System Development of ECG Arrhythmia Signal Classification using Deep learning YOLO Network approach

Abstract

Cardiovascular diseases (CVDs) are often caused by arrhythmia, newlineand severe arrhythmiaresult in abrupt death or heart failure. Therefore, it newlineis vital and essential to identify arrhythmias accurately and promptly. newlineElectrocardiograms (ECGs) are one-dimensional physiological signals newlinethat indicate the condition of the heart. Since the ECG signal is a reflection newlineof the electrical signals of the heart, cardiac rhythm abnormalities or newlinechanges in the ECG waveform can efficiently detect the symptoms of newlinearrhythmias.Therefore the ECG monitoring equipment is paramount newlineimportant in arrhythmia diagnosis. There are generally two types of ECG newlinemonitoring equipment. The first one are those used in used in hospitals newlinewhich are large in size, fixed in position and are mainly used to identify newlinethe congenital heart problems. The second category is the smaller, portable newlineones which are suitable for self-monitoring, mainly used by sports and newlineelderly people. This work deals with the development of small size and newlineportable ECG monitoring equipment. newlineThe primary requirement of portable ECG monitoring equipment newlineincludes accurate real-time detection algorithms and efficient method for newlinepower supply.In the recent days, Deep learning (DL), which is a computeraided newlinerecognition technology is gaining popularity in arrhythmia newlineix newlinediagnosis. This technologyis found to be more accurate and involves fast newlineprocessing time with minimum human usage. Alternative energy solutions newlinehave the potential to redirect energy consumption towards sources with newlinereduced carbon emissions and pollution, thereby offering greater energy newlinediversity. In the proposed work, to power the ECG machines, three types newlineof powering units are usedthat includes two Renewable Energy Sources newlinesuch as Photovoltaic (PV), biogas fuel cells and battery system, as they newlineare the popular methods used for powering portable electronic devices. newlineA new DL-based You Look Only Once(YOLO)-ECG model newlineis proposed as a detection technique aimed at ECG arrhythmias has been newlinedeveloped in this work.