UNSTABLE CARDIAC SINGULARITIES MAY LEAD TO ATRIAL FIBRILLATION

Abstract

A total of 53 atrial electrograms were recorded from 12 human patients diagnosed with different degrees of atrial arrhythmias and fibrillation, but not atrial flutter. The atrial waves were highly complex, noisy, nonuniform, nonlinear, and nonstationary in time and well suited for recurrence quantification analysis (RQA), spectral analysis (FFT) and atrial rate (AR) measurements. Differing degrees of atrial arrhythmias were quantified by measuring singularities in the electrograms. Singularities were defined as the maximum periods of relative isopotential squared (msec2) and presented as unfilled squares along the central line of identity (LOI) on recurrence plots. These nonsolid (unfilled) squares indicate that most singularities were unstable with noisy baselines. All measured variables were plotted against their corresponding unstable singularities. The best correlations were found for variables Vmaxand Laminar over the full range of log10(singularity). That is, the higher the degree of fibrillation the smaller the size of the singularity and the shorter Vmaxand Laminar. The shorter singularities are associated with faster spiral waves. However, since Vmaxand Laminar are direct derivatives of Singularity, this variable remains the sole best quantifier of choice to identify aberrant pacemaker regions.