Corticomuscular Coupling Analysis in Stroke Rehabilitation Based on Variational Mode Decomposition-Transfer Entropy

Area of Science:

• Neuroscience
• Rehabilitation Medicine
• Biomedical Engineering

Background:

• Stroke significantly impairs motor function by disrupting brain-muscle communication.
• Understanding corticomuscular coupling is crucial for effective stroke rehabilitation strategies.

Purpose of the Study:

• To investigate changes in corticomuscular and cortical coupling during stroke patient rehabilitation.
• To identify neuro-remodeling patterns associated with recovery.
• To evaluate the utility of corticomuscular functional networks (CMFN) as a recovery metric.

Main Methods:

• Variational Modal Decomposition (VMD) applied to electromyography (EMG) data.
• VDM-transfer entropy (VMD-TE) quantified electroencephalogram (EEG)-EMG signal coupling.
• Construction of VMD-TE connection matrices and analysis of small-world network properties.
• Random forest algorithm used for feature extraction across rehabilitation phases.

Main Results:

• Beta waves in EEG identified as primary information carriers between cortex and muscle.
• CMFN in the beta frequency band exhibits small-world network characteristics.
• Observed decreased coupling from the affected motor cortex to muscle, with increased frontal-muscle coupling during rehabilitation.
• Evidence of potential neuro-remodeling in stroke survivors post-rehabilitation.

Conclusions:

• CMFN analysis provides valuable insights into corticomuscular coupling dynamics during stroke recovery.
• Rehabilitation induces shifts in brain-muscle communication pathways.
• Beta-band corticomuscular coupling and network topology are significant indicators of neuroplasticity after stroke.