[Comparative Study on the Three Algorithms of T-wave End Detection: Wavelet Method, Cumulative Points Area Method and...

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Accurate T-wave end point detection is crucial for clinical analysis of electrocardiograms (ECGs).
• Existing threshold-dependent methods for T-wave end point detection have limitations in clinical settings.
• Evaluating alternative algorithms is necessary to improve detection performance.

Purpose of the Study:

• To compare the performance of wavelet, cumulative point area, and trapezium area methods for T-wave end point detection.
• To develop and validate an adaptive selection algorithm for T-wave end point detection based on T-wave morphology.
• To enhance the efficiency and accuracy of T-wave end point detection in clinical ECG analysis.

Main Methods:

• Utilized the PhysioNet QT database, comprising 20 records with 3,569 beats each.
• Employed the wavelet method for initial QRS complex and T-wave localization.
• Applied wavelet, cumulative point area, and trapezium area methods for T-wave end point detection, analyzing T-wave morphology variations.

Main Results:

• The proposed adaptive selection algorithm demonstrated superior performance compared to single T-wave end point detection algorithms.
• Achieved high detection performance metrics: 98.93% sensitivity and 99.11% positive predictive value.
• Reported an average time error of (-2.33 ± 19.70) ms, indicating precise detection.

Conclusions:

• The adaptive selection algorithm based on T-wave morphology significantly improves T-wave end point detection efficiency.
• This novel approach offers a more robust and accurate solution for clinical ECG interpretation.
• The findings support the clinical utility of morphology-based adaptive algorithms for ECG analysis.