RL4health: Crowdsourcing Reinforcement Learning for Knee Replacement Pathway Optimization

05/24/2019

∙

Joint replacement is the most common inpatient surgical treatment in the US. We investigate the clinical pathway optimization for knee replacement, which is a sequential decision process from onset to recovery. Based on episodic claims from previous cases, we view the pathway optimization as an intelligence crowdsourcing problem and learn the optimal decision policy from data by imitating the best expert at every intermediate state. We develop a reinforcement learning-based pipeline that uses value iteration, state compression and aggregation learning, kernel representation and cross validation to predict the best treatment policy. It also provides forecast of the clinical pathway under the optimized policy. Empirical validation shows that the optimized policy reduces the overall cost by 7 percent and reduces the excessive cost premium by 33 percent.

READ FULL TEXT

RL4health: Crowdsourcing Reinforcement Learning for Knee Replacement Pathway Optimization

Challenges for Reinforcement Learning in Healthcare

Feature-Based Aggregation and Deep Reinforcement Learning: A Survey and Some New Implementations

Successive Convex Approximation Based Off-Policy Optimization for Constrained Reinforcement Learning

Evaluating the Perceived Safety of Urban City via Maximum Entropy Deep Inverse Reinforcement Learning

On Estimation and Cross-validation of Dynamic Treatment Regimes with Competing Risks

A optimization framework for herbal prescription planning based on deep reinforcement learning

Policy Optimization in Bayesian Network Hybrid Models of Biomanufacturing Processes

RL4health: Crowdsourcing Reinforcement Learning for Knee Replacement Pathway Optimization

Related Research

Challenges for Reinforcement Learning in Healthcare

Feature-Based Aggregation and Deep Reinforcement Learning: A Survey and Some New Implementations

Successive Convex Approximation Based Off-Policy Optimization for Constrained Reinforcement Learning

Evaluating the Perceived Safety of Urban City via Maximum Entropy Deep Inverse Reinforcement Learning

On Estimation and Cross-validation of Dynamic Treatment Regimes with Competing Risks

A optimization framework for herbal prescription planning based on deep reinforcement learning

Policy Optimization in Bayesian Network Hybrid Models of Biomanufacturing Processes