Discovering individual-specific gait signatures from data-driven models of neuromechanical dynamics

Abstract

Locomotion results from the interactions of highly nonlinear neural and biomechanical dynamics. Accordingly, understanding gait dynamics across behavioral conditions and individuals based on detailed modeling of the underlying neuromechanical system has proven difficult. Here, we develop a data-driven and generative modeling approach that recapitulates the dynamical features of gait behaviors to enable more holistic and interpretable characterizations and comparisons of gait dynamics. Specifically, gait dynamics of multiple individuals are predicted by a dynamical model that defines a common, low-dimensional, latent space to compare group and individual differences. We find that highly individualized dynamics - i.e., gait signatures - for healthy older adults and stroke survivors during treadmill walking are conserved across gait speed. Gait signatures further reveal individual differences in gait dynamics, even in individuals with similar functional deficits. Moreover, components of gait signatures can be biomechanically interpreted and manipulated to reveal their relationships to observed spatiotemporal joint coordination patterns. Lastly, the gait dynamics model can predict the time evolution of joint coordination based on an initial static posture. Our gait signatures framework thus provides a generalizable, holistic method for characterizing and predicting cyclic, dynamical motor behavior that may generalize across species, pathologies, and gait perturbations.

Gait signature changes with walking speed are similar among able-bodied young adults despite persistent individual-specific differences

“…The individual-specific gait signatures discovered by our framework are consistent with the idea that individuals maintain consistent, highly characteristic gait dynamics across a variety of walking speeds [3,4,53] . Our study complements previous research by showing that individuals' gaits remain individual-specific across a wide range of speeds, rather than solely at self-selected walking speeds [1] or a more limited range of speeds [6] . Our cohort of able-bodied adults served as a rigorous test case, demonstrating the robustness of individual classification even in a healthy, young population with relatively similar dynamics.…”

“…To assess the individual discriminatory potential of discrete biomechanical variables across speeds, we conducted a comparable SVM classification approach. Specifically, we focused on 26 widely recognized bilateral kinematic and kinetic discrete variables commonly used in gait analyses [6] , as detailed in Supplementary Table T1. Additionally, we explored the discriminatory capacity of only the 18 kinematic-only and eight kinetic-only variables in distinguishing individuals.…”

“…Understanding what data are needed to differentiate gait coordination patterns between individuals may inform experimental design, equipment considerations, and future clinical translation. Our prior work [6] constructed gait signatures using only sagittal-plane joint kinematics (i.e. joint angles).…”

A machine learning approach to quantify individual gait responses to ankle exoskeletons

“…We employ a simple neural network underpinning more complicated deep learning architectures (Goodfellow et al, 2016). Neural networks have been used to model complex behavior from time-series gait data (LeCun et al, 2015; Rosenberg et al, 2020; Winner et al, 2022), making them an ideal candidate to model exoskeleton responses as discrepancies. Neural networks (i) can learn complex patterns in experimental data that coalesce from diverse sources, (ii) do not make explicit assumptions about physiology and motor control, and (iii) learn time-variant features of exploration.…”

Multi-Site Identification and Generalization of Clusters of Walking Behaviors in Individuals With Chronic Stroke and Neurotypical Controls

“…Walking patterns differ between individuals to the point that we can recognize people by how they walk. 1 These individual differences are more marked after a neurologic injury, such as stroke, 2 due to heterogeneity in stroke lesion type, size, location, and differences in recovery. [3][4][5][6] These individual differences in walking patterns make systematizing treatment prescription for walking rehabilitation interventions a difficult clinical endeavor.…”