Learning to predict patient response: what works and what doesn't?

ABSTRACT

Predicting patient response to treatment is key to personalized medicine and the development of new therapeutics. Recent advances in molecular and cell modeling provide new ways to address this task. Surprisingly, in some of these modalities foundation models cannot outperform simple baselines. In this talk, I will summarize the state-of the-art in this field, describe novel multimodal approaches that translate visual tumor characteristics into molecular descriptors and discuss ways to incorporate drug information into response models. 

SHORT BIO

Regina Barzilay is a School of Engineering Distinguished Professor of AI & Health in the Department of Computer Science and the AI Faculty Lead at MIT Jameel Clinic. She was recently named on Time100 Most Influential People in AI 2025 and develops machine learning methods for drug discovery and clinical AI. In the past, she worked on natural language processing. Her research has been recognized with the MacArthur Fellowship, an NSF Career Award, the AAAI Squirrel AI Award for Artificial Intelligence for the Benefit of Humanity, and the IEEE Frances E. Allen Medal for innovative machine learning algorithms that have led to advances in human language technology and demonstrated impact on the field of medicine. Regina is a member of the National Academy of Engineering, National Academy of Medicine, and the American Academy of Arts and Sciences.