[A squeeze approach for electrocardiogram ST-segment detection based on R-wave and T-wave].

Area of Science:

• Cardiology
• Biomedical Signal Processing
• Medical Informatics

Context:

• Electrocardiogram (ECG) signal analysis is vital for diagnosing myocardial ischemia, with ST-segment deviation being a key indicator.
• Accurate ST-segment feature point detection is challenged by signal interferences like baseline wandering and T-wave variability.
• Existing ST-segment detection methods (R+x, J+x) are susceptible to T-wave morphology and J-point location inaccuracies.

Purpose:

• To develop a robust and accurate method for T-wave onset detection independent of T-wave peak location.
• To introduce a novel ST-segment detection approach utilizing R-wave peak and the proposed T-wave onset detection.
• To validate the proposed method's performance using the Long-Term ST (LTST) database.

Summary:

• A new T-wave onset detection algorithm is presented, demonstrating robustness against baseline wander and T-wave morphology variations without needing T-wave peak identification.
• A subsequent ST-segment detection method is proposed, leveraging R-wave peak and the novel T-wave onset detection.
• Validation on the LTST database confirmed the method's timeliness, robustness, and achieved over 92% accuracy in ST-segment detection.

Impact:

• Enhances the accuracy and reliability of ST-segment analysis in ECG, potentially leading to earlier and more precise myocardial ischemia diagnosis.
• Provides a more robust alternative to existing ST-segment detection methods, particularly in the presence of signal noise and morphological variations.
• The developed algorithm offers a valuable tool for automated ECG interpretation systems, improving clinical decision-making in cardiology.