Effective control of glycemia using a simple discrete-delay model

Type-2 Diabetes Mellitus (T2DM) is a metabolic syndrome characterized by low insulin sensitivity, so that higher amounts of insulin are required in order to keep glycemia in a safe range (approximately, 60~110 mg/dl or, equivalently, 3.33~6.11 mM). Although insulin resistance and T2DM are often treated without exogenous insulin administration, the possibility to early treat pre-diabetic states or T2DM patients with insulin administration could be envisaged if the clinical need exists (e.g. surgical stress, infection). The present work introduces a possible new therapeutic insulin administration dosing approach for T2DM patients. The IVGTT glycemia and insulinemia profiles of a diseased patient are collected and, successively, its metabolic parameters are obtained by fitting a compact delay model to those data. Then a controller is designed exploiting the previous model as a tool. Finally, the tuned controller is applied to the patient as an artificial pancreas supplying external insulin administration. The results are shown on a virtual patient, whose behavior is described by a comprehensive, validated extensive model.