An Offset-Free Data-Enabled Predictive Control Approach to Closed-Loop Glucose Management for Subjects With Type 1 Diabetes

Recent advances in hybrid automated insulin delivery (AID) systems have demonstrated promising clinical potential for blood glucose regulation of patients with type 1 diabetes. However, the performance of the closed-loop glucose control algorithms in AID systems is still challenged by the substantial inter-subject variability. In this work, we propose a personalized controller based on an offset-free data-enabled predictive control (DeePC) method. Specifically, different Hankel matrices based on the subject-specific insulin-glucose trajectories for typical glycemic behavioral scenarios are constructed and scheduled based on the current glucose fluctuations to depict the highly nonlinear behaviors of the glucose dynamic system. An incremental form of insulin inputs is adopted in the formulation of DeePC to mitigate the tracking offsets caused by the mismatch in the basal rate profile or the behavioral system representation. Besides, based on the estimated individualized insulin sensitivity, an adaptive penalty mechanism is designed, which utilizes an insulin-sensitivity-dependent input penalty term in the optimization problem to adjust the aggressiveness of the control algorithm in real-time to adapt to the variation of the insulin requirements. The effectiveness of the proposed method is evaluated by the designed comprehensive in silico experiments from the FDA-accepted UVa/Padova T1DM simulator. For the scenario of under-estimated basal rate, the proposed algorithm obtains statistically significant improvement in terms of the percent time in euglycemia 70-180 mg/dL (73.4% vs. 64.8%) and the mean glucose (150.4 mg/dL vs. 166.4 mg/dL) in comparison with the model-based controller. Advisory-mode analysis using clinical data indicates that the proposed controller can mitigate persistent hyperglycemia by recommending additional insulin delivery dosages compared to those administered by the clinician. Note to Practitioners - This work is motivated by the critical need for effective blood glucose control, which is essential for diabetes patients. The closed-loop glucose control algorithm in AID systems can help individuals sustain long-term glycemic stability to mitigate the risks associated with hypoglycemia or hyperglycemia. However, for the state-of-the-art closed-loop glucose control methods, substantial inter-subject variability, such as patient-specific physiological parameters or the different insulin responses caused by the individualized glucose metabolism, is the main cause of performance degeneration. To this end, we propose a personalized controller to achieve the offset-free tracking of the targeted glucose level of patients. Specifically, without identifying a specific parametric system model, a data-enabled predictive control framework is designed for closed-loop glucose management, which utilizes the patient-specific insulin-glucose trajectories data to learn the individualized glycemic behavior system representation and generate the personalized control decision. Besides, the real-time estimated insulin sensitivity information is incorporated in the optimization problem of the control framework, so as to enable the decision behavior to quickly respond to the individualized metabolic variations. The in silico experiments and clinical data advisor-mode analysis demonstrate the effectiveness of the proposed algorithm and its feasibility in clinical application for closed-loop blood glucose management.