Conventional P300 BCIs rely on multi-channel EEG acquisition, which makes setup cumbersome, preparation lengthy and the user workload high, limiting real-world use. The team proposes a fixed few-channel selection framework based on a heuristic algorithm, combining a genetic algorithm with Bayesian linear discriminant analysis to identify a strongly generalizable subset of channels from a traditional eight-channel system, avoiding costly subject-specific recalibration.
In total, 48 healthy subjects completed offline and online virtual reality experiments. The three-channel system maintained accuracy and information transfer rates comparable to the eight-channel system with no significant degradation, the authors report, while equipment preparation time fell by 90%. NASA-TLX workload assessments confirmed a significant reduction in users’ psychological and physical burden. The researchers say the few-channel strategy preserves core interaction performance while improving user experience and practicality, supporting lightweight, user-friendly BCI systems.