Researchers at Illinois Tech are developing the algorithms for a multivariable artificial pancreas (AP) that uses additional information from wristbands to improve insulin administrations decisions for people with diabetes.These multivariable AP algorithms will use information extracted from wearable devices (heart rate, energy expenditure, movement) to mitigate better the effects of exercise, stress and sleep on glucose levels.
Once completed, the artificial pancreas will consist of a glucose sensor, a wristband, a dedicated smartphone and an insulin pump. It will be fully automated and will not expect any manual inputs from the user such as meal amounts or indication of an exercise session.
The research is funded by the National Institutes of Health and the Juvenile Diabetes Research Foundation.
JDRF is at a renaissance time with accelerated advances in research which must continue. We’re on the cusp of some of our most promising breakthroughs, and we thought speaking to someone in the Illinois community whose research and work is directly impacted by this would bring light to conversation.
For any questions, contact JDRF Illinois Chapter at illinois@jdrf.org
Artificial pancreas (AP) holds promise for improving blood glucose homeostasis in patients with Type 1 diabetes (T1D). Several control techniques have been proposed for developing the control algorithms of an AP. The presentation will introduce various control techniques (PID, model predictive control, adaptive control, fuzzy logic) used in AP systems. Research efforts have focused on various control algorithms and estimation techniques for the carbohydrate content and timing of meals. Physical activity has a significant effect on glucose utilization and blood glucose concentrations (GC), and can cause hypoglycemia during or after exercise, with a high likelihood of promoting nocturnal hypoglycemia. A multivariable AP control algorithm that can accommodate the effects of exercise will be presented. It uses physiological data from sports armbands that are incorporated in models to improve the accuracy of glucose concentration predictions and in the control logic of the control system. Hypoglycemia is a major challenge for AP systems. Since patients with T1D do not have natural means to reduce plasma insulin levels and their glucagon response is often impaired, they are unable to prevent hypoglycemia with endogenous means. AP systems with dual hormones (insulin and glucagon) and an integrated multivariable hypoglycemia early alarm and control system for glucose regulation in patients with T1D will be discussed.