What question did this study set out to answer?

The aim is to assess the effectiveness of Implicit Q-Learning for managing blood glucose levels using historical data.

March 18, 2026Open Access

Implicit Q-Learning for Offline Reinforcement Learning in Blood Glucose Management: A Cross-Dataset Evaluation Study

Puntos clave

The aim is to assess the effectiveness of Implicit Q-Learning for managing blood glucose levels using historical data.
Evaluated on three continuous glucose monitoring datasets: OhioT1DM, ShanghaiT1DM, and ShanghaiT2DM.
Conducted cross-dataset transfer experiments to test generalization.
Performed ablation studies to validate reward function design.
Implemented sensitivity and robustness analyses for algorithm stability.
IQL achieved time in range of 68.5%, 54.3%, and 78.9% for each dataset.
Over 98% performance retention was observed across different geographic regions and diabetes types.
Algorithm demonstrated stability against variations in hyperparameters and data quality.

Resumen

Offline reinforcement learning enables learning insulin dosing policies from historical data without risky patient interactions. This study evaluates Implicit Q-Learning (IQL) on three real-world continuous glucose monitoring datasets: OhioT1DM (USA, type 1), ShanghaiT1DM (China, type 1), and ShanghaiT2DM (China, type 2). IQL achieved time in range of 68.5%, 54.3%, and 78.9% respectively. Cross-dataset transfer experiments demonstrated exceptional generalization with over 98% performance retention across geographic regions and diabetes types, suggesting that IQL captures fundamental glucose-insulin dynamics rather than dataset-specific patterns. Ablation studies validated our clinically-motivated reward function design, while sensitivity and robustness analyses confirmed algorithm stability across hyperparameter choices and data quality perturbations.

Me gusta

Guardar

Ver artículo completo