Spinal Cord Stimulation Combined With Motor Imagery Brain-Computer Interface for Chronic Post-Stroke Upper Limb Motor Dysfunction

Upper limb motor dysfunction is a common and disabling sequela of stroke. Many patients enter a chronic phase more than 6 months after stroke onset, during which spontaneous recovery and conventional rehabilitation-related improvement often reach a plateau. Motor imagery brain-computer interface rehabilitation can decode motor intention from electroencephalographic signals and provide closed-loop feedback through external devices, thereby promoting cortical reorganization. However, in patients with impaired corticospinal pathways and insufficient residual motor execution capacity, the efficacy of motor imagery brain-computer interface training alone may be limited.

Spinal cord stimulation may facilitate spinal motor circuits, reduce abnormal muscle tone, and improve the excitability of residual descending motor pathways. Combining spinal cord stimulation with motor imagery brain-computer interface training may provide a synergistic central-peripheral neuromodulation strategy. The brain-computer interface decodes motor intention from the central nervous system, while spinal cord stimulation facilitates peripheral motor pathway execution, potentially enhancing motor recovery and neuroplasticity.

Participants will be assigned, according to patient preference and investigator assessment, to either the experimental group or the control group. The experimental group will undergo spinal cord stimulation implantation followed by individualized stimulation programming and standardized motor imagery brain-computer interface rehabilitation training. The control group will receive the same frequency and duration of motor imagery brain-computer interface rehabilitation training without spinal cord stimulation implantation. Clinical outcomes will be assessed at baseline, after 4 weeks of intervention, 2 months after intervention, and 3 months after intervention. Safety events will be recorded throughout the study. Exploratory assessments will include electroencephalography and neuroimaging to investigate potential mechanisms of neuroplasticity.

• Spinal Cord Stimulation (Device): Spinal cord stimulation will be delivered through epidural electrodes implanted at cervical spinal cord levels, typically C3-C7 for upper limb dysfunction. Stimulation parameters will be individually optimized within clinically safe and device-permitted ranges, including frequency, pulse width, amplitude, electrode configuration, and stimulation mode.
• Motor Imagery Brain-Computer Interface Rehabilitation Training (Device): Motor imagery brain-computer interface training will use a 64-channel medical-grade electroencephalography cap to acquire scalp EEG signals. Participants will perform motor imagery tasks involving the affected upper limb, such as grasping, elbow extension, or wrist lifting. Sensorimotor rhythm features, especially mu rhythm and beta rhythm event-related desynchronization, will be extracted in real time. When significant event-related desynchronization is detected, the system will trigger external feedback, such as a soft robotic glove or functional electrical stimulation, to assist the affected limb in completing the target movement.

Primary outcomes

• Change in Fugl-Meyer Assessment for Upper Extremity Score [Baseline, Week 4, Week 8, and Week 12]: The Fugl-Meyer Assessment for Upper Extremity will be used to evaluate motor function recovery of the affected upper limb.

Secondary outcomes

• Change in Modified Ashworth Scale Score [Baseline, Week 4, Week 8, and Week 12]: The Modified Ashworth Scale will be used to assess muscle tone and spasticity of the affected upper limb.
• Change in Action Research Arm Test Score [Baseline, Week 4, Week 8, and Week 12]: The Action Research Arm Test will be used to assess functional activity of the affected upper limb, including grasp, grip, pinch, and gross movement.
• Change in Modified Barthel Index Score [Baseline, Week 4, Week 8, and Week 12]: The Modified Barthel Index will be used to assess activities of daily living.
• Incidence of Adverse Events and Serious Adverse Events [From enrollment to Week 12]: All adverse events and serious adverse events will be recorded and assessed throughout the study. Spinal cord stimulation-related adverse events may include intraoperative or postoperative bleeding, infection, cerebrospinal fluid leakage, electrode migration or fracture, implant rejection, postoperative pain, and neurological injury. Motor imagery brain-computer interface-related adverse events may include dizziness, visual fatigue, skin allergy related to electrode gel, training-related fatigue, and other discomfort.

Other outcomes

• Change in Electroencephalographic Indicators [Baseline, Week 4, and Week 12]: Resting-state and motor imagery task-state 64-channel electroencephalography will be collected to analyze event-related desynchronization/synchronization, phase-lag-index-based functional connectivity, and graph-theoretical topological properties.
• Change in Functional Magnetic Resonance Imaging and Diffusion Tensor Imaging Indicators [Baseline and Week 12]: Resting-state functional magnetic resonance imaging and diffusion tensor imaging will be performed to assess changes in functional connectivity of the default mode network and sensorimotor network, as well as fractional anisotropy and mean diffusivity of major white matter tracts, including the corticospinal tract.

Sex: All, ages: 18 Years to 75 Years

Inclusion Criteria:

* Age 18 to 75 years.
* First-ever unilateral supratentorial stroke, either ischemic or hemorrhagic, confirmed by computed tomography or magnetic resonance imaging, resulting in hemiparesis, with disease duration longer than 6 months.
* At least one active movement in the wrist or fingers of the affected upper limb, with muscle strength of grade 1 or higher.
* Fugl-Meyer Assessment for Upper Extremity score between 10 and 40, indicating moderate upper limb motor impairment.
* Change in Fugl-Meyer Assessment score less than 10% within the previous month, indicating a functional plateau.
* Clear consciousness and basically normal cognitive function, with Mini-Mental State Examination score of 24 or higher.
* Stable clinical condition and ability to understand and cooperate with simple instructions and rehabilitation training.
* Written informed consent voluntarily signed by the participant or legal guardian.

Exclusion Criteria:

* Other neurological diseases that may cause motor dysfunction, such as Parkinson's disease, multiple sclerosis, or spinal cord injury.
* Severe visual or auditory impairment that prevents cooperation with visual or auditory feedback instructions of the brain-computer interface system.
* Contraindications to spinal cord stimulation surgery, such as severe coagulation dysfunction, infection at the puncture site, severe spinal deformity, or spinal canal stenosis.
* History of epilepsy, intracranial metal implants, cardiac pacemaker, or other contraindications to magnetic resonance imaging.
* Previous neuromodulation surgery for hemiparesis, such as spinal cord stimulation or deep brain stimulation.
* Pregnancy or lactation.
* Any other condition judged by the investigator to make the participant unsuitable for this study.