A method for recognizing high-frequency steady-state visual evoked potential based on empirical modal decomposition and canonical correlation analysis

In most of current studies on SSVEP based BCIs, low-frequency and medium-frequency visual stimuli are used, and subjects are prone to fatigue. The BCI based on high-frequency SSVEP can improve the comfort level of subjects in experiment and reduce the possibility of inducing diseases such as epilepsy. This paper proposed a method combining empirical modal decomposition (EMD) and canonical correlation analysis (CCA) to improve the classification accuracy of high-frequency SSVEP. The experiment results show that the EMD-CCA based method is more suitable for high-frequency SSVEP based BCI, which can achieve a maximum accuracy of 93.68% and an information transmission rate of 15.0236 bit/min
^-1
.