Investigating the effects of visual distractors on the performance of a motor imagery brain-computer interface

Abstract

ObjectivesBrain-computer interfaces (BCIs) allow users to operate a device or application by means of cognitive activity. This technology will ultimately be used in real-world environments which include the presence of distractors. The purpose of the study was to determine the effect of visual distractors on BCI performance. MethodsSixteen able-bodied participants underwent neurofeedback training to achieve motor imagery-guided BCI control in an online paradigm using electroencephalography (EEG) to measure neural signals. Participants then completed two sessions of the motor imagery EEG-BCI protocol in the presence of infrequent, small visual distractors. BCI performance was determined based on classification accuracy. ResultsThe presence of distractors was found to affect motor imagery-specific patterns in mu and beta power. However, the distractors did not significantly affect the BCI classification accuracy; across participants, the mean classification accuracy was 81.5 ± 14% for non-distractor trials, and 78.3 ± 17% for distractor trials. ConclusionThis minimal consequence suggests that the BCI was robust to distractor effects, despite motor imagery-related brain activity being attenuated amid distractors. SignificanceA BCI system that mitigates distraction-related effects may improve the ease of its use and ultimately facilitate the effective translation of the technology from the lab to the home.