From Data to Decisions: Reinforcement Learning and Generative AI for Precision Medicine

Abstract: Precision medicine aims to transform complex patient data and clinical knowledge into individualized treatment decisions. However, this remains challenging when treatments are sequential, patient states evolve over time, and relevant evidence is distributed across real-world data, clinical guidelines, and biomedical knowledge.

In this talk, I will discuss how reinforcement learning and generative AI can provide complementary tools for addressing these challenges. Reinforcement learning offers a framework for learning adaptive treatment strategies from longitudinal data, while large language models and knowledge graphs can help organize, integrate, and reason over heterogeneous clinical knowledge. I will highlight methodological considerations including incorporation of domain knowledge, uncertainty, treatment coverage, and interpretability, and illustrate these ideas through applications in precision HIV care.