Volume 12, Issue 11

ECG Signal Classification with Convolutional Neural Networks

Author

SABIR MUHAMMAD*1, ARSLAN2, IMRAN KHAN3

Abstract

Cardiovascular waves in electrocardiograms (ECG) give significant data on heart conditions and the impacts of heart drugs. Division and analysis of
ECG and its constituent cardiovascular waves are of high importance in cardiology determination and drug studies. Customarily, thoroughly prepared clinicians and cardiologists have performed ECG investigations. In any case, using clinicians in enormous scope ECG screening, for example, drug test stages or populace-based screening programs is not monetarily achievable. Consequently, an automated ECG division approach that can fragment heart waves accurately is of high significance. This thesis concentrates on Deep Learning (DL) based automated ECG division and outline techniques. Because of different shapes and irregularities in the ECG signal, the traditional straightforward component channels neglect to remove the assortment of cardiovascular wave arrangements. Convolutional Neural Networks (CNN) apply multi-facet including channels on the input to remove complex features from the sign. In this way, CNN can be used to remove various levelled highlights from ECG signals. Two Convnet structures are read up and utilized for the restriction of cardiovascular
waves. Their exhibitions are contrasted with one another and different investigations in the writing too. The outcome shows that CNN is capable of extracting heart wave spatial elements with the great execution. Moreover, various long haul notwithstanding transient worldly patterns exists in ECG signals because of arrhythmia and other heart conditions. Generally speaking, momentary memory will be unable to catch the transient highlights in ECG signals, thus a Long Short-Term Memory (LSTM) network is planned and utilized to catch long haul and momentary information dependencies in ECG sequences. This technique has worked on distinguishing proof of worldly highlights, especially in P-wave. To address the crude ECG signal division issue, a hybrid DL compressing of
convolutional auto encoders and LSTM networks is proposed and contemplated. Results show that it performs commonly well for crude ECG signal division.

DOI

https://doi.org/10.62226/ijarst20231117

PAGES : 1182 - 1196 | 8 VIEWS | 5 DOWNLOADS

Download Full Article