The Investigation of Machine Learning and Deep Learning Classification of Internet of Things (IoT) Enabled Medical Devices

Abstract

The present cross-sectional, quantitative study investigates the classification capabilities of ML and DL in processing data from IoT-enabled medical devices: heart rate variability (HRV), blood pressure, blood glucose levels, and physical activity in patients with diabetes and cardiovascular diseases. Eight patients from AIIMS, New Delhi participated in this study. In the current investigation, two types of measurements are taken into consideration: cardiovascular measures and diabetes management metrics, which include BMI, blood pressure, blood glucose, physical activity, and HRV. The outcomes showed a positive association with blood glucose level and age (0.55), negative correlation between blood glucose and physical activity (-0.68), which had significant significance towards exercising importance in diabetes management. Blood glucose correlated moderately favorably with BMI with 0.45, while age and BMI correlated at 0.32. These results show how ML and DL techniques can improve the interpretation of complex health parameters from IoT-enabled devices to better manage patients and develop specific healthcare plans in chronic diseases.