Detection of the QRS complex by linear prediction

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• The electrocardiogram (ECG) records the heart's electrical activity, featuring periodic wave sequences like P, QRS, and T waves.
• The QRS complex is a critical component for analyzing cardiac function and detecting abnormalities.
• Accurate QRS complex detection is essential for reliable ECG interpretation and automated cardiac diagnostics.

Purpose of the Study:

• To develop and evaluate a novel algorithm for precise QRS complex detection in ECG signals.
• To enhance the accuracy of QRS detection by minimizing detection errors.
• To compare the performance of the new algorithm against established methods like Pan and Tompkins.

Main Methods:

• The algorithm incorporates signal-to-noise enhancement, linear prediction for ECG analysis, nonlinear transformation, a moving window integrator, and center-clipping transformation.
• Linear prediction coefficients are determined using a least-squares approach to minimize prediction error.
• The residual error signal from linear prediction is utilized for localizing and detecting QRS complexes.

Main Results:

• The developed QRS detection algorithm was tested on the MIT-BIH arrhythmia database.
• Performance was compared against the widely used Pan and Tompkins QRS detection method.
• The proposed algorithm demonstrated superior performance, yielding fewer false positives and false negatives.

Conclusions:

• The novel algorithm offers improved accuracy and reliability for QRS complex detection in ECGs.
• This method presents a significant advancement over existing techniques, enhancing diagnostic capabilities.
• The algorithm's effectiveness in reducing detection errors makes it a valuable tool for clinical applications.