STAT+: Are we ready to let AI interpret X-rays all by itself?

You’re reading the web edition of STAT’s Health Tech newsletter, our guide to how technology is transforming the life sciences. Sign up to get it delivered in your inbox every Tuesday and Thursday.

Katie Palmer braved frigid temperatures in Chicago last week to learn about the hot, hot world of using artificial intelligence to analyze medical imaging at the Radiological Society of North America annual meeting. She’s already delivered two brilliant readouts:

• Radiology was an early adopter of AI technology but as researchers and startups forge ahead lean with a new wave of the technology, practices and are still scrambling to catch up with the older generation of tech. Read more here.
• With the rise of large vision models that can perform broader interpretation of imaging like chest X-Rays, there’s a growing debate about how much work should be offloaded to AI — especially given the shortage of radiologists. Read more here. 

Schizophrenia DTx is headed to FDA

Speaking of frigid temperatures, last week I drove 60 terrifying, snowy miles on uncleared Vermont highway to attend a digital health summit at Dartmouth College’s Center for Technology and Behavioral Health. On top of some very interesting perspectives on AI regulation throughout the day, the highlight was the first U.S. presentation of data from an RCT of an experimental digital therapeutic, CT-155, that treats negative symptoms of schizophrenia, including lack of motivation and the inability to feel pleasure.

Continue to STAT+ to read the full story…