Introduction and objectives: The aim of this study was to examine the hypothesis that the recording mode (i.e., unipolar or bipolar) affects the information obtained using spectral analysis techniques during ventricular fibrillation by carrying out an experiment using epicardial electrodes.
Methods: Recordings of ventricular fibrillation were obtained in 29 isolated rabbit hearts using a multiple-electrode probe located on the left ventricular free wall. The parameter values obtained in the frequency domain (by Fourier analysis) using unipolar or bipolar electrodes, different interelectrode distances, and different orientations (i.e., horizontal, vertical or diagonal) were compared.
Results: Changing the recording mode (i.e., unipolar to bipolar) or the interelectrode distance significantly altered the mean frequency (P< .0001) and the normalized energy of the spectrum (+/-1 Hz) around the dominant frequency (P< .05), though the changes were small relative to the dominant frequency. Cross-spectral analysis showed that the coherence between unipolar recordings decreased as the interelectrode distance increased, while the opposite occurred with the coherence between unipolar and bipolar recordings. The two coherences were inversely correlated such that the greater the former coherence, the less the coherence between unipolar and bipolar recordings (r=0.29; P< .0001; n=348).
Conclusions: The recording mode (i.e., unipolar or bipolar) used influenced the information obtained using spectral analysis techniques from epicardial recordings of ventricular fibrillation. Differences were observed in the mean frequency and in the frequency distribution, but they were very small relative to the dominant frequency.