Real-time brain-computer interface control of walking exoskeleton with bilateral sensory feedback

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Spinal cord injury (SCI) severely impacts ambulation, necessitating advanced assistive technologies.
• Current brain-computer interfaces (BCIs) for gait are unidirectional, lacking crucial sensory feedback.
• Restoring both motor control and sensory perception is vital for functional recovery.

Purpose of the Study:

• To demonstrate a bidirectional brain-computer interface (BDBCI) for simultaneous real-time brain-controlled walking and artificial leg sensation.
• To investigate the use of electrical stimulation of the sensory cortex for artificial leg sensation.
• To establish the feasibility of an embedded BDBCI system for restoring walking function.

Main Methods:

• Recruited epilepsy patients with bilateral interhemispheric subdural electrocorticography (ECoG) implants.
• Utilized motor mapping for leg motor cortex decoding of stepping intent.
• Employed sensory stimulation mapping in the somatosensory cortex for artificial leg percepts.
• Developed a custom embedded BDBCI to control a robotic gait exoskeleton (RGE) and provide sensory feedback.

Main Results:

• Achieved high decoding performance (ρ = 0.92 ± 0.04) for BDBCI-controlled RGE operation.
• Validated bilateral leg percepts with high accuracy (92.8%) in a blind step-counting task.
• Confirmed that stimulation artifacts did not affect decoding performance; no adverse events reported.

Conclusions:

• Established the feasibility of an embedded BDBCI system for restoring both motor control and artificial sensation of walking.
• Demonstrated that leveraging interhemispheric leg sensorimotor cortices is safe and improves decoding.
• Provided a foundation for translating BDBCI technology into fully implantable systems for SCI patients.