An algorithm for robust and efficient location of T-wave ends in electrocardiograms

Area of Science:

• Biomedical Engineering
• Cardiovascular Signal Processing

Background:

• Accurate T-wave end localization is crucial for electrocardiogram (ECG) analysis.
• Existing algorithms face challenges with noise and waveform variability.

Purpose of the Study:

• To introduce a new, robust algorithm for precise T-wave end detection in ECG signals.
• To validate the algorithm's performance against established databases and methods.

Main Methods:

• The algorithm computes an indicator based on the T-wave's curve area.
• It leverages the T-wave's concavity for formal proof of T-wave end identification.
• The core computation is implementable as a finite impulse response (FIR) filter.

Main Results:

• The algorithm demonstrates high accuracy in T-wave end localization.
• It shows robustness against common ECG signal interferences like noise, morphological variations, and baseline wander.
• Performance evaluation on the PhysioNet QT database indicates superior accuracy compared to other methods.

Conclusions:

• The proposed algorithm offers a simple yet effective solution for T-wave end localization.
• Its robustness and computational efficiency make it suitable for clinical applications.
• This method represents a significant advancement in ECG analysis accuracy.