A P300-speller based on event-related spectral perturbation (ERSP)

Abstract

A brain-computer interface (BCI) P300 speller is a novel technique that helps people spell words using the electroencephalography (EEG) without the involvement of muscle activities. However, only time domain ERP features (P300) are used for controlling of the BCI speller. In this paper, we investigated the time-frequency EEG features for the P300-based brain-computer interface speller. A signal preprocessing method integrated ensemble average, principal component analysis, and independent component analysis to remove noise and artifacts in the EEG data. A time-frequency analysis based on wavelet transform was carried out to extract event-related spectral perturbation (ERSP) and inter-trial coherence (ITC) features. Results showed that the proposed signal processing method can effectively extract EEG time-frequency features in the P300 speller, suggesting that ERSP and ITC may be useful for improving the performance of BCI P300 speller.