Frequency measures obtained from the surface electrocardiogram in atrial fibrillation research and clinical decision-making

Abstract:
Frequency Measures in AF. Introduction: Frequency analysis of fibrillation (FAF) and time-frequency analysis (TFA) were developed recently in order to quantify atrial electrical remodeling in atrial fibrillation (AF) from the surface ECG. This article describes the experience with these two different frequency analysis techniques in consecutive AF patients and discusses possible applications in AF research and clinical decision-making. Methods and Results: Baseline 2-minute, high-gain, high-resolution ECG recordings using three bipolar leads were obtained from 80 consecutive patients with AF lasting >24 hours. A power spectrum was obtained using Fourier analysis following spatiotemporal QRST cancellation. The dominant fibrillatory rate (in fibrillations per minute [fpm]) was derived (FAF). Stability of the instantaneous fibrillatory rate measured in overlapping 1-second segments was expressed as the segment proportion with consecutive rate differences <6 fpm (TFA). An adequate power spectrum that could be submitted for determination of fibrillatory rate was obtained in all patients. Dominant atrial rates ranged between 288 and 534 fpm and showed a high correlation (R = 0.878-0.911, P < 0.001) when assessed from the three different leads. The average instantaneous fibrillatory rate was inversely related with its stability (R = -0.417, P < 0.001). It was closely related with the dominant fibrillatory rate obtained from FAF (R = 0.948, P < 0.001). A literature review revealed that pharmacologic or electrical cardioversion and AF pace termination success rates were highly dependent on fibrillatory rate. Conclusion: Atrial fibrillatory rate and its variability can be reliable obtained from the surface ECG in AF patients. These parameters exhibit a significant interindividual variability allowing individual quantification of the atrial electrical remodeling process and might prove useful for predicting therapy efficacy.