ML-Enhanced IoMT-based Electrocardiogram Monitoring System for Patient-Centric Solution

Cardiovascular health is a global concern today, with Pakistan as no exception. Traditionally, electrocardiogram (ECG) tests are very cumbersome and performed mostly at hospitals. This leaves many people, especially in remote areas, unable to get timely access due to insufficient healthcare facilities. In addition, the discomfort that usually comes with conducting conventional ECGs makes people shy away from repeat screenings. In this paper, we have presented a machine learning (ML) enabled Internet of Things (IoT) based ECG monitoring system to provide accessible, portable, and real-time heart health monitoring. This paper discusses how we have integrated ECG sensor, IoT, and ML technologies to detect anomalies within the ECG signal in real time so that low-cost healthcare provisioning in regions can be achieved. The high accuracy of the system based on the BMD101 sensor in terms of the detection of ECG anomalies shows that it can also be used as a large-scale ECG monitoring device. The work presented in this paper demonstrates comparison between the two convolution neural network models, where two convolutional layers and a filter size of 64 results in better training and testing accuracy whereas a filter size of 128 results in better anomaly accuracy. Overall, this work is a step forward to the development of Internet of Medical Things (IoMT) systems for realizing ease in patient monitoring.