G2P2C - A modular reinforcement learning algorithm for glucose control by glucose prediction and planning in Type 1 Diabetes

Chirath Hettiarachchi, Nicolò Malagutti, Christopher J. Nolan, Hanna Suominen, Elena Daskalaki

Abstract: Highlights•We propose G2P2C, an Reinforcement Learning (RL) algorithm for glucose regulation.•G2P2C automates treatment by eliminating manual meal estimation and announcement.•It is evaluated in-silico based on an FDA-approved Type 1 Diabetes Simulator.•G2P2C improves performance compared to clinical treatment and RL benchmarks.•We release the code base of G2P2C and provide an online demonstration tool (CAPSML).