Impact of wearing a robotic exoskeleton on gait parameters and muscle activity in individuals with incomplete spinal cord injury



Journal Title

Journal ISSN

Volume Title



Robotic exoskeleton devices, such as the Ekso GTTM robotic exoskeleton (EKSO), enable individuals with lower extremity weakness to stand up and walk over ground. Research relevant to the effects of the EKSO on gait parameters and muscle activity in patients with incomplete spinal cord injury (SCI) is limited. Therefore, the purpose of this study was to evaluate whether people with incomplete SCI would walk differently when they wore the EKSO. Specifically, the temporospatial gait parameters as well as kinematics and muscle activity of the lower extremities during level walking were compared between two conditions: with and without wearing the EKSO. SUBJECTS: Ten ambulatory adults (age: 39.3 ± 11.7 years, 9 men and 1 woman) with incomplete SCI completed the study. Average time since injury was 7 ± 5.5 years with a Walking Index for Spinal Cord Injury II (WISCI II) score of 14.5±2.8. METHODS: A 10-camera motion analysis system, synchronized with a surface electromyographic (EMG) unit, was used to obtain temporospatial gait parameters, range of motions (ROMs) of hip flexion-extension, knee flexion-extension and ankle dorsiflexion-plantarflexion, and muscle activity of the lower extremities. Each participant performed walking under two conditions: with and without wearing the EKSO. RESULTS: There were significant differences between the two conditions in gait speed (p = 0.006, no EKSO: 0.56 ± 0.32m/s, EKSO: 0.20 ± 0.03m/s), stride length (p = 0.001, no EKSO: 1.04 ± 0.24m, EKSO: 0.65 ± 0.07m), step length (p = 0.001, no EKSO: 0.51 ± 0.12m, EKSO: 0.33 ± 0.03m) and swing time (p = 0.006, no EKSO: 0.61 ± 0.17s, EKSO: 0.80 ± 0.13), but not in double-limb-support time (p = 0.474, no EKSO: 0.73 ± 0.75s, EKSO: 0.90 ± 0.25s) and stance time (p = 0.413, no EKSO: 2.03 ± 01.66s, EKSO: 2.44 ± 0.42s). When wearing the EKSO, ankle dorsiflexion-plantarflexion ROM was significantly reduced during the stance phase (p = 0.020, no EKSO: 35.90 ± 26.18, EKSO: 11.34 ± 4.55), but there was no difference between conditions in the knee (p = 0.211, no EKSO: 23.92 ± 9.87, EKSO: 27.58 ± 0.4.99) or hip (p = 0.425, no EKSO: 34.14 ± 8.04, EKSO: 32.25 ± 2.64) ROMs. In addition, there were no significant differences between conditions in all three ROMs in the swing phase. Lastly, there were no significant differences in EMG activity of the lower extremities between conditions in both the stance and swing phases. CONCLUSION: Participants walked slower with shorter stride length when walking with a robotic exoskeleton. Although kinematics were unaffected in hip and knee ROM, ankle dorsiflexion-plantarflexion ROM was decreased primarily due to the rigid design of the ankle joint in the EKSO, which limited ankle motions. The finding of no significant change of muscle activity when wearing the EKSO could be due to the reduced walking speed and the heterogeneity of the participants.



Spinal cord injury, Exoskeleton, Robotic