Detection of Arrhythmia for Cardiac Functional Recovery Using Electrocardiogram

Cardiac arrhythmia refers to irregular heartbeat or a disorder in the rhythm or rate of heartbeat. If left untreated, it can lead to fatal heart attack, stroke, or cardiac arrest. However, if detected early and treated on the most urgent basis, the affected person can achieve a better quality of life by undertaking proper rehabilitative aid. For timely detection and prevention, Electrocardiogram (ECG) can be reliably used as a device to monitor the functionality of the cardiovascular system, and it has proved its efficiency in various existing frameworks. In this paper, 11 time-domain features are extracted after preprocessing the raw ECG signals of the publicly available dataset MIT-BIH Arrhythmia using the Chebyshev filter. Based on the obtained feature space, classification has been performed with the help of a recurrent neural network to detect the current state of the heartbeat in five categories, i.e., normal beat, beat signifying left bundle block, right bundle block, ventricular premature beat, and atrial premature contraction. The classification result has been compared with the other competitive algorithms in the existing literature. Statistical testing has also validated the performance to check the proposed model’s applicability in a real-world scenario.