A Novel Index to Evaluate the Blood Glucose Controllability of Type I Diabetic Patients

CAMPETELLI, G.; MUSULIN, E.; BASUALDO, M.; RIGALLI, A.

Proceedings of the 18th IFAC World Congress

IFAC

Lugar: Milano; Año: 2011; p. 1423 - 1428

The development of the Artificial Pancreas has been growing rapidly during the last 10 years with the advent of Glucose Monitoring Systems and insulin delivery pumps. Several works have appeared testing different control algorithms for defining the proper insulin dosage in type I diabetic patients, focused on driving them into a healthy range of blood glucose (BG) concentration. However, practically none of them has previously considered to quantify the inherent degree of difficulty in achieving normal glycaemic values. Hence, in this work a novel index, named "Glucose Controllability" (GC), is presented using the parameters of simplified models and easily measurable patients data. It allows performing a certain classification procedure before the implementation of any control algorithm. To test this methodology, Predictive Functional Control (PFC) is applied for insulin dosage computation. After that, through the well-known Control Variability Grid Analysis (CVGA) plot, it is evaluated if the GC index is able to anticipate how "controllable" each diabetic person could be. The validation was performed with a database of 30 virtual type I diabetic patients. The final simulation results are shown here together with the conclusions and future works.