INTRACARDIAC SIGNAL CLASSIFICATION FOR SVT ARRHYTHMIA DETECTION

Abstract

Electrophysiology study (EPS) has been serving as a diagnostic and curative tool for rhythm-related cardiac diseases for more than few decades now. Clinical EPS requires intense knowledge of numerous procedures and protocols and run-time analyses of various parameters of the intracardiac signals in order to find the rootcause of the problem. The Intracardiac Electrograms available during EPS are investigated for abnormality in cardiac activations. The specific manner in which the abnormal activations occur gives hint of a particular disease. This crucial diagnostic procedure involves manual calculations and thus requires a high level of proficiency and capability. In this thesis, an attempt has been made to facilitate the doctors by reducing the manual working required for diagnosis through development of an automated arrhythmia detection algorithm. The proposed algorithm will assist the beginners and support the decisions of the experts of EPS. The algorithm is designed to detect AtrioVentricular Reentrant Tachycardia (AVRT), a type of Supraventricular Tachycardia (SVT). After studying the underlying medical mechanisms, exploring the electrograms formed by reentrant circuit and analyzing temporal progression of cardiac activations during AVRT, a novel technique exploiting the above relationships is presented.