Proceedings of the International Conference
I W S S I P   2005

22 - 24 September 2005, Chalkida Greece
(from Chapter 1: Invited Addresses and Tutorials on Signals, Coding, Systems and Intelligent Techniques)

 Full Citation and Abstract

0 Title: Level-dependent wavelet denoising: application to very noisy ECG signals
  Author(s): Sid Ahmed Chouakri, Fethi Bereksi-Reguig, Said Ahmaïdi, Odette Fokapu
  Address: Laboratoire Télécommunications et Traitement Numérique du Signal –LTTNS- Université de Sidi Bel Abbès BP 89 Algérie 22000
Laboratoire de Génie Biomédicale Université de Tlemcen Algérie 13000
Laboratoire EA330 «APS et Conduites Motrices : Adaptations et Réadaptations » Faculté des sciences du sport Allée P. Grousset, 80025, Université de Picardie Jules Verne Amiens Cedex, France
Université Technologique de Compiègne Laboratoire de Biomécanique et génie Biomédical UMR CNRS 6600 - BP 20529 60205 Compiègne, France
chouakri.s.a @
  Reference: SSIP-SP1, 2005  pp. 95 - 99
We present in this work an algorithm allowing the filtering of very noisy ECG signal, corrupted by a white Gaussian noise (WGN) with an SNR of around 0 dB. Our algorithm tends to solve the drawbacks faced to the classical wavelet denoising approach using the 'VisuShrink' threshold calculus method and the hard thresholding strategy, in the case of very noisy ECG signal, mainly the R wave distortion. Our key idea is to pass, first, the noisy signal through the classical low pass Butterworth filter, and, next, to use leveldependent threshold calculated basing, mainly, on the 'VisuShrink' methodology given by: Tj=(2≡log(Nj))1/2≡ median('Cj')/0.6745. Our study demonstrates that the optimal value of the Nj is the length of corresponding detail level (j) while the median('Cj') value is kept constant, along the different denoising levels, and is computed at the lowest resolution DWT, i.e. Cj¥cD1 (the 1st level detail coefficients) of the very noisy ECG signal. The obtained results of applying our algorithm to the record '100.dat' of the MIT-BIH Arrhythmia Database, corrupted with a WGN of an SNR of 0 dB, provides an output SNR of around 4.25 dB and an MSE of 0.0016. A comparative study using the classical wavelet denoising process, at 2 successive levels (4th, and 5th) and the classical low pass Butterworth filter provides the output SNRs of (3.65, 3.37, and 3.74 dBs) and mean square error (MSE) values of (0.0017, 0.0018, and 0.0018) respectively. These obtained results demonstrate the superior performance of our algorithm regarded to the set of the tested denoising approaches.
PDF  View Full PDF
 only subscribers
PDF  Order by phone
 or buy On-line

 We welcome your comments about this Article