Gordon Lecture: Clinically Applicable Deep Learning in Radiology and Ophthalmology

James Brown, Ph.D., is a research fellow at the Athinoula A. Martinos Center for Biomedical Imaging at Massachusetts General Hospital. His background and interests include medical image analysis, computer vision and machine learning. Dr. Brown was the guest speaker at a lecture organized by the MGH Gordon Center. Below is his presentation summary.

Deep learning (DL) has reinvigorated public discourse about artificial intelligence and its role within our society. As of September 2018, there were 13 “AI-based” algorithms approved by the FDA that operate on a wide range of imaging modalities. Dr. Brown gave a broad overview of deep learning and how it relates to other kinds of machine learning. He also presented two applications of DL that are beginning to make some impact in their respective clinical domains: retinopathy of prematurity (ROP), a leading cause of preventable childhood blindness, and glioblastoma (GBM), an aggressive primary brain tumor that carries a poor prognosis. Dr. Brown also presented his group’s open-source software package – DeepNeuro – that has been developed to make DL more accessible to clinical communities and reproducible by all.