Glucose-Insulin Pharmacodynamic Surface Modeling Comparison

Abstract

Metabolic system modeling for use in glycaemic control is increasing in importance. Few models are clinically validated for both fit and prediction ability. For such models, this research introduces a new form of pharmaco-dynamic (PD) surface comparison for model validation. These 3D surfaces are developed for 3 validated models, including the well-known Minimal Model, and fit to clinical data. The approach is clearly highlights differences in modeling methods, dynamics utilized and physiological assumptions that may not appear as clearly in other validation approaches. The deficiencies of the Minimal Model in comparison to more physiologically representative models are illustrated in this context.