Trajectory Inspection: A Method for Iterative Clinician-Driven Design of Reinforcement Learning Studies
arXiv:2010.04279
Abstract
Reinforcement learning (RL) has the potential to significantly improve clinical decision making. However, treatment policies learned via RL from observational data are sensitive to subtle choices in study design. We highlight a simple approach, trajectory inspection, to bring clinicians into an iterative design process for model-based RL studies. We identify where the model recommends unexpectedly aggressive treatments or expects surprisingly positive outcomes from its recommendations. Then, we examine clinical trajectories simulated with the learned model and policy alongside the actual hospital course. Applying this approach to recent work on RL for sepsis management, we uncover a model bias towards discharge, a preference for high vasopressor doses that may be linked to small sample sizes, and clinically implausible expectations of discharge without weaning off vasopressors. We hope that iterations of detecting and addressing the issues unearthed by our method will result in RL policies that inspire more confidence in deployment.
arXiv commentTo appear in AMIA 2021 virtual informatics summit
References in corpus (5)
- Continuous State-Space Models for Optimal Sepsis Treatment - a Deep Reinforcement Learning Approach
- Evaluating Reinforcement Learning Algorithms in Observational Health Settings
- Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models
- POPCORN: Partially Observed Prediction COnstrained ReiNforcement Learning
- Understanding the Artificial Intelligence Clinician and optimal treatment strategies for sepsis in intensive care