Power spectrum analysis of endocardial electrograms during automatic defibrillator implantation

Implantable cardioverter/defibrillator devices use different algorithms to recognize ventricular tachycardia and ventricular fibrillation from sinus rhythm. However problems remain, especially differentiating physiologic high rhythms from life threatening situations. In order to test detection abilities of different methods, off-line analysis was performed in time (rate counting, autocorrelation function, statistical methods) and in frequency (digital fast Fourier transform) domains on bipolar electrograms that were obtained in 28 patients during the implantation of an automatic defibrillator. Tachycardia and fibrillation were induced and recordings made of the ECG, right ventricular electrogram and aortic pressure. Results showed that even simple methods perform equally well compared to calculation intensive methods