Compartmental Recurrent Neural Networks for Modeling Glucose-Insulin Dynamics

We introduce the Compartmental Recurrent Neural Network (COMP-RNN), a novel method for modeling glucose-insulin dynamics in type 1 diabetes mellitus patients. By integrating physiological knowledge and topology into recurrent neural networks, the COMP-RNN significantly improves predictive accuracy and parameter efficiency compared to traditional models used in control. Simulated patient data validate its superior performance and demonstrate that the COMP-RNN’s internal states reflect key physiological patterns, proving its potential to improve artificial pancreas systems.