Collaborative Brain-Computer Interfaces to Enhance Group Decisions in an Outpost Surveillance Task

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Cognitive Science

Background:

• Group decision-making under time constraints is challenging.
• Existing methods for aggregating group decisions often lack optimal confidence weighting.
• Brain-Computer Interfaces (BCIs) offer potential for real-time cognitive state assessment.

Purpose of the Study:

• To develop and evaluate a two-layered collaborative BCI (cBCI) for improving group decisions in time-sensitive surveillance tasks.
• To integrate response times and electroencephalography (EEG) neural features for refined decision confidence estimation.
• To enhance both the accuracy and speed of group decision-making compared to conventional approaches.

Main Methods:

• A novel two-layered cBCI architecture was implemented.
• Response times (RTs) were used to estimate initial decision confidence.
• Neural features from EEG were extracted to refine confidence estimates.
• Refined confidence scores were used to weigh individual responses for group decision aggregation.

Main Results:

• cBCI-assisted groups demonstrated significantly higher decision accuracy than groups using majority or reported confidence.
• The cBCI approach led to faster group decision-making processes.
• The refined confidence measure showed better correlation with decision correctness.

Conclusions:

• The proposed two-layered cBCI effectively enhances group decision-making accuracy and efficiency.
• Integrating RTs and EEG-based confidence estimation is a promising strategy for cBCI applications.
• This cBCI represents a novel and effective tool for time-constrained group decision support in realistic settings.