Open Access Highly Accessed Research

Brain-computer interface controlled robotic gait orthosis

An H Do1*, Po T Wang2, Christine E King2, Sophia N Chun3 and Zoran Nenadic24*

Author Affiliations

1 Department of Neurology, University of California, Irvine, CA, USA

2 Department of Biomedical Engineering, University of California, Irvine, CA, USA

3 Department of Spinal Cord Injury, Long Beach Veterans Affairs Medical Center, Long Beach, CA, USA

4 Department of Electrical Engineering and Computer Science, University of California, Irvine, CA, USA

For all author emails, please log on.

Journal of NeuroEngineering and Rehabilitation 2013, 10:111  doi:10.1186/1743-0003-10-111

Published: 9 December 2013



Excessive reliance on wheelchairs in individuals with tetraplegia or paraplegia due to spinal cord injury (SCI) leads to many medical co-morbidities, such as cardiovascular disease, metabolic derangements, osteoporosis, and pressure ulcers. Treatment of these conditions contributes to the majority of SCI health care costs. Restoring able-body-like ambulation in this patient population can potentially reduce the incidence of these medical co-morbidities, in addition to increasing independence and quality of life. However, no biomedical solution exists that can reverse this loss of neurological function, and hence novel methods are needed. Brain-computer interface (BCI) controlled lower extremity prostheses may constitute one such novel approach.


One able-bodied subject and one subject with paraplegia due to SCI underwent electroencephalogram (EEG) recordings while engaged in alternating epochs of idling and walking kinesthetic motor imagery (KMI). These data were analyzed to generate an EEG prediction model for online BCI operation. A commercial robotic gait orthosis (RoGO) system (suspended over a treadmill) was interfaced with the BCI computer to allow for computerized control. The subjects were then tasked to perform five, 5-min-long online sessions where they ambulated using the BCI-RoGO system as prompted by computerized cues. The performance of this system was assessed with cross-correlation analysis, and omission and false alarm rates.


The offline accuracy of the EEG prediction model averaged 86.30% across both subjects (chance: 50%). The cross-correlation between instructional cues and the BCI-RoGO walking epochs averaged across all subjects and all sessions was 0.812±0.048 (p-value <10−4). Also, there were on average 0.8 false alarms per session and no omissions.


These results provide preliminary evidence that restoring brain-controlled ambulation after SCI is feasible. Future work will test the function of this system in a population of subjects with SCI. If successful, this may justify the future development of BCI-controlled lower extremity prostheses for free overground walking for those with complete motor SCI. Finally, this system can also be applied to incomplete motor SCI, where it could lead to improved neurological outcomes beyond those of standard physiotherapy.