Validation and Clinical Application of Time‐Frequency Analysis of Atrial Fibrillation Electrocardiograms

INTRODUCTION: Fibrillatory rates can reliably be obtained from surface ECGs during atrial fibrillation (AF) and correspond with right atrial (RA) and coronary sinus (CS) rates, while both the relation with pulmonary venous (PV) rates and determinants of fibrillatory waveform are unknown. Class III antiarrhythmic drugs prolong atrial refractoriness and decrease its dispersion, effects that may be reflected in ECG parameters. Consequently, this study sought (1) to investigate the relation between ECG fibrillatory rate and waveform characteristics with intraatrial/PV fibrillatory activity and (2) to noninvasively monitor class III antiarrhythmic drug effects in patients with AF. METHODS AND RESULTS: Thirty-six patients with drug-refractory AF who underwent catheter-based pulmonary vein isolation and had AF at the beginning of the procedure were studied. A positive correlation between V1 rates obtained by time-frequency analysis and RA (R = 0.97, P < 0.001), CS (R = .71, P < 0.001), and PV rates (R = 0.65, P = 0.001) was found. Exponential decay defined as decay of the curve that connects power maxima of dominant and harmonic frequency components correlated with RA rate dispersion (R = 0.53, P = 0.004). In amiodarone-treated patients (n = 7), V1 rate (286 +/- 64 vs. 371 +/- 40 fpm, P < 0.001) and exponential decay (1.06 +/- 0.29 vs. 1.38 +/- 0.38, P = 0.034) were lower than in patients without amiodarone (n = 29). In 19 additional patients with persistent AF, oral dofetilide treatment decreased mean fibrillatory rate from 377 +/- 57 to 294 +/- 50 fpm (P < 0.001) and exponential decay from 1.24 +/- 0.43 to 0.85 +/- 0.22 (P = 0.002). CONCLUSIONS: Fibrillatory waves of surface ECG lead V1 closely reflect right atrial, and, to a lesser degree, left atrial activity. Time-frequency analysis allows noninvasive monitoring of antiarrhythmic drug effects on fibrillatory rate and waveform.