Siemens' MR gambit with Duke pays dividends in high-speed cardiology imaging

Symbiosis expands clinical and corporate potential

Siemens has staked a claim on the frontier of noninvasive cardiac testing. The company is building on a collaboration with one of the premier centers in the U.S. dedicated to cardiac MRI, Duke University's Cardiovascular Magnetic Resonance Center.

Heart structure, function, and blood flow are being plumbed on two Siemens 1.5T systems optimized for CVMR. These platforms are charting new territory through qualitative assessment and quantitative measures of heart function, with and without innovative uses of contrast media. High-speed imaging with TrueFISP creates short-axis cine movies, indicating the size of the heart and the thickness of the walls, and, ultimately, helps determine whether tissue damaged by myocardial infarction is stunned or necrotic.

Inversion recovery turboFLASH is used in concert with gadolinium-based contrast media designed initially for use in the central nervous system. These media wash out rapidly from the bloodstream but linger inside dead myocardial tissue whose membranes have torn or disintegrated.

Wolfgang Rehwald, Ph.D., a Siemens senior scientist, is onsite helping to tweak the technology, while experiencing firsthand the challenges that equipment faces in CVMR. Rehwald installs and improves upon pulse sequences based on feedback from the cardiology staff.

His mission is to help Duke researchers implement new ideas. He modifies the computer code and tests it in phantoms. Cardiologists test it on animals and patients, providing clinical feedback. Based on this feedback, Rehwald fine-tunes, hoping to improve the results.

Success in CVMR often depends on speed. And shorter scan times usually breed artifacts.

"I work on understanding artifacts mathematically and finding methods to remove them," he said.

The staff work in a call-and-response mode. Cardiologists sit at the scanner, looking at images as they develop, deciding what other exams to run based on what they see. Rehwald modifies the code to reduce artifacts, improve image quality, or boost speed. Sometimes the goal is simply to make life easier for the cardiologists. Any given pulse sequence can have 20 parameters, he said. Getting the best possible images may require fiddling with some or many of them.

"Whatever you can do to automate this process speeds patient throughput," he said. "Time is money."

Rehwald previously worked for GE Medical Systems onsite at Johns Hopkins University. The two codirectors of the Duke CVMR Center, biomedical engineer Robert Judd, Ph.D., and cardiologist Dr. Raymond Kim, previously worked at Northwestern University. While there, they invented a technique called MR delayed contrast enhancement, which uses contrast media to detect myocardial damage sustained in heart attacks. The dead cells soak up the contrast to appear hyperenhanced. The technique can also be combined with studies of contractile function to help physicians predict which areas of the heart will respond best to bypass surgery, angioplasty, or revascularization.

The Duke center opened June 3, 2002, with a single 1.5T Siemens Sonata in place. A second matching system was added in early 2003, and a third may be installed some time in the future. Eventually the scanners might be used to explore therapeutic applications. For the time being, diagnostic techniques are imposing enough.

Rehwald is working on several projects. One involves the development of single-shot sequences that acquire enough data to determine myocardial viability within a cardiac cycle. Trying to get a lot of data in a short time negatively affects spatial and temporal resolution unless a workaround is applied. The Duke team is now experimenting with Siemens' iPAT (integrated parallel acquisition technique), whereby multiple coils acquire data simultaneously, as one possible solution.

Another project involves MR angiography as a tool to assess the coronary arteries. The images might be different from the ones now being made using MR.

"They might be more like what you get from a cardiac cath," Rehwald said. "We don't know for sure. We're still trying it out."

If successful, its application would not be confined to the cardiovascular system.

"The technique might be used for renal angiography or for the whole body," he said.

Arrhythmia patients are particularly challenging. Gating images to the R wave of the heart does not account for irregular beats. But retrograde imaging-adjusting the scan to select just the correct data depending on the heart rate-might.

"They'd ask, 'Couldn't we do that?' And I would say, 'I can do something,'" he said.

Research at the Duke CVMR Center has underscored MR's strength in evaluating congenital heart disease, particularly in structural variations, such as the transposition of a ventricle, whereby the ventricle is morphologically right when it should be left. MR can be used to measure left ventricular function and is also proving useful in the evaluation of young children, allowing the visualization of anomalous blood vessels without cardiac catheterization.

MR's ability to distinguish subtle detail and measure functional indicators of cardiac health, along with applications that differentiate stunned from necrotic tissue, may one day win the hearts of cardiologists, but only if the techniques are adequately developed and communicated. Siemens is creating joint education symposia with the Duke CVMR Center for physicians who would like to learn cardiac MR techniques during week- and month-long courses, as well as shorter term preceptorships. Duke is also training cardiology fellows in CVMR.

Meanwhile, Siemens is gaining a better understanding of how to evolve its technology by collaborating with Duke. There is a symbiosis between the Duke staff and Rehwald, as well as Duke University and Siemens.

"It's not only me supporting them but them supporting Siemens," he said. "They tell us what is useful and what's not."