A Low-Cost Lower-Limb Brain-Machine Interface Triggered by Pedaling Motor Imagery for Post-Stroke Patients...

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Stroke survivors often experience lower-limb motor deficits, necessitating effective rehabilitation strategies.
• Existing Brain-Machine Interfaces (BMIs) can be costly and computationally intensive.
• Motor imagery (MI) is a cognitive task that can be detected using electroencephalography (EEG) to control external devices.

Purpose of the Study:

• To develop and validate a low-cost, EEG-based BMI for lower-limb motor recovery in post-stroke patients.
• To provide passive pedaling feedback via a Mini-Motorized Exercise Bike (MMEB) controlled by the BMI.
• To assess the system's efficacy and response time in healthy subjects and stroke survivors.

Main Methods:

• Utilized Riemannian geometry for EEG feature extraction.
• Employed Pair-Wise Feature Proximity (PWFP) for feature selection.
• Implemented Linear Discriminant Analysis (LDA) for motor imagery recognition and system control.

Main Results:

• Healthy subjects achieved up to 100% accuracy in triggering the MMEB.
• Post-stroke patients (PS1, PS2) showed improved performance (41.67%, 91.67% accuracy) and reduced latency (2.03s, 1.99s) in the second session.
• The system demonstrated a fast response, creating a closed-loop control experience.

Conclusions:

• The developed low-cost BMI shows promise for lower-limb motor recovery in chronic stroke patients.
• The system facilitates neural relearning and enhances neuroplasticity through motor imagery-based rehabilitation.
• The findings support the potential of this EEG-BMI for accessible and effective stroke rehabilitation.