Related Experiment Video
Updated: May 10, 2025
04:42
Treating Low Back Pain in Failed Back Surgery Patients with Multicolumn-lead Spinal Cord Stimulation
Published on: June 26, 2018
14.4K
Development and evaluation of a non-invasive brain-spine interface using transcutaneous spinal cord stimulation
Carolyn Atkinson1,2, Lorenzo Lombardi1,2, Meredith Lang1,2
1Biomedical Engineering, Washington University in St. Louis, St. Louis, USA.
Journal of Neuroengineering and Rehabilitation
|April 25, 2025
Summary
This study developed a non-invasive brain-spine interface using EEG and spinal cord stimulation to aid motor rehabilitation. The system shows promise for improving functional recovery in spinal cord injury patients, even those with severe injuries.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Rehabilitation Science
Background:
- Motor rehabilitation aims to restore function after spinal cord injury (SCI), but its effectiveness is limited in individuals with severe sensorimotor deficits.
- Spinal cord stimulation (SCS) offers a potential method to enhance rehabilitation by creating a temporary prosthetic effect.
- Developing non-invasive interfaces is crucial for extending SCS benefits to a wider SCI population.
Purpose of the Study:
- To identify electroencephalography (EEG)-based neural correlates of lower limb movement in the sensorimotor cortex.
- To develop and evaluate a linear discriminant analysis (LDA) decoder for detecting movement onset using these neural signals.
- To assess the performance of a non-invasive brain-spine interface (BSI) combining EEG and transcutaneous spinal cord stimulation (tSCS) for motor rehabilitation.
Main Methods:
- Utilized EEG to record neural activity from unimpaired individuals (N=17) during lower limb movement tasks.
- Developed an LDA decoder to identify movement onset based on event-related desynchronization in specific EEG frequency bands (µ, low β, high β).
- Integrated the EEG decoder with real-time tSCS delivery to modulate stimulation based on detected neural activity.
Main Results:
- Initiation of knee extension was linked to event-related desynchronization in central-medial cortical regions (4-44 Hz).
- The offline LDA decoder achieved an average area under the curve (AUC) of 0.83 ± 0.06 for cued movement.
- Real-time BSI with tSCS demonstrated an AUC of 0.81 ± 0.05 for cued movement and 0.68 ± 0.12 for uncued movement.
Conclusions:
- The developed non-invasive BSI effectively detects neural correlates of lower limb movement, showing potential for motor rehabilitation.
- The system's performance suggests that tSCS can be precisely timed with voluntary effort, potentially enhancing rehabilitation outcomes.
- Further research is needed to understand performance variations between cued and uncued movements and optimize the BSI for diverse SCI conditions.
