Patient level simulation and reinforcement learning to discover novel strategies for treating ovarian cancer

TL;DR

This paper develops a reinforcement learning framework to simulate ovarian cancer treatment pathways and discover potentially effective novel treatment strategies using real-world data.

Contribution

It introduces a new reinforcement learning environment tailored for ovarian cancer treatment and applies model-free RL to identify promising therapeutic regimens.

Findings

RL can model complex treatment trajectories

Potential to identify novel treatment strategies

Framework adaptable to other cancers

Abstract

The prognosis for patients with epithelial ovarian cancer remains dismal despite improvements in survival for other cancers. Treatment involves multiple lines of chemotherapy and becomes increasingly heterogeneous after first-line therapy. Reinforcement learning with real-world outcomes data has the potential to identify novel treatment strategies to improve overall survival. We design a reinforcement learning environment to model epithelial ovarian cancer treatment trajectories and use model free reinforcement learning to investigate therapeutic regimens for simulated patients.