A Closed-loop Brain Computer Interface to a Virtual Reality Avatar: Gait Adaptation to Visual Kinematic Perturbations

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Science

Background:

• Human bipedal locomotion control is crucial for lower-body brain-computer interfaces (BCIs) in gait rehabilitation.
• While closed-loop BCI for exoskeleton control is feasible, multi-day closed-loop neural decoding for gait in virtual reality (BCI-VR) remains undemonstrated.

Purpose of the Study:

• To propose and demonstrate a real-time closed-loop BCI system for decoding lower limb joint angles from scalp electroencephalography (EEG).
• To control virtual avatar walking movements and investigate gait adaptation using BCI-VR over eight days with kinematic perturbations.

Main Methods:

• Real-time closed-loop BCI system decoding lower limb joint angles from EEG during treadmill walking.
• Control of a virtual avatar's walking movements in a virtual reality environment.
• Introduction of virtual kinematic perturbations to induce asymmetric gait patterns and assess adaptation.

Main Results:

• Demonstrated feasibility of a closed-loop BCI for controlling a walking avatar under normal and perturbed conditions.
• Observed cortical adaptations in response to altered visuomotor feedback within the BCI-VR system.
• Successful multi-day closed-loop neural decoding and control of gait-related movements.

Conclusions:

• The developed BCI-VR system is feasible for learning to control walking avatars, even with altered visuomotor feedback.
• Findings support the potential application of BCI-VR systems for post-stroke gait rehabilitation.
• The study provides insights into cortical plasticity induced by closed-loop BCI systems.