Detemination of decision moments in EEG signals using DWT

Wavelet transforms and wavelet based tools are widely used in EEG signal analysis. However, they are mainly used for seizure analysis or feature extraction tools for classification. In this paper, discrete wavelet transform (DWT) is used to answer a critical question of Brain Computer Interface (BCI). For a robot navigation application, if the moments that decision of the user changes are known, then the classification performance would increase since the correct time segment will be considered. Moreover, communication cost is reduced because the robot is only controlled at the decision moments.