Verification of a Method of Detecting Life-threatening Arrhythmias from Human Data for Use in Implantable Cardioverter-Defibrillator

Abstract

The implantable cardioverter-defibrillator is an effective therapeutic device for preventing death from life-threatening arrhythmias such as ventricular tachycardia and ventricular fibrillation in patients with cardiac diseases. It is important to prevent their recurrence and treat these arrhythmias as early as possible. It is also essential to accurately distinguish among normal sinus rhythm, ventricular tachycardia, ventricular fibrillation, and supraventricular tachycardia. Therefore, in this study, we proposed a classification method that uses multiple regression model based on information extracted from simultaneous intracardiac electrocardiograms, to identify episodes of sinus rhythm, supraventricular tachycardia, and ventricular tachycardia from human intracardiac electrocardiograms. From the results of validation, we confirmed that the proposed method is able to distinguish shockable cardiac rhythms from nonshockable cardiac rhythms based on indices obtained from simultaneous intracardiac electrocardiograms.