arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.10312cs.AIcs.LGq-bio.QM
收藏

用于血糖控制的端到端离线强化学习

End-to-end Offline Reinforcement Learning for Glycemia Control

Diabeloop(迪亚布洛普公司)

Tristan Beolet, Alice Adenis, Erik Huneker, Maxime Louis

AI总结:

针对闭环血糖控制过度依赖模拟器的问题,提出基于真实患者数据训练的离线强化学习智能体,并利用端到端个性化流程与离线策略评估消除模拟器需求,同时估计临床相关指标。

摘要:

开发用于I型糖尿病血糖控制的闭环系统严重依赖于模拟患者。提高这些闭环的性能和适应性会增加过拟合模拟器的风险。这可能带来严重后果,尤其是在模拟器未能忠实(甚至完全未)捕获的异常情况下。为解决此问题,我们提出使用在真实患者数据上训练的离线强化学习智能体来进行血糖控制。为进一步提升性能,我们提出了一种端到端的个性化流程,该流程利用离线策略评估方法完全消除对模拟器的需求,同时仍能估计与糖尿病相关的临床指标。

英文摘要:

The development of closed-loop systems for glycemia control in type I diabetes relies heavily on simulated patients. Improving the performances and adaptability of these close-loops raises the risk of over-fitting the simulator. This may have dire consequences, especially in unusual cases which were not faithfully-if at all-captured by the simulator. To address this, we propose to use offline RL agents, trained on real patient data, to perform the glycemia control. To further improve the performances, we propose an end-to-end personalization pipeline, which leverages offline-policy evaluation methods to remove altogether the need of a simulator, while still enabling an estimation of clinically relevant metrics for diabetes.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑