Comparison of different orthogonal systems in high resolution ECGs and the significance of the number of averaged beats in detecting late potentials

A H Pietersen,E Gymoese

DOI: https://doi.org/10.1093/oxfordjournals.eurheartj.a059945

Abstract:For non-invasive detection of late potentials in the signal-averaged electrocardiogram (SAECG) most studies use modified orthogonal leads. In the present study the significance of three different lead systems for the detection of late potentials was examined. SAECGs were recorded at five different (32, 64, 128, 256, and 512) numbers of averaged beats in 24 patients with ventricular tachycardia after myocardial infarction and in 76 normal subjects. The original Frank leads and two different modified orthogonal vector lead systems were used. There was no clinical difference between the three orthogonal systems in sensitivity and the difference in specificity (P less than 0.01) seemed to be primarily related to differences in noise content. When the vector systems were compared at equally low noise levels of 0.47-0.50 microV, the apparent differences in specificity between the vector systems disappeared. At decreasing noise levels the sensitivity increased, at the cost of a loss in specificity which at very low noise levels was out of proportion with the gain in sensitivity. A modified approach with a fixed threshold noise level to determine the QRS endpoint is suggested.