Cardiac Arrhythmia Classification Using Convolutional Neural Network

The heart, vital for sustaining life, relies on proper rhythm to function effectively. Arrhythmia, a condition disrupting this rhythm, is a significant medical concern. Using Electrocardiogram (ECG) data, medical professionals diagnose arrhythmias by analyzing the heart's electrical signals. ECGs are widely used in clinical practice to assess cardiac health. This study aims to utilize deep learning techniques specifically Convolutional Neural Network (CNN), on an ECG dataset categorized into multiple classes including Right Bundle Branch Block Beat, Ventricular Escape Beat, Supraventricular Premature Beat, Normal Beat, and others. Automated arrhythmia diagnosis has substantial implications in cardiology and emergency medicine, enabling swift identification by healthcare providers and potentially saving lives. The model trained, validated and tested using ECG signals from the MIT-BIH database. To assess model performance, the confusion matrix used to compute accuracy, precision, recall and F1-score. Experimental results demonstrate outstanding performance, with accuracies of 99.67% and 99.71% in the validation and test sets, respectively. This indicates the model's potential as automated tool for arrhythmia diagnosis.