| Magnetic resonance imaging is
ideally suited for visualizing the cardiovascular
system by providing images with high spatial and
temporal resolution. These images are acquired in
the absence of ionizing radiation, in any
tomographic plane, and without interference from
surrounding bone or soft tissue. Magnetic
resonance imaging can be used to assess cardiac
morphology, function, and valvular disease and to
evaluate the anatomy of the great vessels and
peripheral vasculature. Eventually, it may be
possible to perform myocardial perfusion studies
and to image the coronary arteries noninvasively.
This review addresses the current clinical use of
magnetic resonance imaging in cardiovascular
medicine and briefly discusses future
applications. |
ardiovascular
disease remains the leading cause of mortality in the
USA, accounting for 40% of all deaths. The annual cost
for diagnosing and treating these diseases is staggering,
with some estimates as high as $250 billion. Therefore,
the health care system would benefit from an accurate
means of imaging the cardiovascular system. These
benefits include not only simply making a diagnosis or
demonstrating the effects of treatment or preventive
measures, but also determining the risk for future
cardiac events, predicting survival, providing insight
into physiology and pathology, and perhaps even
decreasing the length of hospitalizations or minimizing
the need for more expensive invasive testing.
MAGNETIC RESONANCE IMAGING
Magnetic resonance imaging (MRI), a noninvasive method
that is ideally suited for visualizing the cardiovascular
system, provides images that have high spatial and
temporal resolution. These images are acquired in any
tomographic plane, without ionizing radiation or
interference from surrounding bone or soft tissue.
For routine clinical applications, image acquisition
is based on observing the interaction or behavior of
hydrogen nuclei, the magnetic field, and radiofrequency
energy. Three basic steps are necessary for imaging.
First, the patient is placed in the magnetic field, and
hydrogen nuclei become aligned with the field. Second,
radiofrequency energy is momentarily applied to the
system to perturb the aligned nuclei. Finally, a
radiofrequency signal is received as the perturbed nuclei
return to equilibrium. The received signal is transformed
(using Fourier methods) into image data.
Unlike ionizing radiation, magnetic fields and
radiofrequency energy do not damage DNA. The imaging
procedure is safe for the patient because no known
harmful biologic effects have been demonstrated at field
strengths of currently available commercial scanners.
Nevertheless, certain metallic objects and devices can
interact with the magnetic field. Contraindications to
MRI include intracranial aneurysmal or hemostatic clips,
pacemakers or implantable defibrillators, pre-6000
Starr-Edwards mitral valve prostheses, Poppen-Blaylock
carotid clamps, McGee stapes piston implants, or any
metal located near a vital structure (for example, a
metal splinter in the eye). Patients with sternal wires
or stents used with percutaneous transluminal coronary
angioplasty can be imaged safely.
A pulse sequence is a series of gradient and
radiofrequency pulses produced by the scanner to acquire
image data. The 2 primary sequences used in cardiac
imaging are the spin-echo and gradient-echo sequences (Figure 1). The
spin-echo, or dark blood, sequence provides excellent
spatial resolution and is used to study anatomy and
morphology. The gradient-echo, or bright blood, sequence
samples multiple data points during the cardiac cycle.
Because there is intrinsic contrast between flowing blood
and static tissue, no contrast agents are required. The
resulting gradient-echo images can be displayed in a
continuous loop of ventricular contraction and
relaxation, so-called cine-MRI, to assess regional and
global ventricular function.
APPLICATIONS
Cardiac function and morphology
Magnetic resonance imaging is unique in its ability to
image in multiple planes (or slices). In fact, the
operator may select any oblique plane. Multiple slices in
any orientation will ensure that the entire structure is
examined. Common orientations used in MRI of the heart
include the vertical long axis (so-called 2-chamber
view), horizontal axis (so-called 4-chamber view), and
short axis (see Figure
1). From these standard views, cardiac
chamber sizes and volumes can be measured and calculated.
Wall thickness can be measured to determine myocardial
mass. Unlike other imaging modalities, MRI can provide
accurate information on the right ventricle without
making any geometric assumptions. Magnetic resonance
imaging is invaluable in visualizing complex congenital
heart disease. Venoarterial, atrioventricular, and
ventriculoarterial connections are readily identified, as
are chamber morphology, position, and relation to other
visceral organs and great vessels. Magnetic resonance
imaging can demonstrate the location, extent, and
attachment of intracardiac and paracardiac masses.
Generalities regarding tissue composition can be made
based on the signal intensity and homogeneity, because
there are different relaxation properties between normal
and diseased tissue. In a similar manner, MRI can help
elucidate infiltrative processes such as hemochromatosis,
amyloidosis, and fatty replacement of myocardium in right
ventricular dysplasia.
Ventricular function
Global and regional ventricular function can be
measured from cine-MRI. The ejection fraction is
calculated from tomographically determined end-systolic
and end-diastolic volumes. Abnormal regional contraction
due to failure of the ventricular wall to thicken can be
seen with cine-MRI. Ventricular remodeling and myocardial
thinning after myocardial infarction can be appreciated
easily (Figure
2).
Valvular assessment
In cine-MRI, turbulent blood flow creates an area of
low signal (signal void) due to dephasing
(asynchronous precession) of the hydrogen nuclei.
Accordingly, regurgitant flow is identified by a jet of
signal loss in the receiving chamber (Figure 3b
and c).
Likewise, turbulent flow is seen with stenotic valve (Figure 3a).
A qualitative assessment of the degree of valvular
disease can be made based on the size of the jet, the
duration of the jet in the cardiac cycle, and the
presence of a proximal convergence zone (an area of flow
acceleration) on the opposite side of the valve.
After a radiofrequency pulse has been delivered, the
received signal from the body has both amplitude and
phase. A technique known as phase velocity mapping can
determine the velocity of hydrogen atoms traveling within
or through a given plane. The scanner can encode
velocities as high as 5 m/s. Whereas previous figures
have used amplitude data to construct anatomic images,
phase images display the velocity of moving blood in much
the same way as Doppler echocardiography (Figure 4). To
interrogate an area of turbulent flow, the operator can
draw a region of interest around turbulent flow from
which peak and mean velocities can be measured and flow
can be calculated. Shunts (using Qp/Qs calculation) can
be evaluated by calculating the flow through the
ascending aorta and main pulmonary artery.
Pericardial evaluation
Pericardial effusions can be defined, and some
generalities about fluid composition can be made, based
on signal intensity. The thickness of the pericardium can
be measured and should be <3 to 4 mm.
Magnetic resonance angiography
Magnetic resonance angiography (MRA) has increasingly
become an accepted method for visualizing both venous and
arterial vessels. Most MRAs can be performed without
administration of a contrast agent because there is
intrinsic contrast between flowing blood and static
tissue, although contrast agents may improve image
quality in certain circumstances. Magnetic resonance
angiography depicts normal laminar flow as bright signal,
whereas turbulent, diminished, or absent flow is depicted
as signal voids.
Intimal flips, true and false lumen identification,
thrombus formation, origin of the tear, involvement of
the brachiocephalic vessels, and the presence of aortic
regurgitation or hemopericardium all can be demonstrated
during an MRI examination for aortic dissection (Figure 5).
Similarly, aortic aneurysms are well delineated with MRA,
and 3-dimensional reconstruction of multislice transaxial
images is helpful in planning surgical repair (Figure 5). Besides
aortic imaging, the carotid, renal, and peripheral
vasculature can be evaluated for stenoses and for
suitability of distal vessels for grafting (Figure 6).
Myocardial perfusion imaging
Myocardial perfusion can be assessed either directly
or indirectly using MRI. In a method similar to stress
echocardiography, ischemic wall motion changes during an
infusion of dobutamine can be observed. Contrast-enhanced
MRI to evaluate perfusion presently uses a first-pass
technique during which a bolus of an MRI contrast agent
is injected intravenously, and serial MRI images track
the bolus through the cardiac chamber and finally into
the myocardium. In the presence of coronary artery
disease, inhomogeneities in perfusion can be induced by
an infusion of dipyridamole or adenosine. Myocardium
perfused by a stenotic coronary artery shows lower signal
intensity as well as a delay in the appearance of
contrast.
Magnetic resonance spectroscopy
Magnetic resonance spectroscopy (MRS) is increasingly
used to assess myocardial metabolism. Other nuclei
besides hydrogen that are MRI sensitive include
phosphorus-31, carbon-13, fluorine-19, and sodium-23.
Magnetic resonance spectroscopy acquires the spectra of a
particular nuclear species in a selected region of
tissue. For example, phosphocreatine and adenosine
triphosphate levels can be measured and the bioenergetic
profile of the myocardium investigated.
Coronary MRA
Imaging the coronary arteries has been one of the most
important advances in the management of coronary artery
disease. Estimates are that 1.4 million cardiac
catheterizations are performed each year in the USA, with
as many as 20% demonstrating no critical stenosis.
Conventional x-ray angiography carries a small risk and
involves admission to a hospital, local anesthesia and
mild conscious sedation, arterial puncture, the passage
of catheters within the vessels, the injection of
radiopaque contrast media, and the exposure of the
patient and staff to ionizing radiation. At present,
there is no satisfactory alternative, and, thus, the
possibility of noninvasive coronary MRA has been
considered by many to be the Holy Grail of
MRI.
Magnetic resonance angiography has challenged
conventional carotid, renal, and peripheral angiography.
However, coronary MRA has been hindered by a combination
of obstacles. These include the small caliber of the
coronary arteries, their tortuous course, and their
encasement within epicardial fat (a high-signal tissue).
Unlike other vessels, the coronary arteries move
considerably during image acquisition from a combination
of both cardiac and respiratory motion. These challenges
are not insurmountable, as MRI techniques continue to
improve over the years.
A technique for imaging the coronary arteries that has
met success is one that employs both cardiac gating and
respiratory compensation using navigator echoes. Cardiac
gating ensures image acquisition of the coronary arteries
during diastole when the heart is relatively still and
coronary blood flow is maximal. Early attempts with
breath-holding to minimize respiratory motion were
disappointing because the patients often became fatigued
after several breath-holding maneuvers, and each
successive breath-hold became less reliable.
Subsequently, a navigator echo approach using a
pencil beam tag to track the motion of the
diaphragm ensures proper alignment and position
registration of the coronary arteries during quiet
breathing (Figure 7a and
b). A
fat-suppression pulse nulls the signal from epicardial
fat, and a surface coil is placed on the patient's chest
to improve signal from the coronary arteries and reduce
noise that would degrade image quality. A block of
multiple transaxial slices is acquired (about 2-mm thick)
through the chest, analogous to the slices in a loaf of
bread. Because a coronary artery will transverse through
multiple slices, a computer-generated reconstruction of
the artery usually is performed to visualize the artery
along its course (Figure
7c).
In its current state of development, coronary MRA can
evaluate only the proximal two thirds of the length of
the major epicardial vessels and will not replace x-ray
angiography as we know it today. Instead, MRA could be
used to diagnose high-grade stenoses of the left main or
proximal left anterior descending arteries (the 2 main
indications for revascularization) and to screen
symptomatic patients having chest pain syndromes or
asymptomatic patients having multiple risk factors.
Coronary MRA continues to be an area of active research
to improve image quality. With the future introduction of
faster imaging sequences, coronary-specific surface
coils, and intravascular MRI contrast agents
(administered intravenously), signal from the coronary
arteries should increase and thus significantly improve
the quality of the images.
Ultimately, for MRI to play a key role in the
understanding and management of coronary artery disease,
MRI should not be as good as conventional x-ray
angiography. Rather, MRI should be better, since
MRI has the potential to image the wall of the
vessel rather than simply to image the lumen of
the vessel. The potential of MRI to characterize plaque
by measuring lipid concentration would have important
prognostic and therapeutic implications.
THE ONE-STOP SHOP
Imagine a single, noninvasive test that assessed
cardiac morphology, function, valvular disease,
myocardial perfusion, and coronary arteries as well as
evaluated the anatomy of the great vessels and peripheral
vasculature. Magnetic resonance imaging has the potential
to provide this comprehensive examination in one sitting
at considerably less risk and cost to the patient. With
rapid imaging techniques, it should be possible to
complete a comprehensive study within an hour (perhaps
costing $1500 to $2000). Although MRI and MRA may seem
expensive as screening tools, they may actually be cost
effective by eliminating the need for an echocardiogram
($800), nuclear perfusion imaging ($1200), and cardiac
catheterization ($4000).
With so many different modalities, it is often
difficult for the clinician to select the appropriate
test or series of tests to evaluate the patient. For a
test to be useful to the clinician, it should provide a
diagnosis, effect a management strategy, and provide
insight into the prognosis. The test or series of tests
should not be redundant (i.e., does not add anything) but
rather be confirmatory (i.e., increases understanding or
confidence) or independent (i.e., provides relevant
information beyond what is already known). So, when
should MRI be considered? This question is best answered
by asking, When should MRI not be
considered? Magnetic resonance imaging should not
be considered if there is a contraindication for imaging.
Magnetic resonance imaging should not be considered in
patients with arrhythmias, especially rapid atrial
fibrillation. Arrhythmias degrade image quality due to
irregular cardiac gating and incomplete relaxation
between excitations. Magnetic resonance imaging should
not be considered for the evaluation of smaller, randomly
moving objects (such as thrombi and valvular vegetations)
that are not well visualized with MRI. This is because
image data are acquired over multiple cardiac cycles, and
an imaged object must be in the identical location in
each successive cardiac cycle. Although it is possible to
measure heart rhythm and rate, blood pressure, and oxygen
saturation while the patient is in the magnet, clinicians
understandably feel uneasy with critically ill patients
who are isolated from them in the bore of the magnet.
Occasionally, metal objects (such as sternal wires and
left internal mammary artery clips) may obscure an area
of interest due to distortion in the magnetic field
caused by these objects. Otherwise, MRI and MRA would be
indicated in all other forms of cardiovascular disease.
So, what has kept cardiac MRI from becoming the
imaging method of choice? It is not the value of the
data, the quality of the data, or the cost. Rather, it is
lack of awareness, lack of imaging expertise, lack of
convincing literature that compares and contrasts imaging
modalities, and lack of trust in the reliability of a new
technology and new consultants.
The role of cardiovascular MRI and MRA will continue
to expand, as they are low-risk, cost-effective
technologies. Clinicians should remain mindful of these
powerful diagnostic imaging modalities.
The reader is directed to the selected review articles
and textbooks listed in the general references for
further readings on cardiovascular MRI.
| General references |
| 1. |
Pohost
GM: Cardiovascular Applications of Magnetic
Resonance. Mt. Kisco, N.Y.: Futura Publishing
Co., 1993. |
| 2. |
Blackwell
GG, Cranney GB, Pohost GM: MRI: Cardiovascular
System. New York: Gower Medical Publishing,
1992. |
| 3. |
Marcus
ML, Schelbert WR, Skorton DJ, Wolf GL: Cardiac
Imaging. Philadelphia: W. B. Saunders,
1991:732976. |
| 4. |
Lawson
MA, Pohost GM: Radionuclide and magnetic
resonance methods. In Kelley WM, ed: Textbook
of Internal Medicine. Philadelphia: J. B.
Lippincott Co., 1997:517527. |
| 5. |
Reichek
N, ed: Cardiac Magnetic Resonance Imaging. Cardiology
Clinics 1998;16:125351. |
| 6. |
Passariello
R, De Santis M: Magnetic resonance imaging
evaluation of myocardial perfusion. Am J
Cardiol 1998;81:68G73G. |
| 7. |
Duerincks
AJ: MRI of coronary arteries. Int J Card
Imaging 1997;13:191197. |
| 8. |
Steffans
JC, Sakuma H, Bourne MW, Higgins CB: Magnetic
resonance imaging in ischemic heart disease. Am
Heart J 1996;123:156173. |
| 9. |
Van
der Wall EE, Vliegen HW, de Roos A, Bruschke AV:
Magnetic resonance imaging in coronary heart
disease. Circulation
1996;94:27232739. |
| 10. |
Globits
S, Higgins CB: Assessment of valvular heart
disease by magnetic resonance imaging. Am
Heart J 1995;129:369381. |
| |