From spinal central pattern generators to cortical network: integrated BCI for walking rehabilitation

Area of Science:

• Neuroscience
• Rehabilitation Engineering
• Biomedical Engineering

Background:

• Locomotion is primarily controlled by spinal mechanisms, but the brain monitors and informs central pattern generation.
• Upper and lower limb coordination is crucial for effective locomotion control.
• Brain-computer interfaces (BCIs) offer a promising avenue for improving locomotor rehabilitation outcomes.

Purpose of the Study:

• To critically investigate various neurophysiological signal-based approaches for controlling assistive exoskeletons in locomotor rehabilitation.
• To explore the integration of electroencephalogram (EEG), electromyogram (EMG), and hybrid signals for exoskeleton control.
• To examine the role of programmable central pattern generators (PCPGs) and dynamic recurrent neural networks (DRNNs) in locomotion control.

Main Methods:

• Utilizing electroencephalogram (EEG) and upper limb electromyogram (EMG) signals, individually or combined.
• Controlling assistive exoskeletons for locomotion using PCPGs or DRNNs.
• Employing plantar surface tactile stimulation and virtual reality to simulate walking sensations in a supine position.

Main Results:

• Different neurophysiological signal processing methods show potential for controlling assistive exoskeletons.
• Hybrid EEG-EMG approaches may offer enhanced control over locomotion-assisting devices.
• Virtual reality and tactile stimulation can facilitate the generation of relevant brain signals for gait training.

Conclusions:

• BCI-driven assistive exoskeletons, controlled by EEG, EMG, or hybrid signals, can significantly improve locomotor rehabilitation.
• These technologies leverage brain plasticity to aid neurorehabilitation for conditions like stroke, spinal trauma, MS, and cerebral palsy.
• The integration of PCPGs, DRNNs, VR, and tactile stimulation presents a multi-faceted approach to gait recovery.