Volume 11, Number 23, December 2021

Proposal for a Methodology to Create an Artificial Pancreas for the Control of Type 1 Diabetes
using Machine Learning and Automated Insulin Dosing Systems (AID)

  Authors

Rodrigo Vilanova¹ and Anderson Jefferson Cerqueira², ¹Avantsec, Brazil, ²University of Brasília (UnB), Brazil

  Abstract

The number of children, adolescents, and adults living with diabetes increases annually due to the lack of physical activity, poor diet habits, stress, and genetic factors, and there are greater numbers in low-income countries. Therefore, the aim of this article is to present a proposal for a methodology for developing a pancreas using artificial intelligence to control the required doses of insulin for a patient with type 1 diabetes (T1D), according to data received from monitoring sensors. The information collected can be used by physicians to make medication changes and improve patients’ glucose control using insulin pumps for optimum performance. Therefore, using the model proposed in this work, the patient is offered a gain in glucose control and, therefore, an improvement in quality of life, as well as in the costs related to hospitalization.

  Keywords

Machine Learning, Artificial Intelligence, Data Mining, Diabetes, T1D, AID, iCGM, Clustering, Regression, Decision-making, Time series, Cybersecurity, Methodology.