Towards An Intelligent Blockchain-Enabled IoMT Framework for Telemedicine

The telemedicine industry faces significant challenges in maintaining data integrity and traceability. This paper examines the integration of Blockchain and Machine Learning (ML) technologies to develop a secure and efficient Internet of Medical Things (IoMT)-based telemedicine system designed to predict potential health risks. The system features two primary mechanisms implemented through lightweight smart contracts: an identity management mechanism to control patient identification and authorization to access the network and a health prediction mechanism to securely store patient health data. The health prediction mechanism is interfaced with an off-chain ML module, which uses oracles to retrieve real-time data, enabling the assessment of potential health risks. The ML prediction module uses one of the ensemble techniques: Bagging, LightGBM, and Stacking, to predict particular medical problem, cardiac arrest. We investigate the performance of these three techniques for the prediction of cardiac arrest. Simulation results indicate that the LightGBM and Stacking models are particularly effective in minimizing false alarms, especially in older patients.