Spasticity is a common comorbidity associated with spinal cord injury (SCI). Robotic exoskeletons have recently emerged to facilitate legged mobility in people with motor complete SCI. Involuntary muscle activity attributed to spasticity, however, can prevent such individuals from using an exoskeleton. Specifically, although most exoskeleton technologies can accommodate low to moderate spasticity, the presence of moderate to severe spasticity can significantly impair gait kinematics when using an exoskeleton. In an effort to potentially enable individuals with moderate to severe spasticity to use exoskeletons more effectively, this study investigates the use of common peroneal stimulation in conjunction with exoskeleton gait assistance. The electrical stimulation is timed with the exoskeleton swing phase, and is intended to acutely suppress extensor spasticity through recruitment of the flexion withdrawal reflex (i.e., while the stimulation is activated) to enable improved exoskeletal walking. In order to examine the potential efficacy of this approach, two SCI subjects with severe extensor spasticity (i.e., modified Ashworth ratings of three to four) walked in an exoskeleton with and without supplemental stimulation while knee and hip motion was measured during swing phase. Stimulation was alternated on and off every ten steps to eliminate transient therapeutic effects, enabling the acute effects of stimulation to be isolated. These experiments indicated that common peroneal stimulation on average increased peak hip flexion during the swing phase of walking by 21.1° (236%) and peak knee flexion by 14.4° (56%). Additionally, use of the stimulation decreased the swing phase RMS motor current by 228 mA (15%) at the hip motors and 734 mA (38%) at the knee motors, indicating improved kinematics were achieved with reduced effort from the exoskeleton. Walking with the exoskeleton did not have a significant effect on modified Ashworth scores, indicating the common peroneal stimulation has only acute effects on suppressing extensor tone and aiding flexion. This preliminary data indicates that such supplemental stimulation may be used to improve the quality of movement provided by exoskeletons for persons with severe extensor spasticity in the lower limb.

This paper describes a controller for a lower-limb exoskeleton that enables variable-geometry stair ascent and descent for persons with lower limb paralysis. The controller was evaluated on a subject with T10 complete spinal cord injury (SCI) on two staircases, one with a riser height and tread depth of 18.4 × 27.9 cm (7.25 × 11 in) and the other 17.8 × 29.8 cm (7 × 11.75 in). The controller enabled ascent and descent of both staircases without explicit tuning for each, and with an average step rate of 12.9 step/min during ascent and 14.6 step/min during descent.

This paper describes a control approach for a lower limb exoskeleton intended to enable stair ascent and descent of variable geometry staircases for individuals with paraplegia resulting from spinal cord injury (SCI). To assess the efficacy of ascent and descent functionality provided by the control approach, the controller was implemented in a lower limb exoskeleton and tested in experimental trials on three subjects with motor-complete SCI on three staircases of varying geometry. Results from the assessments indicate that subjects were able to capably ascend and descend step heights varying from 7.6 to 16.5 cm without changing control settings; the controller provided for step time consistency highly representative of healthy subjects (9.2% variation in exoskeleton step time, relative to 7.7% variation in healthy subjects); and the exoskeleton provided peak joint torques on average 110% and 74% of the healthy-subject peak joint torques during stair ascent and descent, respectively. Subject perceived exertion during the stair ascent and descent activities was rated between “light” and “very light.”