Subject combination and electrode selection in cooperative brain-computer interface based on event related...

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Biomedical Engineering

Background:

• Classical Brain-Computer Interface (BCI) systems primarily focus on individual users.
• There is a growing need for BCI paradigms that cater to healthy individuals and collaborative tasks.
• Cooperative BCIs, involving multiple users in a shared system, present a promising alternative.

Purpose of the Study:

• To explore the potential of cooperative BCIs for healthy users.
• To investigate performance improvements by combining data across multiple subjects.
• To compare cooperative BCI performance against traditional single-subject BCI paradigms.

Main Methods:

• Utilized an event-related potentials (ERP) based experiment with synchronized visual stimuli presentation to multiple subjects.
• Evaluated the performance enhancement achieved by combining decisions across subjects.
• Assessed the impact of varying numbers of subjects and electrodes on cooperative BCI performance.

Main Results:

• Combining trials across subjects significantly improves performance compared to single-user BCI systems.
• Achieved a mean Area Under the Curve (AUC) exceeding 0.95 with 10 subjects and 3 electrodes, or 4 subjects and 6 electrodes.
• Demonstrated the feasibility and effectiveness of cooperative BCI decision-making.

Conclusions:

• Cooperative BCIs offer a novel and efficient paradigm for BCI applications, particularly for problem-solving tasks with shared goals.
• The approach shows significant potential for enhancing BCI performance and expanding its utility beyond clinical settings.
• Further research into challenges and applications can leverage cooperative BCIs with current technologies.