Dr. Lily Peng: AI for Ophthalmology and the Challenges of AI in the Real World
NEJM AI Grand Rounds
00:00
How Deep Learning Can Predict Disease Pathology
Deep warning models are essentially just learning a bunch of shortcut features that are correlated with BMI, age, and gender. So we always do these baselines in the same data set to make sure that whatever we're adding with the image is actually additive or there's actual additional power there. You have to know that those things are potential confounders in the first place though to include them in a baseline model, right?
Transcript
Play full episode
