MR Ebers, MC Rosenberg, JN Kutz, KM Steele (2023) “A machine learning approach to quantify individual gait responses to ankle exoskeletons”

Journal Article in bioRxiv:

Discrepancy modeling is a unique and innovative tool that complements current biomechanical modeling approaches and may accelerate the discovery of individual-specific mechanisms driving responses to exoskeletons, other assistive devices, and clinical interventions.

Aim: This study aims to leverage a neural network-based discrepancy modeling framework to quantify complex changes in gait in response to passive ankle exoskeletons in non-disabled adults. It hypothesized that (i) the Nominal model would predict Exo kinematics and EMG less accurately than for the Nominal condition, and (ii) the Augmented (Nominal+Discrepancy) model would capture greater variance in Exo kinematics and EMG than the Nominal model.

Method: This study analyzed gait data for 12 non-disabled adults during treadmill walking in bilateral passive ankle exoskeletons at a self-selected speed, results of which were used in participant-specific continuous-time neural network with discrepancy models to predict gait responses.

Results: Discrepancy modeling successfully quantified individuals’ exoskeleton responses without requiring knowledge about physiological structure or motor control. However, additional measurement modalities and/or improved resolution are needed to characterize Exo gait, as the discrepancy may not comprehensively capture response due to unexplained variance in Exo gait.

Interpretation: These techniques can be used to accelerate the discovery of individual-specific mechanisms driving exoskeleton responses, thus enabling personalized rehabilitation.

BC Conner, AM Spomer, KM Steele, ZF Lerner (2022) “Factors influencing neuromuscular responses to gait training with a robotic ankle exoskeleton in cerebral palsy”

Journal Article in RESNA:

Our findings underscored the importance of monitoring how users change their gait kinematics when walking with the resistive device, with a specific emphasis on stance-phase lower limb extension. We also highlight the necessity of considering an individual’s functional status and amount of practice with the device, as well as more obvious factors, like device parameters. BART can be used early in the development of robotic gait training interventions to better understand complex and multifactorial user-device interactions.

Aim: Although ankle exoskeletons offer a promising means of augmenting gait training and enhancing independent mobility among individuals with neuromuscular disorders, response to existing paradigms is highly heterogeneous. In this study we aimed to identify factors which may affect how individuals with cerebral palsy (CP) interact with a resistive ankle exoskeleton during multi-day training to inform future device design and individualized tuning.

Method:We evaluated the gait mechanics (kinematics and muscle activity) of eight individuals with CP as they walked with bilateral ankle exoskeletons – designed to promote increased plantar flexor recruitment – during a seven-day training paradigm. These data along with pertinent device and participant parameters were input into a Bayesian Additive Regression Trees (BART) machine learning model to identify factors which were most associated with increased plantar flexor recruitment.

Results: Four themes emerged: 1) AFO provision is a confusing and lengthy process, 2) participants want more information during AFO provision, 3) AFOs are uncomfortable and difficult to use, and 4) AFOs can benefit mobility and independence. Caregivers and individuals with CP recommended ideas such as 3D printing orthoses and education for caregivers on design choices to improve AFO design and provision.

Interpretation: Individuals with CP and their caregivers found the AFO provision process frustrating but highlight that AFOs support mobility and participation. Further opportunities exist to support function and participation of people with CP by streamlining AFO provision processes, creating educational materials, and improving AFO design for comfort and ease of use.

MC Rosenberg, BS Banjanin, SA Burden, KM Steele (2020) “Predicting walking response to ankle exoskeleton using data driven models”

Journal Article in The Royal Society:

This work highlights the potential of data-driven models grounded in dynamical systems theory to predict complex individualized responses to ankle exoskeletons., without requiring explicit knowledge of the individual’s physiology or motor control

silhouette walking on left with purple lines and projections on right elipsoids and colored spheres

Aim: Evaluate the ability of three classes of subject-specific phase-varying (PV) models to predict kinematic and myoelectric responses to ankle exoskeletons during walking, without requiring prior knowledge of specific user characteristics.

Method: Data from 12 unimpaired adults walking with bilateral passive ankle exoskeletons were captured. PV, linear PV (LPV), and nonlinear PV (NPV) models leveraged Floquet theory to kinematics and muscle activity in response to three exoskeleton torque conditions.

Results: The LPV model’s predictions were more accurate than the PV model when predicting less than 12.5% of a stride in the future and explained 49–70% of the variance in hip, knee and ankle kinematic responses to torque. The LPV model also predicted kinematic responses with similar accuracy to the more-complex NPV model. Myoelectric responses were challenging to predict with all models, explaining at most 10% of the variance in responses.

Interpretation: This work highlights the potential of data-driven PV models to predict complex subject-specific responses to ankle exoskeletons and inform device design and control.

NBC Learn: Exoskeletons and Engineering

NBC Learn logo for the Discovering YOU series - engineer your world. Supported by NSF, Chevron, and ASEE.

We partnered with NBC Learn to share some of our work on exoskeletons to help encourage students to consider a career in engineering. What can be more exciting than musculoskeletal modeling, exoskeletons, horses, and stuffed animals?

Check out the video – a lesson plan will also be posted soon for classrooms to use.

Go team!

Jessica Zistatsis Successfully Defends Her Master’s Thesis

Jessica Zistatsis has successfully defended her Master’s Thesis here at the University of Washington, in Dr. Steele’s Ability & Innovation Lab.

To complete her Master’s in full, Jessica will be submitting and disseminating her thesis, A Passive Pediatric Exoskeleton to Improve the Walking Ability of Children with Neuromuscular Disorders.

To begin watching Jessica’s defense, you may view Part 1 on YouTube HERE, or directly below:

A silly congratulations graphic made by the lab depicting the fictional Godzilla stomping through downtown Seattle, while wearing PlayGait, the pediatric exoskeleton Jessica worked on for her masters. Test within the photo reads, "Congrats on destroying your defense, Jessica!"Our lab could not be more proud! To help celebrate her successful defense, the lab drafted a flyer depicting an empowered Godzilla wearing PlayGait, Jessica’s pediatric exoskeleton. Here’s hoping future children will use their newfound superpowers for good, unlike our friend in this picture!