Dynamic Measurement Scheduling for Event Forecasting using Deep RL

TL;DR

This paper introduces a deep reinforcement learning approach to optimize measurement scheduling in critical care, balancing cost and predictive accuracy, outperforming heuristics and physicians in simulations and real ICU data.

Contribution

It presents a novel deep RL framework for dynamic measurement scheduling that adapts to patient history and scales to large action spaces.

Findings

Outperforms heuristic scheduling with higher predictive gain and lower cost.

Reduces measurements by 31% in ICU data.

Improves predictive gain by a factor of 3 over physicians.

Abstract

Imagine a patient in critical condition. What and when should be measured to forecast detrimental events, especially under the budget constraints? We answer this question by deep reinforcement learning (RL) that jointly minimizes the measurement cost and maximizes predictive gain, by scheduling strategically-timed measurements. We learn our policy to be dynamically dependent on the patient's health history. To scale our framework to exponentially large action space, we distribute our reward in a sequential setting that makes the learning easier. In our simulation, our policy outperforms heuristic-based scheduling with higher predictive gain and lower cost. In a real-world ICU mortality prediction task (MIMIC3), our policies reduce the total number of measurements by $31\%$ or improve predictive gain by a factor of $3$ as compared to physicians, under the off-policy policy evaluation.