Characterizing Relationships between Muscle Coordination and Walking Neuromechanics in Stroke Survivors
Huxley, Amelia
Huxley, Amelia
Citations
Altmetric:
License
License
Abstract
We previously demonstrated that machine-learning-based signatures of walking neuromechanics (termed gait signatures) holistically encode post-stroke walking function and biomechanics. Our more recent work suggests that individual differences in post-stroke gait signatures are also related to differences in the complexity of paretic-leg muscle coordination, as defined by the muscle synergies framework. However, it remains unclear how individual differences in muscle coordination complexity in stroke survivors� non-paretic legs are related to differences in their gait signatures. Here, we characterized relationships between non-paretic leg coordination and gait signatures of 43 stroke survivors and 17 able bodied adults at a self-selected speed. We quantified muscle coordination complexity as the number of synergies needed to explain most of the variance in each participant�s electromyography data from eight muscles per leg. We also computed a continuous measure of coordination complexity: the total variance accounted for by one synergy (tVAF1). Gait signatures were computed by stride averaging the individual-specific latent states of a recurrent neural network, trained to model the time-evolution of joint kinematics. Consistent with one prior study, paretic legs had fewer synergies than non-paretic and able-bodied legs, while non-paretic legs had a similar number of synergies as able-bodied legs. Greater tVAF1 in both the paretic (r2=0.12) and non-paretic (r2=0.14) legs was weakly correlated with post-stroke gait signatures being less similar to the averaged able-bodied gait signature. To further characterize multidimensional relationships between muscle coordination and gait signatures, we have developed and verified�and plan to implement�a Partial Least Squares Correlation analysis framework.Advancing our understanding of how muscle coordination impacts neuromechanical gait signatures may be useful for guiding personalized gait rehabilitation for stroke survivors.
Description
Date
1/1/2026
Student Status
Sophomore (Graduating in 2028)
Journal Title
Journal ISSN
Volume Title
Type of presentation
Poster
Collections
Research Projects
Organizational Units
Journal Issue
Citation
DOI
Advisor(s)
Department
Program/Major
Biomedical Engineering
College/School
College of Engineering and Mathematical Sciences
Organization
Research Category
Engineering