Optimizing stimulation frequency for BCI-based color vision assessment: Preliminary results

Brain-computer interfaces (BCIs) have recently been demonstrated as a new method for assessing color vision (CV) that do not require the user's active participation, enabling CV assessment in young children, people with cognitive/motor impairments, and those with complex communication disorders. These BCIs use steady-state visual evoked potentials (SSVEPs) to measure similarities in color between pairs of light sources with different spectral distributions (i.e., they measure metamerism). Differences in the SSVEPs elicited by specific pairs of light sources enable the BCI to discriminate between people with and without CV deficiencies. Present BCI-based CV assessment systems elicit SSVEPs using 10 Hz stimulation. Here, we investigated the effect of stimulation frequency (from 2–38 Hz) on BCI-based CV assessment. In addition, we assessed whether there were differences in this effect for stimuli that had equal hues, but different illuminances versus stimuli that had different hues, but equal illuminances. Results show that SSVEPs elicited by stimuli with different hues are larger at lower frequencies than higher frequencies. Conversely, SSVEPs elicited by stimuli with different illuminances are smaller at lower frequencies than higher frequencies. Overall, 16 Hz was identified as the optimal stimulation frequency for BCI-based CV assessment. Preliminary results show that optimizing stimulation frequency can enhance the performance of BCI-based CV assessment and independent assessment of hue- versus illuminance-dominated differences in light sources may be possible.