D-Resnet: Deep Resnet based approach for ECG classification

The ECG signal represents the electrical activity of the heart and reflects the health of the cardiovascular system. It also contains information that can be used to differentiate cardiovascular diseases. The automatic classification of arrhythmias is an important step in the development of monitoring equipment in the ambulatory or intensive care setting. In this work we propose a deep residual network (DResnet) model which allows a very deep extraction of the characteristics of the ECG signal in order to accurately differentiate between normal and abnormal signals. In the framework of an embedded system project for elderly automated diagnosis, we propose an approach that is designed to classify six types of cardiac rhythms: normal beats, ventricular premature beats, rhythmic beats, atrial premature beats, fusion of ventricular and normal beats or noise. The characteristics and depth of the proposed model make it possible to provide satisfactory precision in comparison with the work of the literature.