An ECG Based CNN Model for Detection of Different Classes of Arrhythmia

Electrocardiograms (ECGs) use electrodes to monitor heart rates and discover tiny electrical changes in each beating. This test can detect abnormal cardiac activity such as arrhythmias and conduction abnormalities. The goal of this research is to offer a method for analyzing and classifying the six micro-classes of cardiac variations found in the MIT-BIH arrhythmia database. Normal (NOR), Left Bundle Branch Block (LBBB), Right Bundle Branch Block (RBBB), Atrial Premature (AP), Premature Ventricular Contraction (PVC), and others (/) are the classifications. The wavelet transform is used to pre-process the input signals, and the peak detection technique is used to identify the R peaks. After preprocessing, oversampling is conducted on the underrepresented classes to improve feature learning by the Convolutional Neural Network(CNN).CNN has 15 levels that are used to process the input. The model achieves a 99.42 accuracy rate.