A comparison of three electrode channels selection methods applied to SSVEP BCI

There has been increasing interest in using steady-state visual evoked potential (SSVEP) in brain-computer interface systems (BCIs). The electrode channels usually used in SSVEP classification are O1, O2 and Oz. However, optimal results of SSVEP recognition could not be obtained from the same electrode setting (O1, O2 and Oz), which was caused by subject variation. In this paper, three methods (Sequential floating forward selection (SFFS), discrete particle swarm optimization (DPSO) and F-score) were employed to select the optimal electrode channels. The electrode channels obtained by SFFS, DPSO and F-score were compared with traditional electrode channels (i.e., O1, O2 and Oz), which are usually used in SSVEP BCI. The results show that SFFS and DPSO can obtain higher classification accuracy than traditional approach. The results also show that SFFS is superior to DPSO in terms of calculating time, and F-score is not good compared to other two methods. Channel selection can not only reduce features for data analysis but also reduce the time for channel installation.