Bimanual attempts improving EEG-based hand movement decoding in stroke patients with hemiplegia

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Medicine

Background:

• Brain-computer interfaces (BCIs) using electroencephalography (EEG) aid stroke patient hand motor function rehabilitation.
• Bilateral arm training promotes affected hand recovery.
• Current BCIs often overlook the unaffected hand's potential in stroke rehabilitation.

Purpose of the Study:

• Investigate neural signatures differentiating bimanual and unimanual motor attempts in hemiplegic stroke patients using EEG.
• Determine if these neural differences enhance movement decoding accuracy for BCIs.
• Explore the utility of movement-related cortical potential (MRCP) and event-related desynchronization (ERD) for decoding.

Main Methods:

• Designed an experimental paradigm for unimanual and bimanual hand opening/closing motor attempts.
• Analyzed neural signatures using MRCP and ERD.
• Compared decoding accuracy between unimanual and bimanual movements.

Main Results:

• Greater MRCP activation observed during bimanual movements compared to unimanual.
• Event-related spectral perturbation (ERSP) varied, with ERD being greater for bimanual movements in specific bands/regions.
• Average decoding accuracy for bimanual movements was 4-12% higher than for unimanual movements.

Conclusions:

• Neural signatures differ between bimanual and unimanual movements in stroke patients.
• Bimanual movement decoding offers improved accuracy for BCIs in stroke rehabilitation.
• This research supports advancing BCI-based active hand motor function rehabilitation for stroke survivors.