Research Combines Machine Learning and Neuroimaging to Aid People with Brain Tumors

Miaomiao Zhang, an assistant professor in the University of Virginia School of Engineering, and her Ph.D. student Jian Wang earned the best poster award at this year’s Information Processing in Medical Imaging international conference, for research to enable better image analysis of the human brain. Their work will benefit patients with brain tumors, for example by reducing the time a patient spends in the operating room and lowering the risk of tissue damage during surgery.

Their imaging research focuses on registration algorithms embedded in a dynamic, 3-D model of the patient’s brain. The registration algorithm allows a surgeon to visualize damage, plan, and evaluate risks of alternative approaches during surgery. The complexity of the brain and how it responds to disease or injury, such as swelling or build up and drainage of cranial fluid, can lead to inaccuracies in the 3-D model. Zhang and Wang’s registration algorithm tracks this “brain shift” more efficiently and with a more accurate representation of the patient’s condition.

Zhang joined the faculty of UVA’s Charles L. Brown Department of Electrical and Computer Engineering, with joint appointment in Computer Science, in August 2019 to pursue interdisciplinary research with UVA’s School of Medicine. “I enjoy doing more than developing algorithms and tools. Real problems are a motivator and driver; we energize each other and stimulate each other’s ideas,” she said.

Several research projects are underway to improve image registration simultaneously with segmentation or reconstruction, not only in neuroscience but also in areas such as maternal and fetal health. This area of research appeals to Wang, who is completing his doctoral research in computer science at UVA Engineering, building on doctoral studies begun at Lehigh University in Pennsylvania and Washington University in St. Louis.

UVA is one of only eight universities in the country where top schools of engineering and medicine are located within a mile of each other, and engineering for health is one of UVA Engineering’s three top research focus areas.

“UVA Engineering provides excellent medical-related resources and opportunities for students like me, who are passionate about exploring challenging bio-imaging topics. Working with a variety of accomplished and talented professors makes research life more interesting, especially when we encounter research bottlenecks,” Wang said. 

Wang had an extraordinary experience at the Information Processing in Medical Imaging conference, in part because this year’s conference marked the organization’s 50th anniversary. “I received valuable suggestions from peers and scientists from all over the world,” Wang said.

In a memorable moment, he discussed the major impacts of his research with professor Stephen M. Pizer from the University of North Carolina at Chapel Hill. Pizer is at the top of Wang’s academic family tree: Pizer was the Ph.D. advisor for Tom Fletcher, UVA associate professor of electrical and computer engineering and computer science. Fletcher was Zhang’s Ph.D. advisor. “This research perfectly fits with my career goal to be a medical image researcher committed to the health care community,” Wang said.

The registration algorithm that earned Zhang and Wang recognition is one of UVA Engineering’s many major contributions to the future of biomedical imaging. This research strength addresses unmet clinical needs and basic science questions to better understand anatomy and function at the molecular, cellular, tissue, and organ scales, and improve guidance and application of medical surgeries.