A Dual-Hormone Multicontroller for Artificial Pancreas Systems

(Hybrid) Closed-Loop Systems: From Announced to Unannounced Exercise

Physical activity and exercise have many beneficial effects on general and type 1 diabetes (T1D) specific health and are recommended for individuals with T1D. Despite these health benefits, many people with T1D still avoid exercise since glycemic management during physical activity poses substantial glycemic and psychological challenges - which hold particularly true for unannounced exercise when using an AID system. Automated insulin delivery (AID) systems have demonstrated their efficacy in improving overall glycemia and in managing announced exercise in numerous studies. They are proven to increase time in range (70-180 mg/dL) and can especially counteract nocturnal hypoglycemia, even when evening exercise was performed. AID-systems consist of a pump administering insulin as well as a CGM sensor (plus transmitter), both communicating with a control algorithm integrated into a device (insulin pump, mobile phone/smart watch). Nevertheless, without manual pre-exercise adaptions, these systems still face a significant challenge around physical activity. Automatically adapting to the rapidly changing insulin requirements during unannounced exercise and physical activity is still the Achilles' heel of current AID systems. There is an urgent need for improving current AID-systems to safely and automatically maintain glucose management without causing derailments - so that going forward, exercise announcements will not be necessary in the future. Therefore, this narrative literature review aimed to discuss technological strategies to how current AID-systems can be improved in the future and become more proficient in overcoming the hurdle of unannounced exercise. For this purpose, the current state-of-the-art therapy recommendations for AID and exercise as well as novel research approaches are presented along with potential future solutions - in order to rectify their deficiencies in the endeavor to achieve fully automated AID-systems even around unannounced exercise.

InsNET: Accurate Basal and Bolus Insulin Dose Prediction for Closed Loop Diabetes Management

It has been demonstrated that closed-loop diabetes management results in better glycemic control and greater compliance than open-loop diabetes management. Deep learning models have been used to implement different components of artifical pancreas. In this work, a novel deep learning model InsNET has been proposed to estimate the basal and bolus insulin level and insulin bolus in patients with type I diabetes utilizing subcutaneous insulin infusion pumps for closed loop diabetes management system. The proposed InsNET is formed with a Wide-Deep combination of LSTM and GRU layers. Additionally, physical activity level has been included as an input in comparison to previous models where only past glucose levels (CGM), meal intake (CHO) and past insulin dosage were used as inputs. The proposed model was tested on In-silico data, and it achieved a Mean Absolute Error (MAE) of 0.002 and Root Mean Squared Error (RMSE) of 0.007 for UVA/Padova Dataset and MAE of 0.001 and RMSE OF 0.003 for mGIPsim Dataset.Clinical relevance- Insulin dose determination is an important as aspect of artificial pancreas. This work describes a deep learning model to determine accurate basal and bolus insulin dosage.