Combination of frequency and phase to characterise the spatiotemporal behaviour of cardiac waves during persistent atrial fibrillation in humans

The spatial distribution of atrial dominant frequency (DF), phase and phase singularity points (PSs) may reflect mechanisms driving and maintaining persistent atrial fibrillation (persAF). Here we developed an automatic algorithm that combines the three parameters and depicts the complex spatiotemporal patterns of fibrillation. For 9 patients undergoing left atrial persAF ablation, noncontact virtual unipolar electrograms (VEGMs) were simultaneously collected using a balloon array (Ensite Velocity, St. Jude Medical). After removal of the far field ventricular influence, we used fast Fourier transform and Hilbert transform to detect the DF and phase of each VEGM PSs are detected by finding the curl of the spatial phase gradient. DF along with phase and PSs were plotted for each window and the behaviour of the trajectory of HDF 'clouds' was observed. Our results indicate that spatial and temporal organization correlating HDF and phase exists during persAF. Generating and analysing the maps of HDF and phase may prove helpful in understanding the spatial and temporal activation dynamics during persAF.