The Sinusoidal Assisted MEMD based CCA Method for SSVEP based BCI Improvement

Although the canonical correlation analysis (CCA) algorithm has been applied successfully to SSVEP detection, artifacts and unrelated brain activities may influence the performance of the steady state visual evoked potential (SSVEP) based brain-computer interfaces (BCI) system. Extracting the characteristic frequency sub-bands is an effective method to enhance the signal-to-noise-ratio of SSVEP signals. The sinusoid-assisted MEMD (SA-MEMD) algorithm is a powerful method for spectral decomposition. In this study, we propose an SA-MEMD based CCA method for SSVEP detection. The results suggest that the SA-MEMD based CCA algorithm is a useful method in the detection of typical SSVEP signals. The classification accuracy achieved 88.3% in a 4 s time window and there was a 2.8% improvement compared with the standard CCA algorithm.