Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics

Gait impairment post-stroke is highly heterogeneous. Prior studies classified heterogeneous gait patterns into subgroups using peak kinematics, kinetics, or spatiotemporal variables. A limitation of this approach is the need to select discrete features in the gait cycle. Using continuous gait cycle...

Full description

Saved in:
Bibliographic Details
Main Authors: Andrian Kuch, Nicolas Schweighofer, James M. Finley, Alison McKenzie, Yuxin Wen, Natalia Sanchez
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Transactions on Neural Systems and Rehabilitation Engineering
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10994322/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850265744902717440
author Andrian Kuch
Nicolas Schweighofer
James M. Finley
Alison McKenzie
Yuxin Wen
Natalia Sanchez
author_facet Andrian Kuch
Nicolas Schweighofer
James M. Finley
Alison McKenzie
Yuxin Wen
Natalia Sanchez
author_sort Andrian Kuch
collection DOAJ
description Gait impairment post-stroke is highly heterogeneous. Prior studies classified heterogeneous gait patterns into subgroups using peak kinematics, kinetics, or spatiotemporal variables. A limitation of this approach is the need to select discrete features in the gait cycle. Using continuous gait cycle data, we accounted for differences in magnitude and timing of kinematics. Here, we propose a machine-learning pipeline combining supervised and unsupervised learning. We first trained a Convolutional Neural Network and a Temporal Convolutional Network to extract features that distinguish impaired from neurotypical gait. Then, we used unsupervised time-series k-means and Gaussian Mixture Models to identify gait clusters. We tested our pipeline using kinematic data of 28 neurotypical and 39 individuals post-stroke. We assessed differences between clusters using ANOVA. We identified two neurotypical gait clusters (C1, C2). C1: normative gait pattern. C2: shorter stride time. We observed three post-stroke gait clusters (S1, S2, S3). S1: mild impairment and increased bilateral knee flexion during loading response. S2: moderate impairment, slow speed, short steps, increased knee flexion during stance bilaterally, and reduced paretic knee flexion during swing. S3: mild impairment, asymmetric swing time, increased ankle abduction during the gait cycle, and reduced dorsiflexion bilaterally. Our results indicate that joint kinematics post-stroke are mostly distinct from controls, and highlight kinematic impairments in the non-paretic limb. The post-stroke clusters showed distinct impairments that would require different interventions, providing additional information for clinicians about rehabilitation targets.
format Article
id doaj-art-ea2b8fa432ff44c188b24363839288ef
institution OA Journals
issn 1534-4320
1558-0210
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Transactions on Neural Systems and Rehabilitation Engineering
spelling doaj-art-ea2b8fa432ff44c188b24363839288ef2025-08-20T01:54:20ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1534-43201558-02102025-01-01331927193810.1109/TNSRE.2025.356832510994322Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle KinematicsAndrian Kuch0https://orcid.org/0000-0002-8122-4529Nicolas Schweighofer1https://orcid.org/0000-0003-3362-6088James M. Finley2https://orcid.org/0000-0003-2679-2221Alison McKenzie3Yuxin Wen4https://orcid.org/0000-0002-2352-5622Natalia Sanchez5https://orcid.org/0000-0002-6467-1781Department of Physical Therapy, Chapman University, Orange, CA, USADivision of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles, CA, USADivision of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles, CA, USADepartment of Physical Therapy, Chapman University, Orange, CA, USAFowler School of Engineering, Chapman University, Orange, CA, USADepartment of Physical Therapy and the Fowler School of Engineering, Chapman University, Orange, CA, USAGait impairment post-stroke is highly heterogeneous. Prior studies classified heterogeneous gait patterns into subgroups using peak kinematics, kinetics, or spatiotemporal variables. A limitation of this approach is the need to select discrete features in the gait cycle. Using continuous gait cycle data, we accounted for differences in magnitude and timing of kinematics. Here, we propose a machine-learning pipeline combining supervised and unsupervised learning. We first trained a Convolutional Neural Network and a Temporal Convolutional Network to extract features that distinguish impaired from neurotypical gait. Then, we used unsupervised time-series k-means and Gaussian Mixture Models to identify gait clusters. We tested our pipeline using kinematic data of 28 neurotypical and 39 individuals post-stroke. We assessed differences between clusters using ANOVA. We identified two neurotypical gait clusters (C1, C2). C1: normative gait pattern. C2: shorter stride time. We observed three post-stroke gait clusters (S1, S2, S3). S1: mild impairment and increased bilateral knee flexion during loading response. S2: moderate impairment, slow speed, short steps, increased knee flexion during stance bilaterally, and reduced paretic knee flexion during swing. S3: mild impairment, asymmetric swing time, increased ankle abduction during the gait cycle, and reduced dorsiflexion bilaterally. Our results indicate that joint kinematics post-stroke are mostly distinct from controls, and highlight kinematic impairments in the non-paretic limb. The post-stroke clusters showed distinct impairments that would require different interventions, providing additional information for clinicians about rehabilitation targets.https://ieeexplore.ieee.org/document/10994322/Clustering methodskinematicsstrokemachine learningneural networksgait analysis
spellingShingle Andrian Kuch
Nicolas Schweighofer
James M. Finley
Alison McKenzie
Yuxin Wen
Natalia Sanchez
Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics
IEEE Transactions on Neural Systems and Rehabilitation Engineering
Clustering methods
kinematics
stroke
machine learning
neural networks
gait analysis
title Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics
title_full Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics
title_fullStr Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics
title_full_unstemmed Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics
title_short Identification of Subtypes of Post-Stroke and Neurotypical Gait Behaviors Using Neural Network Analysis of Gait Cycle Kinematics
title_sort identification of subtypes of post stroke and neurotypical gait behaviors using neural network analysis of gait cycle kinematics
topic Clustering methods
kinematics
stroke
machine learning
neural networks
gait analysis
url https://ieeexplore.ieee.org/document/10994322/
work_keys_str_mv AT andriankuch identificationofsubtypesofpoststrokeandneurotypicalgaitbehaviorsusingneuralnetworkanalysisofgaitcyclekinematics
AT nicolasschweighofer identificationofsubtypesofpoststrokeandneurotypicalgaitbehaviorsusingneuralnetworkanalysisofgaitcyclekinematics
AT jamesmfinley identificationofsubtypesofpoststrokeandneurotypicalgaitbehaviorsusingneuralnetworkanalysisofgaitcyclekinematics
AT alisonmckenzie identificationofsubtypesofpoststrokeandneurotypicalgaitbehaviorsusingneuralnetworkanalysisofgaitcyclekinematics
AT yuxinwen identificationofsubtypesofpoststrokeandneurotypicalgaitbehaviorsusingneuralnetworkanalysisofgaitcyclekinematics
AT nataliasanchez identificationofsubtypesofpoststrokeandneurotypicalgaitbehaviorsusingneuralnetworkanalysisofgaitcyclekinematics