Wavelet based ECG denoising by employing Cauchy distribution at subbands

This paper presents a new ECG denoising method based on the modeling of wavelet coefficients in each subband with a Cauchy probability density function. By using this statistical model, thresholds are estimated at each level on wavelet coefficients for noise reduction. We evaluated this approach on offline single lead noisy ECG records from Cardiovascular Research Centre of University of Glasgow. Results show that our proposed technique provides better performance factors i.e. signal-to-noise ratio (SNR) and percentage of zeros (PZ).