Wavelet transform in detection of the subject specific frequencies for SSVEP-based BCI

Abstract

One of the paradigms often used to build a brain computer interface (BCI) is a paradigm based on steady state visually evoked potentials (SSVEPs). In SSVEP-based BCI a user is stimulated with a set of light sources flickering with different frequencies. In order to ensure the best performance of the interface built according to this paradigm, the stimulation frequencies should be chosen individually for each user. Usually, during the frequency scanning phase the user-specific stimulation frequencies are chosen according to the power of the corresponding SSVEPs. However, not only the power should be taken into account when choosing the stimulation frequencies. The second very important factor is the time needed to develop the prominent SSVEP. The wavelet transform (WT) seems to be an excellent tool for dealing with this task, since it provides not only the information about the frequency components represented in the signal but also about the time of their occurrence. The aim of this paper is to present a procedure, based on WT, that can be used to determine the user-specific frequencies with respect to the synchronization time and its strength.