�����Ƶ

From CT scans to at-home COVID tests, technology is revolutionizing healthcare. By working across disciplines, faculty at Pace are turning groundbreaking ideas into reality.

Juan Shan, PhD, an associate professor of computer science in the Seidenberg School of Computer Science and Information Systems, is focused on applying artificial intelligence and machine learning to analyzing medical imaging—things like breast ultrasounds and knee MRIs. The basic idea is to take advantage of those advanced models in machine learning and apply those to the medical domain to help solve medical problems.

“In computer science, we know more about machine learning and computer vision techniques than about medical problems,” Shan says. “But we want to apply these techniques to help doctors solve the medical domain problems.”

Shan and her students are using machine learning and computer vision techniques to help doctors solve problems detecting the precise details of a tumor more accurately and efficiently or providing a second opinion on the severity of knee osteoarthritis.

“My students here at Pace get involved in my research projects,” she says. “Research informs my teaching, and I can always bring new ideas back to my classroom and discuss them with my students.”

Shan and her doctoral students design Computer Aided Diagnosis Systems (CADs), systems that are programmed to do particular tasks—pinpoint location, measure joints, estimate size. Her primary focus is developing robust and efficient CAD algorithms to help doctors analyze medical images, discover distinguishing features, and classify data utilizing machine learning methods.

In one recent research endeavor, for example, Shan created a system where a computer was able to estimate the severity of osteoarthritis in a hand after learning how to read and interpret thousands of X-rays—a technique that Shan says helps save on labor and time for medical professionals that would otherwise be manually diagnosing the severity of the osteoarthritis.

“Research informs my teaching, and I can always bring new ideas back to my classroom and discuss them with my students.” —Juan Shan, PhD

At the College of Health Professions, Assistant Professor John Damiao, PhD, has spent the past several years figuring out how to leverage gains in scanning technologies to build customized wheelchairs tailored toward the needs of an individual user. His work makes the wheelchairs more comfortable, and potentially helps reduce the risk of future injury.

“Traditionally, the wheelchair user is seated in a molding bag that makes an imprint of their shape,” says Damiao. “The molding bag is hardened, and that bag shape is scanned, sent off to a company and they make a cushion from that imprint.”

But his work is driving a paradigm shift in how wheelchair seats are constructed.

“The problem with the traditional method is that the person is sitting in a loaded fashion, and their body contours are being distorted from sitting in a loaded fashion,” he says. “The innovation in my research is scanning the person directly, or in an unloaded position, which should make for a more accurate custom contoured seat.”

Custom contoured seating refers to wheelchair seating systems to fit people with severe deformities—whether its postural or a skeletal deformity, and they can’t sit in a typical linear wheelchair seating system because it would cause discomfort, or eventually cause pressure ulcers because of the mismatching of their shape to what is a typical wheelchair seat shape, explains Damiao.

Pressure injuries kill 60,000 people are a year and are, as Damiao describes, a “$10 billion per year problem.” Damiao’s inventive use of custom-contoured seating utilizes ever-improving technology to potentially save lives.

As a 21st century occupational therapist, Damiao understands that, while academic research is vital and will continue to be vital, it is just as important to be able to leverage technological advances to implement changes rapidly. He hopes that increased interdisciplinary collaboration—for instance, better collaboration between those developing healthcare technologies and researchers—can help take the theoretical into the practical much faster, and thus positively impact lives.

As per the work of Shan and Damiao, it seems that when it comes to the intersection of health and tech, �����Ƶ is certainly building an effective algorithm.