High accuracy decoding of user intentions using EEG to control a lower-body exoskeleton

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Brain-Machine Interface (BMI) systems traditionally rely on invasive methods for signal acquisition.
• Decoding user intentions is crucial for controlling assistive devices like robotic manipulators.
• Non-invasive electroencephalography (EEG) offers a less intrusive approach to brain signal measurement.

Purpose of the Study:

• To develop and evaluate a non-invasive Brain-Machine Interface (BMI) for enabling paraplegic individuals to walk using a lower-body exoskeleton.
• To decode a paraplegic subject's motion intentions from electroencephalography (EEG) signals.
• To implement a real-time closed-loop system for exoskeleton control.

Main Methods:

• Utilized a lower-body exoskeleton controlled by brain activity.
• Acquired brain signals using non-invasive electroencephalography (EEG).
• Developed a novel decoding method for motion intentions and implemented a closed-loop system (NeuroRex).

Main Results:

• Achieved high offline evaluation accuracies for motion intention decoding (approximately 98%).
• Demonstrated a closed-loop implementation with a short on-site training time (around 38 seconds).
• Presented preliminary positive results from real-time closed-loop implementation with a paraplegic test subject.

Conclusions:

• The developed non-invasive BMI system shows significant promise for restoring mobility in paraplegic individuals.
• The novel decoding method and efficient closed-loop implementation facilitate practical application of exoskeleton-assisted walking.
• Further research and development of the NeuroRex system could greatly enhance the quality of life for people with lower-limb paralysis.