Neural decoding of unilateral upper limb movements using single trial MEG signals

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Technology

Background:

• Brain machine interfaces (BMIs) offer potential for restoring function in individuals with motor impairments.
• Non-invasive methods for decoding complex motor intentions, particularly for upper limb movements, require further investigation.

Purpose of the Study:

• To assess the decoding accuracy of three distinct unilateral upper limb movements using non-invasive magnetoencephalography (MEG).
• To explore the temporal dynamics of decoding accuracy relative to movement onset.

Main Methods:

• Recorded neuromagnetic activity from 9 healthy subjects performing hand grasping, pinching, and elbow flexion.
• Utilized a support vector machine classifier on single-trial MEG signals.
• Analyzed decoding accuracy across different time windows relative to movement onset.

Main Results:

• Achieved an average decoding accuracy of 66 ± 10% for classifying three upper limb movements.
• Identified significant peaks in decoding accuracy before and after movement onsets.
• Localized decoding-related brain activity to parietal and sensorimotor areas.

Conclusions:

• Non-invasive MEG signals can accurately decode distinct unilateral upper limb movements.
• The findings support the potential of BMIs for controlling assistive devices.
• Understanding the temporal patterns of brain activity is crucial for improving BMI performance.