| The inability to adequately bend the knee during swing phase of walking, known as Stiff-Knee gait (SKG), commonly occurs after stroke. It is believed other compensatory, energy-consuming kinematics such as pelvic obliquity and hip abduction accompany the "stiff" knee to help clear the foot. Models have shown that improving knee flexion torque before the paretic limb leaves the ground will result in greater knee flexion angle during swing, thus greater foot clearance. If these gait compensations are motivated by foot clearance, then assisting preswing knee flexion torque will reduce gait compensations.;To examine this hypothesis, we needed a lightweight, backdrivable knee actuator. After finding existing actuators insufficient, we developed a concept that remotely controlled deflection of a compliant spring through a sheathed cable. This concept was developed into a knee flexion actuator capable of selectively applying torque during gait. Its performance characteristics made this device, known as SERKA (Series Elastic Remote Knee Actuator), optimally suited to our needs.;Previous work has shown that some robotic assistance reduces its users' voluntary drive. Therefore, we examined the effect of isolated preswing knee flexion assistance over 15 minutes of treadmill walking on non-disabled controls. While the assistance created a significant increase in knee flexion angle, it revealed sparse evidence of adaptation, lacking significant changes in outcomes upon removal of assistance. This lack of adaptation to assistance is advantageous for assistive devices.;We then examined the effect of knee flexion assistance on gait compensations in nine stroke subjects, examining changes in peak knee flexion angle, hip abduction and pelvic obliquity. The results were compared to the controls ran earlier. Stroke subjects significantly increased hip abduction during assistance instead of the predicted decrease, with no change in controls. We infer that this is due to abnormal coordination, which may be a primary cause of gait impairment after stroke. In conclusion, rehabilitation science must equally account for neural issues along with mechanical ones in people with neurological disabilities. We hope this work inspires others to investigate underlying mechanisms of disease states and their modes of neural control en route to rehabilitation and assistive device development. |
