Estimation of evoked potentials based on the wavelet analysis

For the effective removal of artifacts and the extraction of an improved evoked potential response, we propose the method using the shrinkage of wavelet coefficients. The wavelet analysis decomposes the measured evoked potentials into scale coefficients with low frequency components and wavelet coefficients with high frequency components as a resolution level, respectively. In the wavelet domain, artifacts are dispersed mainly at the wavelet coefficients rather than the scaling coefficients. Thus, in the course of synthesis evoked potentials, this method shrinks the wavelet coefficients to reduce the effects of artifacts in the wavelet domain and then reproduces the evoked potentials with the constricted coefficients, and lastly averages them. We collected visual evoked responses to simulate the method using the shrinkage of wavelet coefficients and obtained VEP after reproducing evoked responses using the wavelet analysis, and compared it with averaged signal. As a result of simulations, we got VEP with improved SNR in comparison with the averaging method at Daubechies wavelet in the resolution level four. We confirmed that the 32 times of averaging was enough to get normal VEPs