Feature enhancement of P300 based brain computer interface through spatially-constrained ICA

The P300 word speller is one of the important BC´ applications which detects real-time P300 waveforms and translates them into letters (and then words) within a particular BC´ paradigm. However due to the poor SNR of EEG, as well as the presence of other artifacts, the identification accuracy is still not high enough for real-world application. This paper presents two slightly different ´CA approaches to improve character classification performance based on ´CA. When compared with the classification results obtained from the data, the results using these approaches show distinct improvement. Furthermore, the results indicate that it is possible to reduce the number of epochs required to perform stimulus locked averages, whilst still maintaining good performance measures. This has the potential of speeding up the word speller and has further implications for the use on similar ERP based systems.