P371: Selective attention and performance in controlling a P300-based brain computer interface in people with amyotrophic lateral sclerosis

Question: We investigated the involvement of attentional and memory processes in controlling a P300-based brain-computer interface (BCI) for communication purpose in people with amyotrophic lateral sclerosis (ALS). Methods: Eight participants with ALS performed a P300-based BCI spelling task and: (i) a rapid serial visual presentation (RSVP) task, screening the temporal filtering capacity and the speed of the update of such attentive filter, in order to detect two targets in a stream of distractors, and (ii) a change detection (CD) task with the simultaneous presentation of target and distractor stimuli, to screen the spatial filtering capacity of attention for visual working memory consolidation. We performed correlation and regression analyses to investigate the relationship between the indexes of selective attention and visual working memory and the amplitude of P300 event related potential (ERP) elicited during a P300-based BCI task as well as of the performance obtained in such task. Results: Amongst the indexes evaluated, only the temporal filtering capacity was found to significantly correlate with both P300 amplitude (r=0.84, p<0.05), and the accuracy achieved during the BCI task (r=0.79, p<0.05). We therefore conducted a linear regression analysis with such index as independent factor, revealing that it significantly predicted both P300 amplitude and BCI accuracy (p<0.05). Conclusion: The overall findings indicate that the capability to selectively filter a stimulus and to keep such filter active in a temporal domain influences the performance in achieving a P300-based BCI control. The present study partly clarifies the cognitive substrates related to BCI control in people with ALS.