Operate P300 speller when performing other task

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• The P300 speller, a brain-computer interface (BCI), aids motor function restoration but struggles with accuracy during concurrent tasks.
• Maintaining letter recognition accuracy (LRA) in P300 spellers is crucial for practical BCI applications.

Purpose of the Study:

• To implement and validate a dynamic stopping strategy (DSS) for preserving P300 speller LRA during dual-tasking.
• To assess the feasibility of a Bayes-based DSS model in online, real-world BCI scenarios.

Main Methods:

• A dynamic stopping strategy (DSS) was developed using a Bayes-based offline model.
• Simulated dual-task scenarios with varying workloads were used to test the P300 speller system.
• An online P300 speller system was established to evaluate the DSS algorithm's performance.

Main Results:

• The P300 speller with DSS achieved high LRA (96.9%) in dual-tasking, comparable to single-task performance (98.7%).
• DSS dynamically adjusted discriminant confidence based on distraction task workload (r = -0.68).
• The average number of repeated sequences increased significantly (from 4.98 to 6.22) under dual-tasking, compensating for reduced signal-to-noise ratio.

Conclusions:

• The dynamic stopping strategy (DSS) effectively maintains P300 speller performance during concurrent cognitive tasks.
• The DSS model demonstrates robustness and applicability across various dual-task conditions.
• This research supports the real-world implementation of laboratory-developed brain-computer interfaces.