Skip to Main Content

Moving pictures

Medicine@Yale, 2007 - Sept Oct

Contents

Expert on image analysis finds guideposts for doctors in the ever-changing body

Though James S. Duncan, Ph.D., has ready access to the School of Medicine’s state-of-the-art imaging technology, he still grapples with a problem faced by the tintype photographers of old—his preferred subjects just won’t sit still.

No organ is more restless than the heart, and the constant movement of the muscle that forms its chambers only adds to the difficulty of interpreting the speckled, shadowy images produced by echocardiography. For more than 15 years, Duncan, vice-chair and Ebenezer K. Hunt Professor of Biomedical Engineering and professor of diagnostic radiology, has worked with cardiologist Albert J. Sinusas, M.D., professor of medicine and diagnostic radiology, to find better ways to extract information about the heart’s health from these images. In 2006, their efforts received a major boost in the form of a five-year, $7.2 million Bioengineering Research Partnership (BRP) grant from the National Heart, Lung and Blood Institute, the second BRP grant Duncan has received to support his research.

Along with ultrasound expert Matthew O’Donnell, Ph.D., the Frank and Julie Jungers Dean of the University of Washington’s College of Engineering, Duncan and Sinusas are devising computer algorithms that automatically locate the heart wall in echocardiographic images despite the myriad shapes this tissue assumes over time as the heart beats. With image-based models of heart-wall thickness and elasticity such as those Duncan and colleagues are building at the bedside, cardiologists could precisely determine what structures have been damaged by a heart attack and how well patients are healing.

Duncan may inherit his technical bent from his father, who served in the Canadian Air Force and went on to work in the telephone industry in New York City. Duncan, born in the Bronx and still a diehard New York Yankees fan, was the first in his family to graduate from college. He earned a master’s degree in electrical engineering at the University of California, Los Angeles, and a Ph.D. at the University of Southern California, both with the help of fellowships from the Hughes Aircraft Company. Duncan worked at Hughes for 10 years during his schooling, but found that he needed a change.

“I enjoyed the aerospace industry, but military-oriented projects weren’t what I wanted to do as a career,” Duncan says. “I even considered switching careers to medicine, but I decided that if I could take all that I had learned and turn it in a new direction, that would be a better course.”

Duncan also collaborates with Dennis D. Spencer, M.D., the Harvey and Kate Cushing Professor of Neurosurgery, finding ways to guide surgeons through brain operations such as those performed to treat epilepsy. Though the brain doesn’t move as dramatically as the heart, its shape does markedly change during surgery, which alters the location of important anatomical landmarks. Christine DeLorenzo, Ph.D., who just earned her graduate degree working with Duncan, found that by training two cameras on the surface of the brain during surgery and feeding the resulting stereo image into computers it is possible to create a mathematical model that provides an up-to-the-second, three-dimensional rendering of the brain as it shifts in the operative field.

By adding magnetic resonance spectroscopy, which detects biochemical changes In small regions of tissue, to the mix, Duncan foresees a day when neurosurgeons treating epilepsy will use “multimodality” imaging—anatomical, functional and biochemical—to accurately place tiny probes in patients’ brains that detect the onset of a seizure and quickly deliver drugs to interrupt it.

Having once observed surgeries guided by images stuck to the operating room wall with masking tape and highlighted in colored pencil, Duncan says that today’s technology presents an embarrassment of scientific riches.

“The whole idea of how you represent information as images, how you look across scales and modalities and problems, that’s exploding,” says Duncan. “It’s a neat thing for those of us who do the analysis, but there’s a lot out there. You need to grab on to certain pieces of it so you can get your arms around something you can really dig into.”

Previous Article
Expert on protein-folding is named Sterling Professor
Next Article
Lyme disease expert is new section chief and Hughes investigator