Medscape Radiology.  2006;7(1) ?2006 Medscape

Editor's Note:
Vivian Lee, MD, PhD, is one of the premier experts in the arena of body and vascular magnetic resonance imaging (MRI). Professor of Radiology and Physiology and Neuroscience, Vice Chair of Research, New York University School of Medicine, New York City, Dr. Lee is a prolific writer and lecturer who speaks before standing room-only audiences of radiologists and cardiologists at major medical symposia such as the American College of Cardiology (ACC) Annual Meeting to underscore the promise of MRI as a 1-stop-shop technique for assessing the heart and coronary arteries.

Dr. Lee met with Pippa Wysong of Medscape to spotlight how radiologists can align incentives with their cardiology brethren to optimize cardiac imaging studies and cease and desist skirmishes over turf and patients, capitalize on emergent MRI technology and clinical applications for cardiac imaging, and foster the future promise of high-field MRI systems.

Dr. Lee is the author of the recently published book, "Cardiovascular MRI: Physical Principles to Practical Protocols" (Lippincott, 2006).

Medscape: The medical community is increasingly fascinated with cardiac MRI. How far along is cardiac MRI at this point?

Dr. Lee: Early on, cardiac MRI was very challenging and frustrating to do. Acquisitions took a long time, patients had to stay in the machine a long time, and images were often blurry and suboptimal. In recent years the technology has come of age. I think most radiologists, cardiologists, and other referring physicians are now aware that MRI can be used on even the sickest of cardiac patients and total study times can be less than 30 minutes. It's time to revisit the use of cardiac MRI in the clinical setting both for diagnosis and for monitoring therapies.

Medscape: What are examples of some of the challenges facing imagers attempting to perfect cardiac MRI studies?

Dr. Lee: For most cardiac MRI studies, images have to be acquired so that they are coordinated with the motion of the heart to avoid blurring. You can achieve this by synchronizing the acquisition to the heart's electrical activity, usually with electrocardiogram (ECG) tracings. This technique has limitations in some patients.

One issue is that ECG signals are distorted by the fact that the patient is in a magnetic field. With imperfect synchronization, images are blurred. In recent years, vector cardiographic gating (VCG) has emerged as a more advanced form of synchronization that reduces the distortion. VCG is a sophisticated method of using the electrical signal from the heart so that the interference from the magnet is reduced. We now get much more robust cardiac gating and better imaging across the board on those patients.

Another big advance in cardiac MRI is a new pulse sequence, generically called balance steady state free procession (b-SSFP) Imaging. Medscape readers would know this by the brand names of True-FISP, FIESTA, or Balanced FFE. b-SSFP is used to look at the flow and function of the beating heart. The images give superb contrast between the blood pool and myocardium, and in significantly shorter times than the older spoiled cine gradient echo imaging techniques. Because they are acquired in shorter times, patients tolerate the breath-holds well and there is usually less motion artifact.

Medscape: How long does it take to acquire a quality image of the heart?

Dr. Lee: It depends on the type of images you're trying to generate. For the standard beating heart images (usually played in a video loop), the entire set of images spanning systole and diastole can be acquired in under 7 to 10 seconds with b-SSP techniques. If you push the technology, using parallel imaging methods, for example, you can do real-time imaging where each picture of the heart is acquired in about 50 milliseconds. The patient can breathe freely and we can look at the heart in real time; that's the direction cardiac MRI is going.

Medscape: How does cardiac MRI compare to other technologies that are used to assess the heart and coronary arteries?

Dr. Lee: MRI has a key advantage over most other imaging modalities when it comes to imaging the heart, brain, and other body parts. That advantage is the ability to differentiate between different kinds of tissue based on multiple, different tissue characteristics. For instance, with CT scanning and x-rays, we rely on density: the denser the tissue the whiter it appears; the less dense it is, the blacker it is. If tissue densities are not very different, then we will be unable to distinguish them on CT or x-ray, unless we administer contrast. With ultrasound or echocardiography (ECG) we're using the ability of the tissue to propagate sound to generate images. MRI, allows many different ways to look at tissue such as T1- or T2-weighted images, fat-suppressed imaging, perfusion imaging, diffusion imaging and so on. There's even MRI spectroscopy which looks at the biochemical signature of different tissues. With MRI the range of tools, plus the technological advances making things faster gives us much more diagnostic power.

Cardiac MRI Is Poised for Prime Time

Medscape: What, specifically, are the advantages of cardiac MRI?

Dr. Lee: There are a several applications where cardiac MRI is especially powerful. At the most basic level the measurement of cardiac function by MRI is considered to be the gold standard technique compared to other methods. MRI offers very high spatial resolution as well as very high tissue contrast -- we can see the borders between the heart muscle and the blood clearly. Some of the artifacts that interfere with ECG and nuclear medicine stress tests, such as due to a patient's large size or "body habitus" are less of an issue in MRI, provided the patient fits into the magnet bore. The very high resolution images that show a lot of image contrast which give us the ability to get good quantitative data about cardiac function. Details about regional wall motion are exquisite - for example, we can provide quantitative assessment of systolic and diastolic dysfunction of each segment of the left (and right) ventricular myocardium. This is important for the early diagnosis of subtle disease and also for monitoring therapies.

Medscape: Can you expand on the use of cardiac MRI in the evaluation of myocardial infarction (MI)?

Dr. Lee: The evaluation of myocardial infarction (MI) is a really hot area now. With MRI, the technique that has been developed relies on the delayed enhancement of infarcts relative to uninfarcted myocardium. Imaging is performed about 10 minutes after the intravenous injection of gadolinium contrast, which is one of the safest agents we use in medicine. If we image using an inversion recovery sequences set to null the uninfarcted myocardium, then areas of infarct will appear bright on the images. Given the high resolution of the imaging technique, we can now see subtle subendocardial infarcts that were impossible to diagnose in the past. This technique opens up a whole new area of application.

In the past, techniques such as ECG, nuclear medicine stress tests and PET generally pick up infarcts only if they involved the majority of the wall thickness wall of the myocardium, so-called transmural infarcts. Subendocardium infarcts are often difficult to detect. For example, a patient might come into the ER and a blood test or an ECG suggests they've had some damage to their heart, yet imaging tests don't show it. We are often not sure whether to send these patients for catheterization or home. With viability MRI, we can detect new infarcts right away, even if they are not transmural. The ability to diagnose these subtle infracts is of huge interest to both the cardiology and emergency medicine communities. Studies in humans and animals have shown that the MR images correlate well with pathologic evidence of infarcts and irreversible damage to the myocardium.

Medscape: Can you provide an example of how the information provided by cardiac MRI findings can affect clinical management decisions in the setting of myocardial infarction?

Dr. Lee: We had a patient with recurring chest pain who had undergone catheterization, but the lesion in the left main coronary artery couldn't be easily treated with angioplasty.

He had left ventricular dysfunction and the question was whether there was enough viable cardiac tissue to justify sending him for coronary artery bypass graft. If the dysfunctional myocardium was merely ischemic, then surgery would be indicated, but if the myocardium was substantially infracted then the benefit of surgery would be more controversial. In this case, we found there were regions of subendocardial infarct in the left anterior descending territory that spanned about half the wall thickness. Plus, he had a few small focal areas of infarct in the left circumflex territory indicated that the amount of myocardium, in jeopardy with the left main coronary artery was enough to warrant surgery.

It's On: Cardiologists and Radiologists Vie for Cardiac MRI Supremacy

Medscape: Are cardiologists and radiologists still embroiled in a turf war over cardiac MRI?

Dr. Lee: The politics of cardiac imaging vary from place to place. In some centers, cardiac MRI has been driven by radiologists, in others the cardiologists are doing all the imaging. At our institution, we have within our department a cardiology-trained imager who works with several radiology-trained imagers to perform cardiac imaging studies. An interdisciplinary approach is rewarding for everyone, especially the patient.

In the broader area of cardiac MRI, a bigger issue is that there is limited expertise in cardiac MRI. It's a pretty new technology, both for radiologists and cardiologists and has a steep learning curve. In my view, it's less of a turf issue and more of an underutilization one.

Medscape: Why hasn't cardiac MRI caught on to a greater degree, given the advantages you've outlined?

Dr. Lee: More than with most other imaging technologies, the quality of cardiac MR images that can be acquired depends on an understanding of the physics of how MR images are generated. Within the field of MRI, cardiac imaging is technically one of the most challenging of MRI areas and most radiologists and cardiologists don't have the background. The machines are there, it's more of a lack of expertise in how to use them for this particular purpose. Techniques such as the steady-state precession techniques and viability imaging require a certain amount of expertise. The best way to become a master of those techniques is to understand some of the actual technology.

Medscape: You obviously have a keen interest in seeing radiologists and cardiologists align their incentives and goals on behalf of improving the standard of care for patients with heart disease. Is that one of the goals of your new book?

Dr. Lee: Precisely. For several years I taught the basic principles of MRI to fellows and residents, and even to some more senior faculty, with a view to explaining how different pulse sequences work and how decisions about setting up the pulse sequences can impact image quality. I've tried to put these lessons together in my book. It starts with basic physics, which includes enough material to be relevant to the current techniques used, and finally through to discussions of the actual protocols that should be used for different cardiac MRI applications. I try to explain how best to acquire the images based on a knowledge of the physics.

Medscape: Are there 1 or 2 clinical pearls from the book that you can share with Medscape readers?

Dr. Lee: One example is in the area of infarction imaging. As I mentioned, with delayed contrast-enhanced viability imaging, the key is that the areas of infarct should look bright, and uninfarcted tissue should look dark. But to make the uninfarcted tissue look dark you have to "null" the uninfarcted myocardium using an inversion recovery time sequence and set the inversion time parameter correctly. A common problem in infarct imaging is that the inversion time is slightly off, and then image contrast can look strange and the infarcts not as apparent as they should be. The book explains how and why that inversion time is important, and how to adjust it if the images aren't coming out the way you think they should.

Another example is the commonly used beating heart "cine" images using the b-SSFP methods. These sequences can be tricky to set up depending on a patient's heart rate and breath-holding ability. During a breath hold, the heart rate may change and image quality can deteriorate. The book explains the principles of these sequences and takes readers through a series of exercises to practice on different patient scenarios. There are also tables than can be copied and used at the console to help guide parameter selection. The difference between a superb imaging study and a mediocre or poor one may be just the selection of 1 or 2 parameters. The devil is in the details.

Medscape: There are a number of cardiac MRI-related books on the market that are targeting radiologists and cardiologists. How is your book different from the rest?

Dr. Lee: There are several outstanding cardiac MRI books out there, but mostly their primary focus is on diagnosis. They illustrate and discuss many examples of different kinds of pathology to aid image interpretation. There are also many excellent MRI physics books out there that focus on the basic techniques of MRI in general, but don't spend much time on imaging protocols, particularly cardiovascular applications. What I've tried to do is focus on cardiovascular imaging and explain the physics at a level that works for clinicians. My goal is to bring clinical imagers up to speed with state-of-the-art MRI techniques that are used in daily practice of cardiac imaging. I also include sections of pitfalls of imaging, including various artifacts that are often seen. My book also covers vascular imaging and MR angiography.

Why Cardiac MRI May Not Fulfill Its Promise

Medscape: Are there limitations to cardiac MRI?

Dr. Lee: Certain subjects cannot tolerate the claustrophobia they feel in the magnet. Others may be too large to fit in the magnet bore. In general, most patients can be studied. One particular limitation is that we do not image patients who have implanted defibrillators or pacemakers. But as with everything, things change. As MRI becomes increasingly important for diagnosis, the research community is trying to solve some of those restrictions. There have been reports from Europe where researchers are looking at putting certain patients with pacemakers in the MRI scanner. These are non-pacemaker-dependent patients and are very carefully monitored. This approach has not been widely accepted (this is definitely a case of don't try this at home), but it signals a new direction that will be interesting to follow

Medscape: In an era where people are increasingly concerned about medical costs, is cardiac MRI considered a cost-effective technology?

Dr. Lee: Cost-effectiveness studies in cardiac MRI studies are still in their infancy. Cardiac MRI has triggered a lot of interest because it's clearly a very powerful tool and could change treatment patterns in many different areas. One interesting question to ask is whether using cardiac MRI in the emergency department for suspected infarcts more cost-effective than use of traditional tests and keeping patients in for observations? Those are studies that need to be done.

Medscape: A lot of people want to know whether 3-Tesla imaging is going to blossom into widespread use. Can you outline your thoughts regarding the future utility of high-field imaging?

Dr. Lee: This is a question I get asked a lot. Three-Tesla whole body scanners are a recent development, and experience in doing cardiac MRI at 3-Tesla is just starting to emerge in the research literature. At NYU, we have two 3-Tesla scanners and do almost all of our cardiac MRI on them. Three-Tesla has some advantages, such as improved signal to noise ratios which can translate into better image contrast, better spatial resolution, shorter acquisition times, or any combination of those. The downside of 3-Tesla imaging, particularly for cardiac applications, is that there are greater susceptibility artifacts. Research scientists in academia and industry are working hard to reduce these artifacts, and there are improvements being made all the time.

In our experience, the theoretical advantage of higher signal-to-noise ratios at 3-Tesla has not been fully realized yet, mostly because this technology is no where nearly as mature as the 1.5-Tesla systems. Right now, I would say 3-Tesla cardiac imaging is probably just about as good as a really good 1.5 Tesla system. Within the next few years, as all the manufacturers and research scientists turn their efforts toward advancing 3-Tesla technology, we will undoubtedly see some major breakthroughs.