A robotic knee exoskeleton for walking assistance and connectivity topology exploration in EEG signal

Disability of lower limb affects voluntary movements and results in a low quality of life. Although people with disabled lower limb can restore their movements through assistive devices like wheelchair, such manner is not as natural as that healthy people perform locomotion. It would be better to regain movements by restoring the function of lower limb itself. Hence, an orthosis that can be worn on the disable lower limb might be a good choice. In addition, it is of interest to explore brain activities during walking with and without an orthosis. In this paper, we introduced a robotic knee exoskeleton that can provide assistive torque to facilitate walking. Three different conditions were employed here, namely free walking (FW, without exoskeleton), zero force (ZF, with the exoskeleton but no torque was provided), and assistive force (AF, torque was provided by the exoskeleton). During the walking, electrophysiological signals were simultaneously recorded. Partial directed coherence (PDC) was employed to measure connectivity strengths between channels and the obtained effective connectivity network was quantitatively assessed through a standard graph theoretical analysis framework. We found that the clustering coefficient was significantly increased when assistive torque was provided by the exoskeleton compared with FW and ZF. A decrease in path length was also found when assistive torque was provided. Our findings demonstrated that topological patterns of brain activities were distinctly different when people received assistive torque during walking. This study could be of meaningful significance on further orthosis development from the perspective of brain electrophysiological activity and insight into the understanding for brain plasticity and neural rehabilitation.