Fast Fourier transformation of the entire low amplitude late QRS potential to predict ventricular tachycardia
DL Pierce,
AR Easley Jr,
Windle JR,
and
TR Engel
Department of Internal Medicine, University of Nebraska College of Medicine, Omaha.
Signal-averaged electrocardiograms (X, Y and Z leads) were acquired from 24 patients with coronary artery disease and recurrent ventricular tachycardia, 24 control patients with coronary artery disease and 23 normal subjects to assess the discriminant value of fast Fourier transformation of the entire late potential period of the QRS complex. Analysis of the vector magnitude in the temporal domain (25 to 250 Hz bandpass filters) measured high frequency QRS duration, the duration of terminal signals less than 40 microV and the root mean square voltage of the last 40 ms. Late potentials were defined as terminal signals greater than 25 Hz that were less than 40 microV. Analysis in the frequency domain used a 120 ms window that encompassed (had onset with) all of the late potential, but the mean value was first subtracted to eliminate a direct current component. High frequency spectral areas (60 to 120 Hz) and the percent high frequency (100 x [60 to 120 Hz/0 to 120 Hz]) were calculated. Results in both temporal and frequency domains were similar in control patients with coronary artery disease and normal subjects. Patients with ventricular tachycardia had a longer high frequency QRS complex (p less than 0.0001) and longer high frequency terminal signals less than 40 microV (p less than 0.0004), but not significantly lower voltage in the last 40 ms. The most useful temporal domain measurement was high frequency QRS duration (if greater than or equal to 120 ms, odds ratio = 8.2). Patients with ventricular tachycardia had increased high frequency spectral areas (p less than 0.0002) in the late potential, and the percent high frequency was especially increased (p = 0.0000; if percent high frequency greater than 3.1%, odds ratio = 18.4). The odds ratio and the area under the receiver operating characteristic curve were both greater for percent high frequency than for high frequency QRS duration (p less than 0.03). All patients with ventricular tachycardia had a high frequency QRS complex greater than or equal to 107 ms or percent high frequency greater than or equal to 3.1% (sensitivity 100%). For a high frequency QRS complex greater than or equal to 107 ms and percent high frequency greater than or equal to 3.1%, specificity was 96%. Therefore, high frequencies in late potentials, not their duration or reduced voltage, most usefully identify patients with coronary artery disease who are prone to ventricular tachycardia.
