old woman with a lower likelihood of stenotic
disease, cardiac CT would be beneficial if it saved her a trip to
the catheterization lab."
"By looking at a patient's age, gender, and symptoms, you can
begin to predict who is more likely to have significant narrowing,
and that's best answered with catheterization, because you might
intervene with catheterization at the same time. If you don't
suspect significant narrowing, then you should probably go with CT
to help decide the appropriate level of medical therapy."
A somewhat related issue that has attracted a lot of attention
involves the use of cardiac CT for the evaluation of patients who
come to the hospital emergency department chest pain centers. CT
provides a means of evaluating the patients quickly, thus speeding
the time to discharge or admission, depending on the findings.
"Between 85% and 90% of patients who come to the emergency
department for chest pain do not have cardiac disease," said Dr.
Kazerooni. "Patients might have to stay at the hospital for up to
24 hours just to rule out heart disease. With CT, they can undergo
cardiac imaging and be discharged 3 hours later if the images have
shown the patient doesn't have significant cardiac disease. CT
offers a way to reduce the length of stay and the cost of
Although CT might not replace cardiac catheterization, it very
well could provide competition for other noninvasive imaging
modalities, such as transesophageal echocardiography and MR.
"In my opinion, transesophageal echocardiography really isn't
noninvasive," said Dr. Michael Poon, Director of Cardiovascular
Medicine and Integrated Imaging at Cabrini Medical Center, New
York, NY. "MR is much more time consuming and costs more. If I can
do a test in 5 minutes, I think that will be better than something
else that requires half an hour or an hour to do. I think a
cost-effectiveness analysis will favor CT."
Toward that end, Dr. Poon is evaluating CT for several emerging
clinical applications currently within the purview of other imaging
modalities. In particular, he is investigating the utility of CT
for examination of atrial fibrillation, for identification and
assessment of the pulmonary vein, and for detection of atrial
"These are all emerging applications for faster, high-resolution
CT, and I think they will become important applications," said Dr.
Dr. Kazerooni noted that one of the most exciting clinical
potentials for cardiac CT is characterization of atherosclerotic
plaque. Anticipated improvements in its resolution might allow CT
to be used to evaluate and quantify plaque vulnerability. The
results could then be used to plan a patient's therapy.
"You could take a patient with hyperlipidemia and other risk
factors for cardiac disease, quantify the soft plaque burden, and,
potentially, put the patient on very expensive pharmaceuticals to
stabilize or shrink the plaque," said Dr. Kazerooni. "Then you
could scan the patient again in 3 to 6 months and decide whether to
continue that therapy. That could be a very powerful use of CT
technology, which we'll be able to determine once the software
algorithms for the calculations have been validated and the
resolution of CT is another step better than it is now. We're just
playing around with plaque characterization and quantification
right now. It might take another generation of multidetector CT
evolution to make that a reality."
Cardiac MRI imaging
Magnetic resonance also has a considerable history in the
pursuit of cardiac imaging. Beginning in the early 1990s,
researchers demonstrated the potential of cardiac MR angiography
(MRA) as a noninvasive technique for imaging the coronary arteries.
Cardiac MRA required no ionizing radiation and offered the
possibility of visualizing the coronary arteries without the need
for potentially nephrotoxic contrast agents (Figure 3). It could
also obtain tomographic imaging planes in any direction and 3D
MR images are derived from signals produced by protons in the
body. Protons align parallel and antiparallel to the direction of
the primary field, and a small excess of parallel protons gives
rise to a net magnetization vector, which can be altered by a
secondary temporary radiofrequency pulse. When the magnetization
vector recovers to its former position, it releases a radiowave
signal. The relaxation of net vector results from simultaneous
longitudinal (T1) and transverse (T2) relaxations.
The basic pulse sequences used in cardiac MRI are spin-echo (SE)
and gradient echo (GE) sequences or their faster hybrids. Spin-echo
sequences tend to be used for morphologic assessment, and GE
sequences are used to assess valvular lesions, shunts, great
vessels, ventricular function, and wall-motion characteristics.
Recent technologic developments include new scanners that have
numerous coils that can be combined to perform highly parallel
arrays, which offer significant advantages in terms of "massively
increased imaging speeds," said Dr. David Bluemke, Associate
Professor of Radiology and Medicine and Clinical Director of MRI at
The Johns Hopkins Hospital, Baltimore, MD.
"Some of the routine applications in terms of real-time imaging
involve patients who can't hold their breath," said Dr. Bluemke.
"In other cases, we can't do ECG gating for one reason or another,
so we turn on the real-time mode on the scanners and capture the
beating heart. The images are at a lower resolution, but often they
are very adequate for diagnosis and are still better than echo
images in terms of spatial resolution and soft-tissue
"Multiple improvements in MRI hardware and software over the
past several years have substantially enhanced the modality's
imaging capabilities," said Dr. David Bello, Assistant Professor of
Medicine and Director of Cardiac CT and MRI at the University of
California, Irvine, CA. The improvements, including faster
gradients, have enhanced the use of MRI for cardiac imaging, he
Dr. Bello pointed out that the transition to 3T machines has
major implications for imaging of the heart and vasculature. He
noted that high field- strength MRI affords unprecedented
structural delineation, including visualization of small blood
vessels. Newer 3T MRI machines have a 3- to 4-fold higher
signal-to-noise ratio compared with 1.5T machines. Not only do 3T
MRI scanners provide sharper images, they also provide them
"Vast improvements in real-time capability have major
implications for cardiac imaging," said Dr. Steven Wolff, Assistant
Professor of Radiology and Medicine and Director of Cardiovascular
CT and MRI at Columbia University, New York, NY. Two key
improvements have been the ability to image patients who cannot
maintain a breath-hold and the ability to acquire sharper, more
accurate images in patients with irregular heart rhythms.
"Because the heart rate is irregular, no matter how well we
gate, we still get a lot of artifacts in the image," said Dr.
Wolff. "In real time, we don't have that problem. We can also
diagnose certain diseases that we couldn't diagnose before because
of the breath-hold requirement. For example, patients with
constricted pericarditis have paradoxical motion of the septum
through interventricular coupling, which can be seen only during
inspiration. Previously, since all cardiac MR images required a
breath-hold, we would never have an opportunity to see that."
Another enhancement relevant to cardiac MRI is the development
of a new perfusion sequence for evaluating ischemic disease. The
old echoplanar imaging (EPI)-based sequence provided the necessary
speed, but at the cost of multiple artifacts. A new sequence uses
parallel imaging to increase speed and allow GE imaging instead of
EPI. Much higher doses of contrast can be used, resulting in
brighter images that allow better visualization of dark areas.
A third area of cardiac MR improvement relates to visualization,
analysis, and reporting. Software enhancements permit more accurate
assessment of cardiac function and volume.
"For years, MR has been said to be the gold standard for
assessing cardiac function, but one of the dirty little secrets was
that no one really knew when to stop tracing the base of the heart,
because it was hard to tell where the left ventricle ended," said
Dr. Wolff. "That would introduce errors into the calculation. Now
there are tools that allow us to look at long- and short-axis
slices and use the information from the long-axis slice to
determine the correct number of short-axis slices and even what
fraction of a short-axis slice we should use."
Cardiac MR applications
The cardiac capabilities of MR continue to evolve. However, the
modality might already be considered to be the preferred test for
the evaluation of heart function and the precise measurement of
ejection fraction, according to Dr. Bluemke. An example is the
evaluation of patients who are candidates for an implantable
cardioverterdefibrillator (ICD). If the ejection fraction is below
the established threshold, a patient is eligible for ICD
"Most of that work has been done with echocardiography, and the
ejection fraction and volumes have just been estimated visually,"
said Dr. Bluemke. "Some patients are right on the borderline.
Usually, the differences between MR and echo are large, reflecting
estimation with echo versus quantitation with MR."
Evaluation of adult patients with congenital heart disease is
another area in which MR has made its presence known (Figure 4).
Dr. Bluemke noted that echocardiography provides little help in the
evaluation of those patients. Evaluation of an adult patient with
pulmonary artery abnormalities, for example, requires a large
field-of-view. MRI should be chosen for these evaluations because
of its superior ability to assess function, physiology, and flow
Dr. Bello also prefers MR over echocardiography for the
evaluation of patients with cardiomyopathy and heart failure. Every
patient who presents with symptoms of heart failure, an estimated 5
million in the United States, requires assessment of ejection
fraction. MR has proven to be as good as, or better than,
multigated radionuclide angiography, angiography, and
echocardiography for assessment of left ventricular function and
mass. "Cardiac MR clearly surpasses echocardiography because of its
superior spatial resolution for assessment of the endocardial
surface," he said.
A second area in which the use of cardiac MR is growing
considerably is for the assessment of myocardial viability.
Sufficient data have accumulated to support the ability of MR to
determine scar volume and predict a patient's likelihood of
responding to therapy. The accuracy of MR in assessment of
viability on the basis of scar size will be tested in a
10,000-patient study that will begin in the fall, noted Dr. Bello.
The study will involve patients who are potential candidates for
ICD placement, the determination of which is normally based on
ejection fraction criteria. Patients will be randomized to ICDs on
the basis of scar volume, as determined by MR.
Thrombus visualization is a third area of expanding MR use.
Conventionally, ultrasound has been used for this application, but
improvements in inversion recovery times in cardiac sequences allow
MR to assess the presence or absence of thrombus with much greater
accuracy, Dr. Bello noted.
Looking to the future, imaging authorities expect the cardiac
applications of MR to continue to expand as the technology
improves, and the imaging modality proves its value in clinical
studies. Investigations of MR for the characterization of
atherosclerotic plaque are ongoing, similar to the work being done
with CT. "MR has been shown to detect plaque regression in the
aorta, but the relationship between aortic plaque and coronary
calcification or atherosclerosis is not clear," said Dr. Bluemke.
The goal is to improve MR technology and imaging techniques to the
point at which the modality can be used to visualize and assess
plaque in coronary arteries.
Dr. Wolff envisions MR as having the potential to offer a
comprehensive cardiac examination with a single, 30-minute study.
Patients with suspected ischemic disease would be able to undergo a
stress perfusion study, functional study, viability study, and
valve study in a single sitting.
"Evaluation of valvular insufficiency is another area that holds
great potential for MR," Dr. Wolff continued. That potential is
being examined in an ongoing clinical trial.
"While echo may be good for many patients, there clearly are
some patients for whom it is not good," said Dr. Wolff. "Those are
patients with eccentric jets of mitral regurgitation, typically
patients with mitral valve prolapse. Agreement between MR and echo
is good when the jet is central and easy to measure by
echocardiography, but when the jet is eccentric, there can be wide
discrepancies between MR and echo. Because the discrepancies occur
only when the jet is difficult to measure with echo, I think MR
probably is better than echo in these patients."
Dr. Wolff explained that the use of MR to evaluate the coronary
arteries remains a work-in-progress but that progress is being
made. Application of pulse-sequence algorithms to the MR study may
well hold the key to correcting the motion artifact created by
respiration and the beating heart. "Once the motion problem has
been adequately addressed, MR could prove to be as good as or
better than CT for assessment of the coronary arteries," said Dr.
Cardiac imaging turf wars
The expanding use of CT and MR for cardiac applications has,
perhaps inevitably, sparked some controversy about the "ownership"
of the imaging modalities. At this point, the most heated turf
battles revolve around cardiac CT, where radiologists have squared
off with cardiologists in a number of areas. The battle could
spread beyond those 2 specialties.
"Cardiac CT is quite controversial, and it might not be limited
to just those 2 parties," said Dr. White. "I've seen interest in
cardiac CT by surgery groups and internal medicine groups. My
feeling is that, in any given town, the most successful projects
will be those that are collaborative efforts. I think it will be
hard to stay current technologically because obsolescence is
reached so quickly. It will take parties who can justify the
replacement of scanners to stay technologically current. If you
have a partnership between radiologists and cardiologists, for
instance, you could better justify the replacement of the equipment
because it would be used for multiple purposes."
"With respect to cardiac MR, radiologists perform the studies at
some sites, cardiologists at others, and the 2 specialties work
together at still other sites," said Dr. Wolff. Though less
frequent than the conflicts surrounding CT, turf battles
occasionally have arisen.
"Just as in any other area of medicine where the possibility
exists for 2 or more specialties to do the work, there are
conflicts," said Dr. Wolff. "I think the conflicts are manageable.
In the future, I think we'll see both specialties doing cardiac
Dr. Poon said that the debate over who should be in charge
should be decided by training, experience, and competency. Anyone
who has the appropriate training and experience should have the
right to perform cardiac CT or cardiac MR studies.
Toward that end, earlier this year the American College of
Cardiology and the American Heart Association issued a joint
"competency statement," addressing knowledge criteria for cardiac
imaging with CT or MR.
The statement was developed in collaboration with the American
Society of Echocardiography, the American Society of Nuclear
Cardiology, the Society of Atherosclerosis Imaging, and the Society
for Cardiovascular Angiography and Interventions and was endorsed
by the Society of Cardiovascular Computed Tomography.
With technologic advances in scanner hardware, 64-slice CT and
3T MRI offer tremendous utility for noninvasive cardiac imaging.
With faster scanning and higher spatial resolution, both CT and MRI
are increasingly being used to image the coronary arteries and the
heart itself. The features and capabilities of each modality are
applied to specific evaluations, providing improved cardiac
imaging. Future evolutions of CT and MR scanners and reconstruction
choices will offer even faster, clearer, and more useful