QRS detection using S-Transform and Shannon energy

Area of Science:

• Biomedical Signal Processing
• Cardiovascular Diagnostics
• Time-Frequency Analysis

Background:

• Accurate QRS complex detection is crucial for electrocardiogram (ECG) analysis.
• Existing methods face challenges with complex ECG patterns and noise.
• Novel time-frequency representations offer potential for improved QRS detection.

Purpose of the Study:

• To introduce a new QRS detection method utilizing the S-Transform.
• To leverage the frequency-dependent resolution of the S-Transform for enhanced QRS complex isolation.
• To improve the accuracy and reliability of R-wave localization in ECG signals.

Main Methods:

• The proposed method employs the S-Transform, a time-frequency representation (TFR).
• QRS complexes are isolated in the time-frequency domain by exploiting S-Transform properties.
• Shannon energy of local spectra is computed for precise R-wave localization.

Main Results:

• The S-Transform based method demonstrated significant performance enhancement on the MIT-BIH arrhythmia database (MITDB).
• Achieved a sensitivity of 99.84%, positive predictivity of 99.91%, and an error rate of 0.25%.
• Detection parameters were illustrated for complex ECG segments, confirming robustness.

Conclusions:

• The S-Transform provides a powerful tool for QRS detection in ECG signals.
• The proposed method offers superior accuracy and reliability compared to existing techniques.
• This approach holds promise for advanced cardiovascular diagnostics and arrhythmia monitoring.