Brain-Computer Interfaces for AAC: Hybrid SSVEP + Eye-Tracking for Faster Selection Rates

Augmentative and alternative communication systems are essential for people with severe speech and motor impairments, yet many interfaces remain slow, effortful, and error-prone. This study evaluated a hybrid brain-computer interface that combines steady-state visually evoked potentials (SSVEP) with eye-tracking to improve selection speed, reliability, and usability. Participants completed symbol selection, short phrase construction, and correction tasks under three conditions: SSVEP-only, eye-tracking-only, and hybrid SSVEP plus eye-tracking. In the hybrid condition, gaze rapidly narrowed the likely target region, while SSVEP classification confirmed user intent. Outcomes included selection rate, accuracy, errors, correction time, communication throughput, workload, effort, trust, and preference. The hybrid interface achieved the strongest performance, with faster selection than SSVEP-only and better control than eye-tracking-only. It also reduced accidental selections and correction burden, particularly in dense layouts and longer tasks. Results further emphasized personalized gaze thresholds, SSVEP windows, and visual layouts for sustained, practical AAC use across diverse user needs.