[Limitations of the common pre-processing method in the signal-averaged electrocardiogram]

Area of Science:

• Cardiology
• Biomedical Signal Processing
• Computational Biology

Context:

• Isolating micro potentials from electrocardiograms (ECG) necessitates precise signal alignment.
• Coherence averaging is a technique used for signal isolation, requiring accurate temporal alignment.
• The maximum correlation coefficient (MCC) method is a potential tool for achieving this alignment.

Purpose:

• To evaluate the reliability of the maximum correlation coefficient (MCC) method for ECG signal alignment through computer simulation.
• To analyze the influence of correlation-window-width and threshold settings on the accuracy of temporal alignment and averaged results.
• To assess the MCC method's performance in signal recognition and its sensitivity to amplitude and duration variations.

Summary:

• Computer simulations demonstrated that the MCC method is insensitive to amplitude and duration variations but sensitive to window-width and threshold settings.
• Narrower correlation window widths reduce sensitivity to misalignment, yet specific widths (40 and 100 ms) with a 0.99 threshold caused significant deformation in averaged results.
• The MCC method's deficiencies, particularly concerning temporal alignment accuracy, limit its reliability for isolating atrial and ventricular late potentials.

Impact:

• Highlights significant limitations of the MCC method in accurately aligning ECG signals for micro-potential isolation.
• Suggests that current MCC method parameters may compromise the integrity of averaged ECG data in both time and frequency domains.
• Underscores the need for improved signal alignment techniques to reliably detect and analyze cardiac late potentials.