Electrocardiographic Signal Processing for Automated Diagnosis of Cardiovascular Disorders

Abstract

Cardio vascular diseases (CVDs) are the leading cause of sudden cardiac death in the newlineworld. Thus, specialized medical services such as diagnostic tools for the study and treatment are in great demand. The electrocardiogram (ECG), which is frequently used to diagnose car-diac diseases, offers a crucial clinical insight. Computer-based algorithms have been created in this thesis study for automated heart health prediction and detection of different CVDs. newlineThis study shows effective methods created for ECG signal filtering. Additionally, methods for the rapid and precise identification of several cardiac diseases which can be life-threatening newlinehazard have been created. The ECG is a signal that contains information about the heart s elec-trical activity. ECG signal is corrupted by different kinds of noises like Powerline interference (PLI) and baseline wander (BW) which obscure the quality of the signal, which may mislead the clinical diagnosis. As a result, eliminating these noises from the ECG signal is critical. newlineTo reduce CVD-related mortality, it is essential to identify life-threatening arrhythmias as early as possible. This thesis investigation focuses on filtering of ECG signal, identification of ventricular fibrillation (Vfib), ventricular flutter (Vfl), ventricular tachycardia (Vta), atrial newlinefibrillation (Afib), myocardial infarction (MI) and COVID-19. Sudden cardiovascular arrest is an unexpected loss of heart function that need immediate medical attention. newlineFive objectives were developed for the current thesis work based on the identified research gaps in the surveyed literatures: First objective is to perform ECG signal filtering for elimi-nating BW and PLI noises. Second objective is detection of ventricular arrhythmias. Third objective is detection of atrial and ventricular disorders. Fourth objective is efficient detection newlineof myocardial infarction. Fifth objective is to categorize COVID-19 ECG images in presence newlineof other cardiovascular disorders. newlineIn the first objective investigation, we used a newly de