ANALYSIS OF ATRIAL FIBRILLATION BY AUTOCORRELATION FUNCTION

摘要

Introduction: The purpose of this study was to investigate the usefulness of the autocorrelation function (reversed fast Fourier transform analysis) in determining the atrial fibrillation cycle length (AFCL) during human atrial fibrillation (AF). Methods and Results: From 30 episodes of atrial electrograms recorded for 30 seconds from the high right atrium during type I AF in 16 patients, the mean, 5th percentile (p5) and 95th percentile (p95) of the AFCLs were measured by using a computer picked activation time. The peak, minimum and maximum AFCLs were also measured by using the autocorrelation function. The mean AFCL was retrieved at the point of the maximum peak of the coefficient of the first positive autocorrelogram. The minimum AFCL (min AFCL) was chosen as the point where the first positive autocorrelogram crossed the baseline from negative to positive, and the maximum AFCL (max AFCL) was chosen as the point where the first positive autocorrelogram crossed the baseline from positive to negative. There was a significantly strong correlation between the mean and peak AFCLs (r=0.995, p<0.0001), p5 and min AFCLs (r=0.953, p<0.0001), and p95 and max AFCLs (r=0.98, p<0.0001). Conclusion: The autocorrelation function was useful in determining the AFCLs, at least during type I AF.