A Calibration-Free Hybrid BCI Speller System Based on High-Frequency SSVEP and sEMG

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Hybrid brain-computer interfaces (hBCI) combining steady-state visual evoked potential (SSVEP) and surface electromyography (sEMG) offer improved performance.
• Existing hBCI systems predominantly use low-frequency SSVEP, leading to significant visual fatigue.
• There is a need for hBCI systems that utilize high-frequency SSVEP to mitigate visual discomfort.

Purpose of the Study:

• To develop and evaluate a novel hBCI speller system utilizing high-frequency SSVEP and sEMG.
• To design a 32-target speller system capable of operating without user-specific calibration.
• To assess the performance and feasibility of the proposed high-frequency SSVEP-based hBCI.

Main Methods:

• A 32-target hBCI speller was designed, dividing targets into two groups, each with distinct high-frequency visual stimuli (31-34.75 Hz).
• Surface electromyography (sEMG) was used for group selection, while high-frequency SSVEP identified targets within groups.
• Signal processing employed filter bank canonical correlation analysis (FBCCA) for SSVEP and root mean square (RMS) for sEMG.

Main Results:

• The developed hBCI system demonstrated high performance in online experiments with 12 healthy subjects.
• Average accuracy reached 93.52 ± 1.66%, and the average information transfer rate (ITR) was 93.50 ± 3.10 bits/min.
• All 12 participants successfully completed free-spelling tasks, indicating practical usability.

Conclusions:

• The proposed hybrid BCI system, integrating high-frequency SSVEP and sEMG, is feasible and practical for BCI speller applications.
• The system effectively reduces visual fatigue associated with low-frequency SSVEP.
• The calibration-free design enhances user convenience and potential for real-world deployment.