The Role Of Ecg For Identifying Cardiac Involvement In Patients Post Covid19

Abstract

Sudden Cardiac Death (SCD) in COVID is a major risk due to the alterations in Electrocardiogram (ECG) signal. SCD makes the regular monitoring of the ECG signals necessary for tracing the events of distraction. Thus the ultimate aim of the paper is to design an algorithm for in and out hospital monitoring of cardiac patients regularly. But regular monitoring will generate a large volume of data and hence increases the bandwidth requirements. In this state the work also proposing compression of signals at the transmission end. For compression, the Discrete cosine transform (DCT) combined with run-length encoding.  ECG conducts through atrial and ventricular chambers of the heart successively. The arrhythmia can be originated in any of these chambers following the COVID situation. Hence both kinds of signals are considered. The average value of Compression Ratio (CR) is calculated 6.89152 & 4.410718 and Percent root mean square difference (PRD) achieved is 0.976787 & 0.795441 for malignant ventricular arrhythmia (MVA) and Supraventricular or Atrial arrhythmia (SA) respectively. Four classification Error-correcting output code (ECOC), K-nearest neighbor (KNN), Naïve Bayes and Tree classifiers are compared and accuracy of 97%, 92.86%, 89.2%, and 89.2% respectively achieved through each. The performance has been checked on the database taken from MIT-BIH Supraventricular Arrhythmia (SA), and Malignant ventricular ectopic beats (MVA) database and that constitutes 100 signals.