Quantifying changes in individual-specific template-based representations of center-of-mass dynamics during walking...

Area of Science:

• Biomechanics
• Robotics
• Neurorehabilitation

Background:

• Ankle exoskeletons influence walking by altering center-of-mass (CoM) motion.
• Understanding and modeling individual-specific CoM dynamics, especially post-neurological injury, is crucial but challenging.
• The impact of ankle exoskeletons on CoM dynamics remains largely unknown.

Purpose of the Study:

• To evaluate individual-specific changes in CoM dynamics with passive ankle exoskeletons in unimpaired adults and a post-stroke individual.
• To identify optimal, physically interpretable mechanisms (template signatures) describing CoM dynamics using data-driven methods.
• To investigate the robustness of CoM dynamics to passive ankle exoskeletons and the effects of neurological injury.

Main Methods:

• Utilized hybrid sparse identification of nonlinear dynamics (Hybrid-SINDy), an equation-free, data-driven technique.
• Collected walking data from unimpaired adults and one individual with post-stroke hemiparesis using shoes-only and passive ankle exoskeletons (zero-stiffness and high-stiffness).
• Applied Hybrid-SINDy to infer sparse hybrid dynamics and identify template signatures for CoM motion.

Main Results:

• Hybrid-SINDy identified spring-loaded inverted pendulum-like template signatures in unimpaired adults, which were robust to exoskeleton use (except for minor leg resting length changes).
• In the post-stroke individual, paretic-leg rotary stiffness mechanisms significantly increased (37-50%) with zero-stiffness exoskeletons.
• CoM dynamics in unimpaired individuals showed resilience to passive ankle exoskeletons, unlike the observed changes in the post-stroke participant.

Conclusions:

• Individual-specific CoM dynamics appear robust to passive ankle exoskeletons in unimpaired adults.
• Neurological injuries may alter how ankle exoskeletons impact CoM dynamics, warranting further investigation.
• Hybrid-SINDy is a promising tool for discovering mechanisms of individual-specific CoM dynamics when using assistive devices.