Diagnosis and Classification of Cardiac Arrhythmias Using Convolutional Neural Networks

Cardiac arrhythmia stands as a prominent contributor to cardiovascular diseases, resulting in significant global mortality rates. The methods for diagnosing diseases, including cardiovascular conditions, through the processing of biomedical signals, continuously advance and develop over time. The progress in signal processing techniques has enabled the extraction of various types of information from Electrocardiography (ECG) signals. The manual analysis of ECG signals is very difficult because of its complexity. However, the use of artificial intelligence algorithms simplifies this process. In this study, we used a convolutional neural network (CNN) algorithm that mainly facilitates the categorization of the type of arrhythmia and reduces the time of this process. The dataset used in this study is MIT-BIH arrhythmia which is widely popular among scientists from this field. 4 classes were chosen in this classification: N, SVEB, VEB, and F. As a result, the overall classification accuracy, sensitivity, and specificity of the proposed approach are 99.24%, 99.24%, and 96.94% respectively. The obtained outcomes are being evaluated against the highest-level or pioneering norms within the discipline.