An artificial intelligence (AI) system developed by Shinjini Kundu, PhD, a physician and medical researcher at the University of Pittsburgh Medical Center, could find patterns of developing diseases as much as three years earlier than imaging experts.
Her new technology, called Transport Based Morphometry (TBM), uses generative modeling to extract disease markers from medical images to be analyzed separately. The system could help medical imagers recognize signs of disease three years before symptoms present themselves.
“TBM bridges the gap between machine, logic and our fundamental biology to ultimately benefit you, the patients,” Kundu said in a presentation at the 2018 MIT EmTech conference on Sept. 12 in Cambridge, Massachusetts.
Sundu trained the AI system with knee MRI scans of osteoarthritis patients and with the system detected changes in the cartilage that were early signs of the disease.
“TBM allows patterns that were once invisible to be visible,” Kundu said. “It engenders trust in AI based diagnosis and it gives doctors new clues to look for doctors would never have known to even look for and would never have even known existed.”
Kunda was named one of this year’s 35 Innovators Under 35 by MIT Technology Review.
A recent graduate from Dominican University (IL) with a bachelor’s in journalism, Melissa joined TriMed’s Chicago team in 2017 covering all aspects of health imaging. She’s a fan of singing and playing guitar, elephants, a good cup of tea, and her golden retriever Cooper.