Fast-Fourier transform analysis of signal-averaged electrocardiograms for identification of patients prone to sustained ventricular tachycardia.

Abstract

Electrocardiograms obtained from patients during arrhythmia-free intervals do not identify those prone to sustained ventricular tachycardia (VT) despite the occult delayed activation that is presumably present. To determine whether frequency-domain analysis facilitates detection of this hallmark of predisposition to VT, fast-Fourier transform analysis (FFTA) procedures were developed and tested with a computer-generated mathematical model. The FFTA approach developed allows inherent limitations of high-gain amplification and a priori filtering used commonly for time-domain analysis to be avoided. After demonstrating that FFTA detected low-amplitude oscillatory waveforms in signal-averaged recordings in the frequency domain, the procedure was applied to signal-averaged X, Y, and Z lead recordings from the following three groups of patients: group I, patients with prior myocardial infarction and episodic sustained VT (n = 16); group II, patients with prior myocardial infarction without overt sustained VT (n = 35); and group III, normal control subjects (n = 10). Results of FFTA demonstrated significant (p less than .0001) differences in the decibel drop at 40 Hz and the area under the curve from the fundamental frequency to the frequency at which the spectral amplitude was decreased by 60 dB for both the terminal 40 msec of the QRS and ST segment in patients in group I compared with those in groups II and III, in whom results were similar. Results were independent of QRS duration (r = .2), left ventricular ejection fraction (r = .19), and complexity of spontaneous ventricular ectopy. Thus, patients known to manifest sustained VT also exhibited relatively greater high-frequency content in arrhythmia-free intervals in the terminal QRS and ST segment than those without VT (88%, 15%, and 0% in groups I through III, respectively). FFTA offers promise for the noninvasive detection of patients at risk for the development of sustained VT.