Reducing the Calibration Time in Somatosensory BCI by Using Tactile ERD

Area of Science:

• Neuroscience
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Somatosensory brain-computer interfaces (BCIs) require extensive calibration, often leading to user fatigue and prolonged setup times.
• Existing calibration methods rely on imagined sensations, which can be variable and time-consuming to elicit accurately.

Purpose of the Study:

• To introduce and evaluate a novel tactile-induced-oscillation approach for reducing calibration time in somatosensory BCIs.
• To compare the performance of a new sensory calibration strategy against conventional methods using tactile stimulation.

Main Methods:

• Investigated the similarity between tactile-induced event-related desynchronization (ERD) and imagined sensation-induced ERD.
• Evaluated BCI performance using conventional (imagined data) and novel (sensory stimulation data) calibration strategies.
• Subjects performed imagined sensation tasks while receiving tactile stimuli on their wrists.

Main Results:

• Sensory calibration demonstrated significantly better performance than conventional calibration on an imagined sensation dataset (p=0.0038).
• The novel approach achieved an average accuracy improvement of 5.1% and was 39.3% faster in reaching 70% accuracy.
• Tactile ERD from the sensory cortex proved effective in reducing somatosensory BCI calibration time.

Conclusions:

• The proposed tactile-induced-oscillation method offers an effective strategy for reducing calibration time in somatosensory BCIs.
• This approach is particularly beneficial before BCI use, mitigating fatigue associated with difficult mental tasks.
• Tactile ERD holds promise for BCI applications in patients with intact afferent pathways.