Nonlinear dynamic extraction of cognitive evoked potential based on wavelet transform

Cognitive evoked potential (CEP) is an important description of brain cognitive function. At present average CEP of several measurements is used to extract CEP. However, as the latency of CEP varies at a large area, averaging brings a serious distort of shape and amplitude. So, nonlinear dynamic extraction of CEP using one or a few trials has been widely discussed. In this paper, we will introduce three typical extraction methods circumfusing wavelet transform (WT). They reflect the research trend of CEP extraction and the great function of WT. They are: wavelet de-noising, subspace method and wavelet de-noising, wavelet transformation and blind source separation.