Does Inter-Stimulus Distance Influence the Decoding Performance of SSVEP and SSMVEP BCI?

Brain-computer interfaces (BCIs) based on steady-state visually evoked potentials (SSVEP) are considered as one of the most successful noninvasive BCI paradigms. Recently, an evolution of SSVEP was introduced with repetitive motion as visual stimuli instead of the conventional flickering stimuli. This novel paradigm is steady-state motion visually evoked potential (SSMVEP), and it has been proposed to address some inherent limitations of the flickering stimulus. One important factor that affects the decoding performance of BCIs based on repetitive visual stimulus is the change in inter-stimulus distance (ISD) or competing stimuli. In this study, a preliminary analysis was performed to compare the influence of change in ISD on the decoding performance of SSVEP and SSMVEP paradigms. The experiments were performed on thirteen healthy participants in which two visual stimulus types: flickering stimuli and checkerboards stimuli, were presented. Three different ISDs were studied. The decoding performance was quantified by the canonical correlation analysis (CCA) on O1, O2, and Oz channels. With 1-second window length, the average decoding accuracy across all ISDs was 75% for the SSVEP stimulus and 83.6% for the SSMVEP stimulus, respectively, and the corresponding average variation across ISDs was 0.13 and 0.03. The results demonstrated striking robustness of SSMVEP against competing stimuli while maintaining an overall superior accuracy to that of the conventional SSVEP.