Hierarchical Codebook Visually Evoked Potentials for fast and flexible BCIs.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Brain-Computer Interfaces (BCIs) offer direct neural control of devices.
• Current BCIs, especially Codebook Visually Evoked Potentials (cVEP), face limitations in communication speed and stimulus flexibility.
• Existing cVEP applications are primarily limited to spelling tasks.

Purpose of the Study:

• To enhance the flexibility of stimulus presentation in cVEP-based BCIs.
• To investigate a novel hierarchical cVEP (hcVEP) approach using hierarchical codebook vectors and varied color schemes.
• To improve the information transfer rate and reduce response time in non-invasive BCIs.

Main Methods:

• Development of a novel hcVEP approach utilizing hierarchical codebook vectors.
• Implementation of varying color schemes (black/white, green/red) for stimulus presentation.
• An off-line study to evaluate the performance of the hcVEP system.

Main Results:

• The hcVEP approach successfully discriminated target groups within 250 ms with 81% accuracy.
• Final target identification within a group was achieved after 1s with 67% accuracy.
• Different color schemes demonstrated comparable effectiveness, indicating robustness.

Conclusions:

• The proposed hcVEP method significantly increases stimulus flexibility for BCI applications.
• This approach offers a promising avenue for faster and more versatile non-invasive Brain-Computer Interfaces.
• The hcVEP system shows potential for applications beyond traditional spelling tasks, enhancing user interaction.