Novel electrotactile brain-computer interface with somatosensory event-related potential based control.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Brain-computer interfaces (BCIs) enable device control via brain signals.
• Non-invasive electroencephalography (EEG) is a common BCI recording method.
• Electrotactile stimulation offers a novel sensory feedback channel for BCIs.

Purpose of the Study:

• To develop and test a novel electrotactile BCI prototype.
• To utilize somatosensory event-related potentials (sERPs) as control signals.
• To implement a tactile attention task as a BCI control paradigm.

Main Methods:

• A custom system integrated EEG, electrotactile stimulation, and processing software.
• A tactile attention task involved distinguishing target from distractor stimuli on the forearm.
• Single-channel EEG data was analyzed using cross-validation for classifier and feature optimization.

Main Results:

• Achieved classification accuracy between 75.1% and 88.1% across ten healthy subjects.
• Identified optimal parameters for sERP extraction, feature selection, and classification.
• Demonstrated the feasibility of a single-channel electrotactile BCI.

Conclusions:

• The novel electrotactile BCI paradigm is feasible for real-time control.
• This BCI approach shows promise for restorative applications (e.g., sensory retraining).
• Potential for assistive BCIs in communication for severely disabled individuals.