Low-frequency component of body surface potential maps identifies patients at risk for ventricular tachycardia

Aims. To investigate the ability of spectral features of signal-averaged body-surface potential maps in identifying post-infarction patients who are at risk of developing ventricular tachycardia. Methods and Results. We recorded 120 lead body surface potential maps during sinus rhythm in 135 subjects (45 patients with healed myocardial infarction but no history of venticular tachycardia, 45 patients with both healed myocardial infarction and at least one episode of sustained ventricular tachycardia, and 45 normal subjects) and analysed spectral features of body surface potential maps selected on the basis of isoharmonic maps for given bands of the frequency spectrum. We found that in the low-frequency band (1-11 Hertz), the group-mean power spectra of leads located at isoharmonic map maxima were significantly different (P < 0.0001) between the two groups of myocardial infarction patients. We estimated that this single feature alone can prospectively identify myocardial infarction patients at risk for ventricular tachycardia with a predictive accuracy of 74 ± 6%. Conclusion. Our results suggest that the bulk of diagnostic information associated with arrhythmogenicity resides in the low-frequency band of the power spectrum. This finding is at variance with the established notion that only the high-frequency component of signal-averaged electrocardiograms carries such information.