Brain-computer interaction for online enhancement of visuospatial attention performance

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Cognitive Science

Background:

• Visuospatial attention is crucial for reacting to visual stimuli.
• Existing methods for tracking attention can be intrusive or unreliable.
• Brain-computer interfaces (BCIs) offer a novel approach to monitor and potentially enhance cognitive functions.

Purpose of the Study:

• To reliably decode self-directed shifts in visuospatial attention using electroencephalography (EEG) data.
• To investigate whether decoded attention information can improve visuospatial task performance.
• To explore the potential of BCIs for enhancing human-machine interaction.

Main Methods:

• An experiment extended the Posner paradigm using novel, imperceptible ambiguous cues to elicit endogenous attention shifts.
• Two protocols were developed: an 'adaptive' protocol using decoded attention to display targets and a 'warning' protocol alerting users to target-locus mismatches.
• Visuospatial performance was assessed via reaction time and error rate in a target orientation discrimination task.

Main Results:

• Both the adaptive and warning protocols led to improvements in reaction time.
• The adaptive protocol specifically resulted in a reduction of errors.
• Online experiments with ten subjects confirmed the efficacy of the developed protocols.

Conclusions:

• Real-time decoding of EEG signals can accurately capture self-directed visuospatial attention shifts.
• Visuospatial brain-computer interfaces show promise for enhancing human performance in tasks requiring rapid responses to visual events.
• This proof-of-concept study supports the development of BCIs for improved human-machine interaction.