Empowering health data protection: machine learning-enabled diabetes classification in a secure cloud-based IoT framework

Abstract

Smart medical devices and the internet of things (IoT) have enhanced healthcare systems by allowing remote monitoring of patient's health. Because of the unexpected increase in the number of diabetes patients, it is critical to regularly evaluate patients' health conditions before any significant illness occurs. As a result of transmitting a large volume of sensitive medical data, dealing with IoT data security issues remains a difficult challenge. This paper presents a secure remote diabetes monitoring (SR-DM) model that uses hybrid encryption, combining the advanced encryption standard and elliptic curve cryptography (AES-ECC), to ensure the patients' sensitive data is protected in IoT platforms based on the cloud. The health statuses of patients are determined in this model by predicting critical situations using machine learning (ML) algorithms for analyzing medical data sensed by smart health IoT devices. The results reveal that the AES-ECC approach has a significant influence on cloud-based IoT systems and the random forest (RF) classification method outperforms with a high accuracy of 91.4%. As a consequence of the outcomes obtained, the proposed model effectively establishes a secure and efficient system for remote health monitoring.