Model-Based Personalization Scheme of an Artificial Pancreas for Type I Diabetes Applications

An internal model-based proportional-integral-derivative (IMC-PID) controller adapted for an artificial pancreas (AP) system which controller requires only a single tuning parameter to modify controller performance and comprises a discrete control-relevant model with a priori subject information as design criteria, wherein a subject's basal insulin injection characteristic is incorporated, in addition to the total daily insulin (TDI) clinical parameter to further attune the controller's action in cases of model and patient mismatch, and an insulin feedback scheme (IFB), which accelerates the apparent insulin pharmacokinetic profile.

A controller for an artificial pancreas (AP) system adapted to maintain blood glucose concentrations of people with type 1 diabetes mellitus within the euglycemic zone (80-140 mg/dl) comprising a control algorithm, and which automatically directs the delivery of insulin to be administered to a subject with type 1 diabetes, comprising an internal model control (IMC) based design of a proportional-integral-derivative (PID) controller or a model predictive control (MPC) feedback controller, wherein the controller is configured with individual gain personalization based on each subject's clinical characteristics, using a control-relevant discrete model with a priori subject information, wherein the subject's basal insulin is incorporated into the lower order model to further personalize the controller's aggressiveness and take into account variations in insulin sensitivity, wherein the personalization allows the controller to be aggressive in cases where the subject is insensitive to insulin and requires a large basal amount to maintain euglycemia, while preventing hypoglycemic risk by reducing the controller's aggressiveness in cases where the subject requires a low basal amount to maintain normal blood glucose levels.