Adaptive nonlinear principle component analysis based multilayer neural network for P300 detection

In the experiment, four different inter-stimulus intervals (ISIs) are utilized: 325 ms, 350 ms, 375 ms, and 400 ms. The applicability of an adaptive nonlinear principle component analysis method for extracting the P300 waves included in the EEG signals without down-sampling and averaging of the original signals was demonstrated. Back-propagation neural networks were used as the P300 classifier. After a short time of practice, most participants could learn to extract and classify the P300 wave with greater than 80% accuracy. The experiment using different ISI shows the related variations of P300 wave to visual stimuli in normal human subject.