JMIR Publications

ABSTRACT

Over the past few decades, diabetes has become a serious public health concern worldwide, particularly in Bangladesh. This study aimed to increase the predictability of blood glucose levels in Bangladesh using machine learning (ML) techniques and a cutting-edge web application. Based on basic non-invasive health checkup test results, dietary information, and socio-demographic characteristics, five well-known ML models—Linear regression, Boosted Decision Tree Regression, Neural Network, Decision Forest Regression, and Bayesian Linear Regression—are used to predict blood glucose. Continuous blood glucose data were used in this study to train the model, which then used the trained data to predict new blood glucose values. Data were collected from 271 employees of the Bangladeshi Grameen Bank complex. Boosted Decision Tree Regression demonstrated the greatest predictive performance of all evaluated models (Root Mean Squared Error, RMSE = 2.30). The next step is to create a web application for blood glucose prediction. The online application is a simple, practical, and effective option for users. This application helps the users to monitor their own health, especially in remote areas of underdeveloped countries where there are not enough qualified doctors and nurses. The results of this study can assist save money on medical expenses, time, and health management expenses. The created system also aids in achieving the Sustainable Development Goals (SDGs), particularly the third SDG that aims to ensure that everyone in the community enjoys good health and wellbeing and lowering total morbidity and mortality in Bangladesh and other countries with similar settings.