Adapted Assistance and Resistance Training With a Knee Exoskeleton After Stroke

Abstract

Studies on robotic interventions for gait rehabilitation after stroke require: (i) rigorous performance evidence; (ii) systematic procedures to tune the control parameters; and (iii) combination of control modes. In this study, we investigated how stroke individuals responded to training for two weeks with a knee exoskeleton (ABLE-KS) using both Assistance and Resistance training modes together with auditory feedback to train peak knee flexion angle. During the training, the torque provided by the ABLE-KS and the biofeedback were systematically adapted based on the subject’s performance and perceived exertion level. We carried out a comprehensive experimental analysis that evaluated a wide range of biomechanical metrics, together with usability and users’ perception metrics. We found significant improvements in peak knee flexion ( ${p} = {0}.{0016}$ ), minimum knee angle during stance ( ${p} = {0}.{0053}$ ), paretic single support time ( ${p} = {0}.{0087}$ ) and gait endurance ( ${p} = {0}.{022}$ ) when walking without the exoskeleton after the two weeks of training. Participants significantly ( ${p} < {0}.{00025}$ ) improved the knee angle during the stance and swing phases when walking with the exoskeleton powered in the high Assistance mode in comparison to the No Exo and the Unpowered conditions. No clinically relevant differences were found between Assistance and Resistance training sessions. Participants improved their performance with the exoskeleton (24-55 %) for the peak knee flexion angle throughout the training sessions. Moreover, participants showed a high level of acceptability of the ABLE-KS (QUEST 2.0 score: 4.5 ± 0.3 out of 5). Our preliminary findings suggest that the proposed training approach can produce similar or larger improvements in post-stroke individuals than other studies with knee exoskeletons that used higher training intensities.

Keywords

• Rehabilitation robotics
• prosthetics and exoskeletons
• wearable robotics