An IOT framework for detecting cardiac arrhythmias in real-time using deep learning resnet model

A cardiac arrhythmia poses a serious health risk to patients and can have serious consequences for their health. A clinical assessment of arrhythmia disorders could save a person's life. The Internet of Things (IoT) will revolutionize the healthcare sector by continuously monitoring cardiac arrhythmia diseases remotely and minimally invasively. We propose a frame-work that will facilitate the development of a practical diagnostic tool for the identification of cardiac arrhythmias in real-time in this work. An Electrocardiogram (ECG) signal is processed using the Pan Tompkins QRS (Quantum Resonance System) detection method in order to extract the dynamic properties of the signal. The inter beat (RR) intervals are derived from an ECG signal in order to determine the characteristics of heart rate variability. The electrocardiogram is primarily used to identify irregular heartbeats (cardiac arrhythmias). Therefore, in our study, we evaluated other factors such as the heartbeat of the individual. As part of our IoT deployment, we are storing and analyzing data collected by the Pulse Sensor on the ThingSpeak IoT platform. The designed circuit's real-time collection of heartbeat and beats per minute values was uploaded to Thingspeak. Over the course of more than a week, we collected a variety of heart data. We propose Multi Channel Residual Network (MCHResNet) a deep-learning based solution that relies on multi-channel convolutions to detect both spatial and frequency features from electrocardiograms to facilitate the classification process. Based on the well-known Massachusetts Institute of Technology-Beth Israel Hospital Arrhythmia (MIT-BIH-AR) database, we evaluate the proposed framework against MCH ResNet. Our IoT-based framework has been shown to be effective based on the results reported in this paper.