Fiducial segment averaging to improve cardiac time interval estimates

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Accurate fiducial point estimation in electrocardiogram (ECG) signals is critical for diagnosing cardiac conditions.
• Traditional methods can be susceptible to noise and variability, impacting measurement precision.

Purpose of the Study:

• To develop and validate a new technique for improving the accuracy of fiducial point estimates in ECG signals.
• To enhance the precision of measurements for key ECG intervals, including PR, QRS-duration, and QT.

Main Methods:

• A novel technique utilizing beat-to-beat coherence of P-QRS-T complex segments was developed.
• QRS complexes were identified using matched filtering, followed by iterative cross-correlation for fine-tuning.
• Fiducial points (P-onset, QRS-onset/offset, T-peak/end) were refined using a 60 ms segment correlation approach.

Main Results:

• The technique demonstrated improved accuracy in fiducial point estimation across various ECG sources and leads.
• Precise measurements of PR-interval, QRS-duration, and QT intervals were obtained.
• The method preserved beat-to-beat variations, enabling the analysis of heart-rate-dependent relationships like QT/RR.

Conclusions:

• The developed technique significantly enhances the accuracy of ECG fiducial point and interval measurements.
• This improved precision facilitates more reliable analysis of cardiac electrophysiology, including heart-rate dependencies.
• The methodology offers a robust approach for both single and multi-lead ECG analysis.