An advanced hybrid automated insulin delivery system for insulin-dependent diabetes based on successive linearization...

Hybrid automated insulin delivery (hAID) systems represent the most advanced therapy for type 1 diabetes (T1D). Current systems rely on linear or linearized models of glucose homeostasis, which may compromise prediction accuracy and, in turn, timely decision-making by the controller. Physiological variability further complicates insulin requirements, underscoring the need for controllers that adapt dynamically and reduce user burden.

We introduce the University of Bern (UniBE) hAID system, a framework based on successive linearization model predictive control (MPC). The controller integrates basal insulin infusion with the insulin bolus delivery module for meal-related and corrective bolus dosing, adapting bounds in real time to glucose dynamics while accounting for both automated and user-initiated inputs. In-silico evaluation was conducted using the commercial version of the FDA-accepted UVa/Padova metabolic simulator across nine scenarios involving persistent and time-varying errors in meal timing, carbohydrate estimation, and basal insulin profiles.

In the baseline scenario, UniBE achieved a median [interquartile ranges] time in range of 97.3 [90.8 99.1]%, with time below range at 0 [0 0.3]% and time above range at 2.4 [0.9 9.2]%. Across perturbation scenarios, median time in range remained between 81.7%-95.0%, with low hypoglycemia incidence, demonstrating resilience to clinically relevant disturbances.

The UniBE hAID system maintained stable glycemic regulation across diverse error conditions, highlighting its robustness and adaptability. These findings establish proof-of-concept feasibility and provide a foundation for further refinement, although these do not constitute a full pre-clinical validation. Future work must extend beyond in-silico testing to include more preclinical and clinical studies to rigorously assess safety, efficacy, and real-world performance.