Guest lecture by Prof Ali Cinar: "Adaptive Personalized Multivariable Artificial Pancreas"
Welcome to this guest lecture 9 May held by Professor Ali Cinar (Illinois Institute of Technology). He will visit NTNU, Trondheim as an opponent for Odd Martin Staal’s PhD defence, scheduled for 10 May.
Title: Adaptive Personalized Multivariable Artificial Pancreas
Speaker: Professor Ali Cinar from Illinois Institute of Technology. He will be in Trondheim as an opponent for Odd Martin Staal’s PhD defence (10 May).
Location: D251, Elektrobygget D, NTNU Gløshaugen. Map: http://bit.ly/2XPqOFq
Time: Thursday 9 May, 13:00–14:00.
The automation of insulin infusion through closed-loop glucose control strategies is challenged by day-to-day variations in the users’ daily activities that affect the glucose-insulin metabolism of people with Type 1 diabetes. An artificial pancreas (AP) that relies only on glucose concentration information has limited effectiveness because of its reactive nature to glucose concentration changes that have already taken place. Feedforward control based on information about disturbances to glucose homeostasis such as meals and physical activities can improve the performance of APs. Feedforward control could be based on manual information entry (meal time and content, exercise) or automated by streaming data from wearable devices reporting physiological variables.
An adaptive and personalized multivariable AP (mAP) system is developed to efficiently accommodate major disturbances to glucose concentration. Accurate adaptive glycemic models are developed with a recursive subspace identification technique by using physiological measurements from a wristband along with continuous glucose concentration signals and estimates of unannounced meal effect and plasma insulin concentration (PIC) to characterize the glucose dynamics under various conditions such as food consumption and physical activity. These models with recursively updated time-varying parameters are utilized in an adaptive model predictive control (MPC) system that is cognizant of the PIC. These algorithms that interpret various signals and adapt the controller parameters and constraints, enable the mAP system to effectively compute the optimal insulin infusion over diverse variations without manual meal and exercise announcements.
Simulation studies using a multivariable glucose-insulin-physiological variable simulator (mGIPsim) demonstrate the performance of the mAP system.
Ali Cinar is the Hyosung S.R. Cho Endowed Chair professor in the Chemical and Biological Engineering Department at Illinois Institute of Technology, with dual appointment in the Biomedical Engineering Department. He also has a courtesy Research Associate (Professor) appointment in the Section of Endocrinology, Diabetes, and Metabolism at the University of Chicago. He is the founding director (2004) of the Engineering Center for Diabetes Research and Education at Illinois Tech. His research is focused on system modeling, supervision and control, funded by NIH, JDRF, NSF, FDA and industry. Current research includes development of multivariable artificial pancreas systems for automated insulin delivery to people with diabetes, detection of physical activity, stress and sleep to predict their effects on glucose concentrations, modeling immune system activities to simulate beta cell loss and onset of diabetes, modeling and control of biopharmaceutical mammalian cell reactors. He has published three books and more than 250 technical papers. His research lab at Illinois Tech has graduated 24 PhDs and currently has 8 PhD students. He received his Ph.D. degree in chemical engineering from Texas A&M University. Dr. Cinar is a Fellow of the American Institute of Chemical Engineers, a Senior Member of the Institute of Electrical and Electronics Engineers (IEEE) and a member of American Diabetes Association.