Deep Reinforcement Learning for Dynamic Treatment Regimes on Medical Registry Data

Ying Liu, Brent Logan, Ning Liu, Zhiyuan Xu, Jian Tang, Yangzhi Wang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

19 Scopus citations

Abstract

In this paper, we propose the first deep reinforce-ment learning framework to estimate the optimal Dynamic Treat-ment Regimes from observational medical data. This framework is more flexible and adaptive for high dimensional action and state spaces than existing reinforcement learning methods to model real life complexity in heterogeneous disease progression and treatment choices, with the goal to provide doctor and patients the data-driven personalized decision recommendations. The proposed deep reinforcement learning framework contains a supervised learning step to predict the most possible expert actions; and a deep reinforcement learning step to estimate the long term value function of Dynamic Treatment Regimes. We motivated and implemented the proposed framework on a data set from the Center for International Bone Marrow Transplant Research (CIBMTR) registry database, focusing on the sequence of prevention and treatments for acute and chronic graft versus host disease. We showed results of the initial implementation that demonstrates promising accuracy in predicting human expert decisions and initial implementation for the reinforcement learning step.

Original languageEnglish (US)
Title of host publicationProceedings - 2017 IEEE International Conference on Healthcare Informatics, ICHI 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages380-385
Number of pages6
ISBN (Electronic)9781509048816
DOIs
StatePublished - Sep 8 2017
Event5th IEEE International Conference on Healthcare Informatics, ICHI 2017 - Park City, United States
Duration: Aug 23 2017Aug 26 2017

Other

Other5th IEEE International Conference on Healthcare Informatics, ICHI 2017
CountryUnited States
CityPark City
Period8/23/178/26/17

ASJC Scopus subject areas

  • Health Informatics

Fingerprint Dive into the research topics of 'Deep Reinforcement Learning for Dynamic Treatment Regimes on Medical Registry Data'. Together they form a unique fingerprint.

Cite this