Noninvasive assessment of Wedensky modulated signal-averaged electrocardiograms

Subthreshold stimulation without capture reduces the stimulation threshold and changes the action potential of subsequent suprathreshold stimulation, a phenomenon known as Wedensky modulation (WM). Patients with ventricular tachycardia (VT) inducible during electrophysiological testing (n = 47, mean age 63 +/- 13 years, 83% men), and healthy controls (n = 30, mean age 44 +/- 16 years, 60% men) were subjected to transthoracic external subthreshold stimulation between surface precordial and left subscapular patch electrodes. Stimuli of 5, 10, 20, and 40 mA were delivered for 2 ms, in synchrony with, or 20 ms after, R wave detection. A total of 60-200 subthreshold stimulated QRS complexes were averaged and compared with averaged nonstimulated complexes recorded during the same experimental session. To detect transient changes within the QRS complex, both signals were decomposed with 54 scales of Morlet analyzing wavelets (central frequencies 40-250 Hz). Wavelet vector magnitude was obtained for stimulated and nonstimulated complexes. Their difference created a wavelet residuum (WR) that characterized WM numerically. The surface area of the three-dimensional envelope of WR was measured and statistically compared between VT patients and healthy controls. WR showed a significantly greater increase in the spectral power of the stimulated complex in healthy controls than in VT patients (P < 0.01). In conclusion, (1) wavelet decomposition is a suitable tool to analyze WM, (2) WM in the late QRS complex is short, and (3) VT patients are less sensitive to WM, particularly at low subthreshold energies.