Enhancing Cardiovascular Health Diagnosis: A Deep Learning Approach Utilizing ECG Signals for Accurate Heart Attack Detection and Arrhythmia Identification

Abstract

The World Health Organization reports that approximately 17 million individuals succumb to cardiovascular disorders, a consequence of detrimental lifestyle choices such as excessive alcohol and tobacco con- sumption, obesity, heightened stress levels, and alterations in dietary habits. These factors present formidable challenges for timely heart fail- ure diagnosis, complicating surgical interventions. A heart attack tran- spires when the blood flow supplying oxygen to the heart muscle is severely restricted or completely obstructed. The Electrocardiogram (ECG) serves as a precise, reliable, swift, and uncomplicated method for detect- ing heart attacks, measuring heart rates. An ECG signal encapsulates

a single heart cycle, with the QRS complex being a pivotal component. Utilizing the MIT-BIH Arrhythmia Dataset, which encompasses heart- beat signals, this study employs a 1-D Convolutional Neural Network model achieving an impressive 95% accuracy. The discernment of heart rate facilitates the identification of various arrhythmias, demonstrating the potential of deep neural networks in leveraging extensive datasets for improved diagnostic capabilities