A Cloud Application for ECG Arrhythmia Classification Using Deep Learning and N-Square Approaches

Abstract

According to the World Health organization, coronary heart disease (sometimes called coronary artery disease) is the biggest cause of mortality worldwide (WHO). Around 17.7 million individuals died from Cardiovascular disease (CVDs) every year, and roughly 31% occurred in low and middle-income nations around the world. Arrhythmia is a variety of cardiovascular diseases that interrupts the heart’s normal rhythms. Some common types of irregular heartbeats include: There is no immediate danger from a single arrhythmia, but prolonged arrhythmia abnormalities can be life-threatening. Consuming these drugs increases the risk of developing heart disease. To treat and prevent cardiovascular disease, regular cardiac monitoring is essential. The heart’s rhythm and health can be visualized by the non-invasive device. Utilizing a deep two-dimensional convolution neural network, this sorting mechanism successfully categorizes electrocardiograms into the following five classes: fibrillatory, supraventricular, ventricular, ventricular, intraventricular, supraventricular, ventricular, and ventricular above, and finally unknown beats. In this piece, we try to extract the class pattern from an electrocardiogram (ECG) image using the N-Square technique and compressed image data.