Biomechanical differences between able-bodied and spinal cord injured individuals walking in an overground robotic exoskeleton

BACKGROUND: Robotic assisted gait training (RAGT) uses a powered exoskeleton to support an individual’s body and move their limbs, with the aim of activating latent, pre-existing movement patterns stored in the lower spinal cord called central pattern generators (CPGs) to facilitate stepping. The parameters that directly stimulate the stepping CPGs (hip extension and ipsilateral foot unloading) should be targeted to maximise the rehabilitation benefits of these devices. AIM: To compare the biomechanical profiles of individuals with a spinal cord injury (SCI) and able-bodied individuals inside the ReWalk(TM) powered exoskeleton and to contrast the users’ profiles with the exoskeleton. METHODS: Eight able-bodied and four SCI individuals donned a ReWalk(TM) and walked along a 12-meter walkway, using elbow crutches. Whole-body kinematics of the users and the ReWalk(TM) were captured, along with GRF and temporal-spatial characteristics. Discreet kinematic values were analysed using a Kruskall-Wallis H and Dunn’s post-hoc analysis. Upper-body differences, GRF and temporal-spatial characteristics were analysed using a Mann-Whitney U test (P<0.05). RESULTS: Walking speed ranged from 0.32–0.39m/s. Hip abduction, peak knee flexion and ankle dorsiflexion for both the SCI and able-bodied groups presented with significant differences to the ReWalk(TM). The able-bodied group presented significant differences to the ReWalk(TM) for all kinematic variables except frontal plane hip ROM (P = 0.093,δ = -0.56). Sagittal plane pelvic and trunk ROM were significantly greater in the SCI vs. able-bodied (P = 0.004,δ = -1; P = 0.008,δ = -0.94, respectively). Posterior braking force was significantly greater in the SCI group (P = 0.004, δ = -1). DISCUSSION: The different trunk movements used by the SCI group and the capacity for the users’ joint angles to exceed those of the device suggest that biomechanical profiles varied according to the user group. However, upright stepping with the ReWalk(TM) device delivered the appropriate afferent stimulus to activate CPGs as there were no differences in key biomechanical parameters between the two user groups.