Unannounced meal analysis of the ARG algorithm

EMILIA FUSHIMI; PATRICIO COLMEGNA; HERNÁN DE BATTISTA; FABRICIO GARELLI; RICARDO SANCHEZ PEÑA

Conferencia; American Control Conference (ACC); 2019

IEEE

One of the main challenges in automatic glycemicregulation in patients with type 1 diabetes (T1D) is to dispensewith carbohydrate counting. In this context, we propose toequip a previously introduced switched Linear Quadratic Gaussian(LQG) controller?the so-called Automatic Regulation ofGlucose (ARG) algorithm?with an automatic switching signalgenerator (SSG). The ARG algorithm not only regulates thebasal insulin infusion rate but also generates feedback insulinspikes at meal times, i.e., no open-loop insulin boluses areneeded to mitigate postprandial glucose excursions. However, inits former version, it was required to announce the meal time.In this work, the performance of the ARG algorithm combinedwith the proposed SSG is assessed in silico with unannouncedmeals. In addition, the response of the SSG is estimated usingclinical data obtained with the ARG algorithm from the firsteverartificial pancreas (AP) trials carried out in Latin America.