Prediction of Type-2 Diabetes Level with the Smart Sensor System

Abstract

There are a number of type-2 diabetes mellitus models available that can be used in different ways, including estimating long-term clinical outcomes, estimating the costs of clinical trials, as well as assisting in making decisions to determine which interventions are best suited for these populations. The current research aims to present an algorithm for mobile or smart sensor systems to predict the blood glucose level in type 2 diabetic patients.

Dexcom system was considered as a study tool. The efficiency of a revised Dexcom system with an advanced algorithm was assessed. The prediction of future glucose concentration is predicted by using suitable modeling approaches.

The accuracy and dependability of the Dexcom system can be considerably enhanced by using advanced denoising and calibration algorithms. The precision and consistency of the continuous glucose monitor improved over time, with the greatest improvement. According to the findings, signal-processing-induced time delays have been decreased, and performance at low plasma glucose has been enhanced. Advances in sensor systems' performance are expected as a result of these enhancements.

This application effectively maintains a stable blood glucose level because blood glucose level reaction times are delayed when glucose is consumed. Furthermore, the findings of this work can be used to improve closed-loop systems and provide information to insulin pumps.