Utilizing IoT-based sensors and prediction model for health-care monitoring system

Abstract

The Internet of Things (IoT)-based sensors together with smartphone can be utilized as personal health devices to gather vital signs data, so that current health condition of patient can be presented. In this study, we propose a health-care monitoring system by utilizing an IoT-based sensor device and prediction model, so that diabetes patients can better self-manage their chronic condition. The IoT-based sensors such as blood pressure monitor, weight scale, smartwatch, and glucose meter are utilized to gather the vital sign data, while the prediction model is utilized for predicting hypoglycemia and hyperglycemia events in the next 12 and 24hours. The results showed that commercial versions of the IoT-based sensors and the prediction model based on random forest are sufficiently efficient to monitor the vital signs data of diabetes patients and to predict future hypoglycemia and hyperglycemia events of patients, so that it can provide the appropriate preventive actions and avoid critical conditions in the future. © 2021 Elsevier Inc. All rights reserved.