In their study involving 91 patients with known or suspected coronary artery disease, the researchers found that the traditional nuclear imaging technique, "single photon emission computed tomography" (SPECT) detected only 53 percent of these microinfarcts that were detected by cardiac MRI. Additionally, 13 percent of study patients with microinfarcts were shown to have none when SPECT alone was used.

"While both cardiac MRI and SPECT are extremely accurate in detecting large infarcts, our study shows that only cardiac MRI systemically detected smaller infarcts that are missed by SPECT," said lead researcher Robert Judd, Ph.D., co-director of the Duke Cardiovascular Magnetic Resonance Center (DCMRC). "The smaller the infarct, the more likely that SPECT will miss it."

"However, in these smaller microinfarcts, the cell death may only travel a short distance," Judd explained. "Since the spatial resolution of SPECT is roughly equivalent to the thickness of the heart chamber wall, it can only detect those infarcts that have traversed a good portion of the chamber wall."

The spatial resolution of cardiac MRI is 60 times greater than SPECT, Judd said, allowing it to pick up these microinfarcts.

During a cardiac MRI examination, which is non-invasive and radiation-free, a patient is guided through the cavity of a large doughnut-shaped magnet. The magnet causes hydrogen nuclei in cells to align, and when perturbed by radio waves, they give off characteristic signals, which are then converted by computers into three-dimensional images of the heart and its structures. While MRI technology itself is 20 years old, only in the past few years has technology improved to the point where accurate images of moving tissues can be taken.

SPECT technology creates a series of "slices" of the area to be studied, with a computer assembling the slices to create an image. Patients are usually given a radioisotope, which provides information about blood flow and metabolism of tissues being studied. In the study, both groups of patients were given a contrast-enhancement agent.

To verify the findings on human subjects, the researchers then performed the same tests on a series of animals, some without heart disease and others with known disease; and then analyzed the heart tissue after testing. The scientists performed the confirming studies in the animals to ensure that the smaller areas detected by cardiac MRI were actual areas of cell death, and not something else.

"In the animals, cardiac MRI and SPECT detected infarcts involving greater than 75 percent of the chamber wall," Judd said. "However, for infarcts involving less than 50 percent of the chamber wall, cardiac MRI detected 92 percent, while SPECT detected only 28 percent."

Judd said that further studies are needed to establish the role of cardiac MRI in the diagnosis of heart disease.

"If we see these spots on cardiac MRI, the patient likely has coronary artery disease," Judd said. "However, if they don't have spots, we can't say for certain they don't have coronary artery disease."

Duke cardiologists estimate that about 30 percent of patients with heart disease find that conventional methods for imaging the heart fall short in providing accurate information by which to guide treatment. They said that MRI provides crisp 3-D views of cardiac anatomy with no interference from adjacent bone or air.

The image quality of cardiac MRI also surpasses that of echocardiography -- another common imaging technique, Judd said. Cardiac MRI can better show physicians how well the heart muscle is contracting, as well as precisely reveal areas of damaged tissue.

"It wasn't until a few years ago that engineers developed scanners fast enough to clearly capture a beating heart," Judd said. "The discipline is still defining itself. We want to advance the field by improving existing cardiovascular imaging techniques and also by creating entirely novel ways to look at the heart and its vessels."

The experiments reported in this study were conducted at Northwestern University. Some team members have since come to Duke to establish the DCMRC earlier this year, including Judd, co-director Raymond Kim, M.D., Anja Wagner, M.D., and Heiko Mahrholdt, M.D. Team members still at Northwestern are Thomas Holly, M.D., Michael Elliott, M.D., Matthias Regenfus, M.D., Michele Parker, Francis Klocke, M.D., and Robert Bonow, M.D.

The study was funded by grants from National Institutes of Health, the Deutsche Forschungsgemeinschaft and the Robert Bosch Foundation.