ECG signal processing using multiresolution analysis

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Electrocardiogram (ECG) signal processing is crucial for diagnosing cardiac conditions.
• Traditional methods face challenges with noise, baseline drift, and distinguishing QRS complexes from other wave components.
• Wavelet-based multiresolution analysis offers a promising approach for advanced ECG signal analysis.

Purpose of the Study:

• To evaluate the efficacy of multiresolution analysis using wavelets for ECG signal denoising.
• To develop and assess a wavelet-based method for accurate QRS complex detection in ECG signals.
• To improve the reliability and accuracy of automated ECG interpretation.

Main Methods:

• Implementation of a two-step algorithm utilizing multiresolution analysis and discrete wavelet transform (DWT).
• Step 1: Denoising of ECG signals using DWT-based multiresolution analysis.
• Step 2: Application of multiresolution analysis for precise QRS complex detection, differentiating it from P/T waves and artifacts.

Main Results:

• The denoising algorithm demonstrated good performance on MIT-BIH ECG signals.
• The QRS complex detection method achieved high accuracy, with a detection rate exceeding 99.8% (sensitivity=99.88%, predictivity=99.89%).
• The overall detection failure rate for QRS complexes was found to be as low as 0.24%.

Conclusions:

• Multiresolution analysis using wavelets is a robust technique for both denoising ECG signals and detecting QRS complexes.
• The developed approach effectively handles common ECG signal interferences like noise, baseline drift, and artifacts.
• This wavelet-based method significantly enhances the accuracy and reliability of automated ECG analysis, aiding clinical diagnosis.