Research on electrocardiogram signal noise reduction based on wavelet multi-resolution analysis

Electrocardiogram (ECG) signal´s noise reduction is critical for ECG automatic diagnosis and analysis, and the de-noising result directly affects the accuracy of ECG parameter extraction even the patients´ illness diagnosis and analysis. ECG signals are often seriously distorted by noises with the characteristics of low signal-noise ratio and non-stationary stochastic property, especially for P wave and T wave because they have lower amplitude and are sensitive to noise. This paper presents an ECG signal noise reduction method utilizing coif3 wavelet and notch filter. Experiment results show that this algorithm effectively solved the serious ECG signal distortion problem caused by baseline drift, power line interference and EMG interference, greatly inhibits the noise in P wave and T wave, and improves the accuracy of ECG parameter extraction.