Implementación de un Sensor Virtual para Estimación de Parámetros y Estados de un Modelo de Diabetes Tipo 1

Abstract

Currently, the most conventional and widely accepted way to know the status of glucose metabolism is the capillary measurement of glucose concentration in the blood. Glucose metabolism is a highly complex process of the human body dealing with the acquisition, usage, and storage of chemical energy, mainly coming from dietary carbohydrates. To achieve this important task in the preservation of life, metabolism requires the interaction of many organs and processes of the human body; therefore, having a single systemic measurement, while providing a global vision of metabolism, does not provide details of the state of relevant processes in glucose regulation such as liver and pancreas functions. To deal with the technological problem of obtaining information on the metabolism, in this paper we propose a virtual sensor to estimate glucose and insulin concentration in relevant organs, as well as parameters that could be useful to characterize the dynamical response of the glucose metabolism. The structure of the virtual sensor is an adaptive observer based on a mathematical model of glucose metabolism in type 1 diabetes. In addition to its usefulness in estimating the state of metabolism in relevant organs, this information may be useful in the treatment of type 1 diabetes, where automatic and real-time insulin dosing requires information on metabolism, beyond just systemic blood measurement.