WICA And Its Application In Extracting Brain Evoked Potential

With the development of the brain electrical signal processing technology, using effective and accurate method to analyze the mystery of the brain becomes one of the biggest science challenges. Meanwhile, the extraction of brain evoked potential has been widely used in the neurology department and other domains. But a series of recently electrical signal processing technology have not enough perfect.Some useful ingredients which brain electrical signal contains generally all are the transient state weak signal, so according to the respective characteristic of the brain electrical signal and the signal analysis method, applies in extracting brain evoked potential by wavelet transform (WT) and the Independent Component Analysis (ICA) union method (WICA). After wavelet transform, in the sub- belt reorganization multi-channel ICA input signal, the non- target signal ingredient and the unwanted signal ingredient generally changes weakly, therefore, the ICA algorithm can effectively separate stronger target signal ingredient.This article first applies the WICA algorithm in the epilepsy characteristic wave enhancement and the alpha brain electrical signal wave enhancement. Separately according to the epilepsy wave and alpha wave characteristic, eliminate the labor frequency falls the wave, EMG artifacts and the corresponding noise through the continual wavelet decomposes, then reorganizes the ICA input to carry on the analysis again. After the experiment, obtains a better result.Next application in global BCI competition III . First uses the wavelet filter, strengthens P300 which the frequency band concentrate in 2 ~ 8Hz, then uses the ICA algorithm training and the testing, finally the component projection which selects returns on the scalp, causes the P300 peak to be more prominent in order to examines. The experimental result confirmed the WICA algorithm playing roles in extraction of brain evoked potential.