DconvNET Based Classification of ECG for heart attack detection: A Survey

Abstract

The categorization of electrocardiogram (ECG) signals holds significant importance in the diagnosis of cardiac conditions. It's difficult to classify an ECG accurately. An overview of the categorization of ECG arrhythmias is presented in this work. Selecting the right course of treatment for a patient and identifying cardiac conditions depend on the timely and precise identification of various arrhythmia types. For the classification of ECG data, several classifiers are available. Artificial neural networks (ANNs) are the most extensively utilized and well-liked classifiers among all of them when it comes to ECG categorization. In order to address the challenges raised by ECG classification, this work gives a thorough assessment of preprocessing methods, ECG databases, feature extraction methods, ANN-based classifiers, and performance measurements. Additionally, our research provides a thorough analysis of the classifiers' output and input beat selection for each surveyed paper.