Response to Letter Regarding Article, “Dominant Frequency of Atrial Fibrillation Correlates Poorly With Atrial Fibrillation Cycle Length”

Abstract

Frequency-domain analysis has been considered as an alternative way to visualize atrial fibrillation (AF) electrograms for estimating the rate and regularity of activation without having to measure intervals in the time domain. In a response to our article, 1 Platonov et al counter that AF cycle length (AFCL) estimation is too complex and that a method for finding the fundamental or dominant frequency (DF) is to be preferred.Although we agree that it is not simple to create an algorithm that reliably detects activation under all circumstances, we disagree that this implies that DF computation must be preferred for guiding AF ablation. There are 2 main reasons why AFCL is to be preferred in a clinical application for this.One is that, as shown in Figure 8 of our article, DF is a reliable tool as long as the activations are regular enough. 1 This figure also shows that accidentally selecting the double frequency is not the only problem. The frequency spectrum is also affected by difference in cycle length, amplitude, and morphology.This means that precisely in cases that are clinically interesting, DF is unreliable. There is a famous quote attributed to H.L. Mencken that "for every complex problem, there is a solution that is simple, neat, and wrong." We are sure that he was not referring to DF-guided ablation of AF.Our second reason is that DF analysis is performed on electrograms that are heavily filtered and distorted before applying the Fourier transform. Not many cardiologists are experienced in thinking in frequency domain. Richter et al 2 use a much more complex method that, although it might perform better under certain circumstances, is even less accessible to electrophysiologists. If DF analysis gives a suspicious or even wrong result, it is therefore hard to judge where and why the analysis went wrong. AFCL analysis, on the other hand, can at all times show the signals and the automatically detected activation times. Interpreting electrograms is exactly the expertise of clinical electrophysiologists. Therefore, AFCL is a method that helps the physician to understand the substrate, whereas DF analysis is just a computational tool that computes a number and is often not a reliable measure for AF activation times. DisclosuresNone. Arif Elvan, MD, PhD