Unannounced meal analysis of the ARG algorithm

FUSHIMI, EMILIA; FABRICIO GARELLI; DE BATTISTA, HERNÁN; COLMEGNA, PATRICIO; RICARDO S. SÁNCHEZ PEÑA

Philadelphia

Conferencia; American Control Conference; 2019

One of the main challenges in automatic glycemicregulation in patients with type 1 diabetes (T1D) is to dispensewith carbohydrates counting. In this context, an automatic mealdetection algorithm based on a Kalman filter and continuousglucose monitoring (CGM) readings is proposed to be integratedto the previously introduced Automatic Regulation ofGlucose (ARG) algorithm. This closed-loop strategy not onlyregulates the basal insulin infusion rate but also generatesinsulin spikes at meal times, i.e., no open-loop insulin bolusesare needed to mitigate postprandial excursions. However, it isstill required to announce the meal time and classify the mealsize between three categories: small, medium or large. In thiswork, the performance of the proposed meal detector integratedinto the ARG algorithm is evaluated in silico and with clinicaldata obtained from the first-ever artificial pancreas (AP) trialscarried out in Latin America.