Comparison of on-line wavelet analysis and reconstruction: with application to ECG

LESTUSSI, F.; DI PERSIA, L. E.; MILONE, D.H.

Wuhan

Conferencia; 5th International Conference on Bioinformatics and Biomedical Engineering; 2011

The wavelet transform is a consolidated tool for signal analysis. However, it has limitations for processing on-line and long duration signals. In this paper we applied three alternatives for on-line processing based in the wavelet transform which are suitable for biomedical signals. The methods for on-line wavelet processing was compared with the off-line DWT as baseline reference. The output for each methods was analyzed in detail by using the mean square error and the normalized maximum amplitude error. We concluded that the best performance is obtain with Hamming windows and overlap-and-add for reconstruction. As an application example, we show good results for the on-line processing of ECG signals.