Reinforcement Learning For Survival, A Clinically Motivated Method For Critically Ill Patients

TL;DR

This paper introduces a clinically motivated reinforcement learning approach for critically ill patients, providing interpretable value functions and demonstrating its effectiveness on sepsis data, aligning with medical expertise.

Contribution

It proposes a new control objective with medical interpretability and adapts it into a practical Deep RL algorithm for critical care treatment optimization.

Findings

Results align with clinical knowledge

Method applicable to any value-based Deep RL

Demonstrated on large sepsis cohort

Abstract

There has been considerable interest in leveraging RL and stochastic control methods to learn optimal treatment strategies for critically ill patients, directly from observational data. However, there is significant ambiguity on the control objective and on the best reward choice for the standard RL objective. In this work, we propose a clinically motivated control objective for critically ill patients, for which the value functions have a simple medical interpretation. Further, we present theoretical results and adapt our method to a practical Deep RL algorithm, which can be used alongside any value based Deep RL method. We experiment on a large sepsis cohort and show that our method produces results consistent with clinical knowledge.