The signal-averaged ECG becomes late potential-positive at low noise levels in healthy subjects.

Late potentials are detected at various noise levels in clinical studies. The aim of this study was to quantify the effect of residual noise level on the signal-averaged ECG. Thirty minutes of raw ECG (lead -X, +X, -Y, +Y, -Z, and +Z) from 10 healthy volunteers were digitized and stored on optical discs. Each ECG was analysed by four signal averaging procedures to noise level 0.1, 0.2, 0.3, and 0.4 microV, respectively. At each noise level, time domain analysis of the filtered vector (40-250 Hz) included determination of the QRS duration (QRS), late potential duration (LPD, duration of terminal signals below 40 microV), and root-mean-square voltage of the terminal 40 ms of the filtered QRS (RMS40). On average, the measured signal-averaged QRS duration was prolonged by 7.0 ms (range 0.9-12.5 ms) per 0.1 microV reduction in noise level. Late potential duration increased by 5.9 ms (range 0.2-10.0 ms) per 0.1 microV reduction in noise level, and RMS40 was reduced by 9.1 microV (range 0.5-20.2 microV) per 0.1 microV reduction in noise level. At noise level 0.4 microV, 0.3 microV, 0.2 microV, and 0.1 microV the number of late potential-positive subjects were 1, 2, 4, and 6, respectively. Late potential parameters are significantly dependent on noise level after signal averaging. Using conventional criteria for late potentials, healthy subjects become false-positive at low noise levels. Establishment of standards for noise level is recommended.