Exercise-induced hypoglycemia in type 1 diabetes: in-silico comparison between announced and unannounced strategies in closed-loop control

Abstract

Abstract It is recommended that those with Type 1 Diabetes Mellitus participate in physical activity However, insulin management during and after exercise remains challenging for both patients and physicians. The artificial pancreas is a closed-loop system which aims to control blood glucose levels automatically however, with physical activity the system is unable achieve optimal glycemic control. In this work, two different strategies of announced and unannounced exercise are combined with a closed-loop algorithm to mitigate the risk of exercise-induced hypoglycemia caused by aerobic activity. In the announced strategy, patients must inform the system that they are planning to exercise, this allows the modification of various controller parameters and the suggestion of a controller calculated amount of carbohydrates before the onset of exercise. The unannounced strategy removes the need of exercise announcement and utilizes an algorithm to detect exercise and trigger a carbohydrate suggestion and changes to the controller parameters. Both strategies are evaluated in-silico using the UVa/Padova simulator, in a realistic scenario considering sessions of aerobic exercise with 60% of VO2max for a duration of 50 minutes. Results show that both strategies are able to reduce the occurrence of hypoglycemia events, with the unannounced strategy achieving better outcomes.