A combined method for the detection of ventricular fibrillation

Abstract

A combined algorithm is proposed for the detection of ventricular fibrillation and its accurate discrimination from other rhythms. A total of 95 sinus rhythms, 15 monomorphic ventricular tachycardias, 33 atrial fibrillations, 39 supraventricular tachycardias and 30 ventricular fibrillations were analyzed. The algorithm tests three characteristics of the ventricular fibrillation through the estimation of the isoelectric potential, the power spectrum and the coherence spectrum. The algorithm allowed a reliable distinction of ventricular fibrillation from the other rhythms. The aim of this algorithm is to be used in any external device processing surface ECG signals.