How AI, Facial Recognition Could Spot Stroke, Other Diseases

by Wall Street Journal

Researchers are investigating the use of artificial intelligence and facial recognition to identify various diseases. Physicians at Johns Hopkins Hospital are training a computer algorithm to detect changes in facial features that might signal stroke-induced brain damage. They upload video images of patients into the training database, and a preliminary study showed the algorithm could identify stoke in 40 previously diagnosed patients with 70 percent accuracy. Biotechnology company FDNA's Face2Gene software program is designed to diagnose 1,500 rare genetic disorders in children via scans of patients' faces. FDNA's Erik Feingold said the goal is early detection. Meanwhile, experts at the Massachusetts Institute of Technology, Massachusetts General Hospital, and nonprofit EverythingALS are exploring facial recognition-based diagnosis/monitoring of amyotrophic lateral sclerosis by analyzing patients' facial movements in videos.

Read more.