Study of ECG Signal Classification based on Deep Learning Techniques

Abstract

Cardio-vascular disease can become a major cause of death if it is ignored or not diagnosed properly. The treatment can be the significant expense due to the late diagnosis. What's more, the number of specialist working in this cardiology domain are continually diminishing. Hence, it is necessary to design an develop an automatic  algorithm  for the classification of heart-beats from electro-cardiogram (ECG) registers. This classification is tough process with many edges. Eventhough a few people worked on it to address the issues, it isn't as clear that which of them has good performance, due to the inconsistencies in the comparisons and only the researchers who applied the similar paradigms (Intrapatient, Interpatient or patient-specific) are can be compared.The waveforms have various deflections. These signals are analysed and interpreted automatically and it uses a computer assistance signal processing and pattern recognition techniques. In this paper, we have presented an hypothetical survey of deep learning and traditional methods based electro-cardiogram signals classification.