Automated System Classification of ECG Heartbeat based on Support Vector Machine and Convolutional Neural Network

Detecting cardiac arrhythmias is a significant public health concern, but manual identification and classification of arrhythmias using electrocardiogram (ECG) signals is a challenging task. In this paper, we proposed an automatic arrhythmia classification system using Machine Learning and Deep Learning algorithms. The proposed pipeline used a wavelet transform technique to eliminate signal noise. Morphological and statistical features were extracted to classify ECG signal beats into five classes: normal,ventricular, supraven-tricular, fusion, and paced beats. The system was evaluated on the MIT-BIH arrhythmia database using two classification algorithms, SVM and CNN, achieving accuracy rates of 99.34% and 99.72%, respectively.