Comparison of the performance of correlation waveform analysis in controlled environments

Maximizing the efficiency of antitachycardia devices (ATDs) requires minimizing their computational complexity. Reduction of the sampling rate of intracardiac signals favorably impacts ATD power and storage constraints as long as accurate arrhythmia detection is maintained. This study determined the upper and lower limits of acceptable bandpass filtering for an established morphometric method of intraventricular arrhythmia classification: correlation waveform analysis (CWA). CWA accurately discriminated sinus rhythm from ventricular tachycardia and ventricular fibrillation on an intra-patient but not an inter-patient basis. The optimum filter settings were 1-50 Hz for unipolar recorded data and 10-50 Hz for bipolar recorded data sampled at 120 Hz. In conclusion, morphometric analysis remains feasible with narrow bandpass data. Reduction in signal sampling rates with concomitant decreases in ATD computation and data storage is possible without adversely affecting the accuracy of intraventricular arrhythmia classification