Topographic Somatosensory Imagery for Real-Time fMRI Brain-Computer Interfacing

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Medicine

Background:

• Real-time functional magnetic resonance imaging (fMRI) is a non-invasive brain-computer interface (BCI) method.
• Current BCIs often rely on visual or motor imagery, which can be difficult for patients with locked-in syndrome.
• Somatosensory imagery is a robust alternative, but its single-trial detection and classification in BCIs were previously unclear.

Purpose of the Study:

• To investigate the reliability and accuracy of single-trial decoding of somatosensory imagery for BCI applications.
• To explore the potential of somatosensory imagery as an information-encoding strategy for BCIs, particularly for individuals with severe impairments.
• To develop and validate a proof-of-concept somatosensory imagery-based fMRI-BCI communication system.

Main Methods:

• Utilized ultra-high field 7-Tesla fMRI to record brain activity during left-foot vs. right-hand somatosensory imagery tasks.
• Applied advanced decoding techniques to analyze single-trial brain responses.
• Developed and tested a binary (yes/no) communication BCI using somatosensory imagery in simulated and real-time conditions.

Main Results:

• Achieved high-accuracy, single-trial decoding of left-foot and right-hand somatosensory imagery.
• Found that greater spatial separation of decoding-weight patterns in the primary somatosensory cortex correlated with higher decoding accuracy.
• Demonstrated successful binary BCI communication using somatosensory imagery in both simulated (7 subjects) and real-time (1 subject) settings.

Conclusions:

• Body part-specific somatosensory imagery differentially activates the somatosensory cortex topographically.
• A novel somatosensory imagery-based fMRI-BCI strategy offers a promising communication method for visually and motor-impaired patients.
• This BCI approach has potential for transfer to lower field MRI, functional near-infrared spectroscopy, and could aid in somatosensory rehabilitation.