Targeting post-stroke walking automaticity with a propulsion-augmenting soft robotic exosuit: toward a biomechanical and neurophysiological approach to assistance prescription

Abstract

Human locomotor control ranges on a spectrum of automaticity, from highly automatic strategies that require minimal cognitive input, to attention-demanding executive-control strategies. The neural circuitry that facilitates automaticity is impaired by stroke, resulting in a compensatory shift toward executive-control, as well as reduced paretic propulsion and increased step-to-step variability. We have developed a soft robotic exosuit to augment paretic propulsion by providing paretic plantarflexor assistance during the propulsive phase of walking. For this preliminary study, we hypothesized that changes in walking automaticity would accompany changes in paretic propulsion. When plantarflexor assistance timings were tuned to reduce propulsion variability-a biomechanical measure of automaticity-a -14.7±2.5 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup> variability reduction was accompanied by increased paretic propulsion (%Δ+6.4±6.3%) and prefrontal cortex activity (A oxygenated hemoglobin:+ 1.08E-04±1.05E-04 M mm). When plantarflexor assistance timings were instead tuned to reduce prefrontal cortex activity-a neurophysiological measure of automaticity-a -1.3E-05±1.1E-05 M mm decrease in oxygenated hemoglobin was accompanied by both increased paretic propulsion (%Δ:+4.4±8.1 %) and reduced propulsion variability (%Δ:-3.7±19.3%). Biomechanical and neurophysiological measures of automaticity are sensitive to exosuit assistance timing changes, but are differentially affected, highlighting the need for individualized tuning.