DiffuseGaitNet: Improving Parkinson’s Disease Gait Severity Assessment With a Diffusion Model Framework

Assessing the severity of gait impairment in Parkinson’s disease (PD) using the Movement Disorder Society’s Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) is typically performed by clinical experts, but this process is time-consuming, subjective, and costly. To...

Full description

Saved in:
Bibliographic Details
Main Authors: Arshak Rezvani, Nasrin Ravansalar, Mohammad Ali Akhaee, Andrew J. Greenshaw, Russell Greiner, Maryam S. Mirian, Muhammad Yousefnezhad, Martin J. McKeown
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/11080070/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849419932388818944
author Arshak Rezvani
Nasrin Ravansalar
Mohammad Ali Akhaee
Andrew J. Greenshaw
Russell Greiner
Maryam S. Mirian
Muhammad Yousefnezhad
Martin J. McKeown
author_facet Arshak Rezvani
Nasrin Ravansalar
Mohammad Ali Akhaee
Andrew J. Greenshaw
Russell Greiner
Maryam S. Mirian
Muhammad Yousefnezhad
Martin J. McKeown
author_sort Arshak Rezvani
collection DOAJ
description Assessing the severity of gait impairment in Parkinson&#x2019;s disease (PD) using the Movement Disorder Society&#x2019;s Unified Parkinson&#x2019;s Disease Rating Scale (MDS-UPDRS) is typically performed by clinical experts, but this process is time-consuming, subjective, and costly. To address these challenges, we propose a Guided Diffusion Model with an encoder-only transformer that automatically predicts gait severity by learning the underlying distribution of PD gait and leveraging domain knowledge critical for clinical evaluations. Our diffusion model enables us to generate synthetic PD gait video frames conditioned on clinical features determined by experts to assess disease severity. These synthetic samples contain novel movement patterns not present in the observed data; systems trained on this information have better prediction performance. In addition, we propose a novel classification algorithm that can learn a predictive model, from both observed training data and synthetic samples, to accurately assess PD severity. We evaluate the effectiveness of the proposed method using two human motion datasets across two tasks: PD severity prediction and action classification. Our approach in predicting PD, and our action classification is sufficiently accurate that it can be applied to general applications with healthy subjects performing similar tasks. The full codebase is available on GitHub: <uri>https://github.com/arshakRz/DiffuseGaitNet</uri>
format Article
id doaj-art-aa3c7b78b0c9419085d39db788c846df
institution Kabale University
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-aa3c7b78b0c9419085d39db788c846df2025-08-20T03:31:55ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1534-43201558-02102025-01-01332858286910.1109/TNSRE.2025.358907411080070DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model FrameworkArshak Rezvani0https://orcid.org/0009-0008-4225-016XNasrin Ravansalar1https://orcid.org/0009-0009-8541-5670Mohammad Ali Akhaee2https://orcid.org/0000-0003-3753-5618Andrew J. Greenshaw3Russell Greiner4https://orcid.org/0000-0001-8327-934XMaryam S. Mirian5https://orcid.org/0000-0001-5411-9902Muhammad Yousefnezhad6https://orcid.org/0000-0002-9402-5027Martin J. McKeown7https://orcid.org/0000-0002-4048-0817Department of Medicine, The University of British Columbia (UBC), Vancouver, CanadaElectrical and Computer Engineering (ECE) Department, University of Tehran, Tehran, IranElectrical and Computer Engineering (ECE) Department, University of Tehran, Tehran, IranDepartment of Psychiatry, University of Alberta (UoA), Edmonton, CanadaDepartment of Computing Science, UoA, Edmonton, CanadaDepartment of Medicine, The University of British Columbia (UBC), Vancouver, CanadaDepartment of Computing Science and the Department of Psychiatry, UoA, Edmonton, CanadaDepartment of Medicine, The University of British Columbia (UBC), Vancouver, CanadaAssessing the severity of gait impairment in Parkinson&#x2019;s disease (PD) using the Movement Disorder Society&#x2019;s Unified Parkinson&#x2019;s Disease Rating Scale (MDS-UPDRS) is typically performed by clinical experts, but this process is time-consuming, subjective, and costly. To address these challenges, we propose a Guided Diffusion Model with an encoder-only transformer that automatically predicts gait severity by learning the underlying distribution of PD gait and leveraging domain knowledge critical for clinical evaluations. Our diffusion model enables us to generate synthetic PD gait video frames conditioned on clinical features determined by experts to assess disease severity. These synthetic samples contain novel movement patterns not present in the observed data; systems trained on this information have better prediction performance. In addition, we propose a novel classification algorithm that can learn a predictive model, from both observed training data and synthetic samples, to accurately assess PD severity. We evaluate the effectiveness of the proposed method using two human motion datasets across two tasks: PD severity prediction and action classification. Our approach in predicting PD, and our action classification is sufficiently accurate that it can be applied to general applications with healthy subjects performing similar tasks. The full codebase is available on GitHub: <uri>https://github.com/arshakRz/DiffuseGaitNet</uri>https://ieeexplore.ieee.org/document/11080070/Diffusion modelsgenerative AItransformersattention-based networksParkinson’s diseasegait impairments
spellingShingle Arshak Rezvani
Nasrin Ravansalar
Mohammad Ali Akhaee
Andrew J. Greenshaw
Russell Greiner
Maryam S. Mirian
Muhammad Yousefnezhad
Martin J. McKeown
DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model Framework
IEEE Transactions on Neural Systems and Rehabilitation Engineering
Diffusion models
generative AI
transformers
attention-based networks
Parkinson’s disease
gait impairments
title DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model Framework
title_full DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model Framework
title_fullStr DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model Framework
title_full_unstemmed DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model Framework
title_short DiffuseGaitNet: Improving Parkinson&#x2019;s Disease Gait Severity Assessment With a Diffusion Model Framework
title_sort diffusegaitnet improving parkinson x2019 s disease gait severity assessment with a diffusion model framework
topic Diffusion models
generative AI
transformers
attention-based networks
Parkinson’s disease
gait impairments
url https://ieeexplore.ieee.org/document/11080070/
work_keys_str_mv AT arshakrezvani diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT nasrinravansalar diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT mohammadaliakhaee diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT andrewjgreenshaw diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT russellgreiner diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT maryamsmirian diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT muhammadyousefnezhad diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework
AT martinjmckeown diffusegaitnetimprovingparkinsonx2019sdiseasegaitseverityassessmentwithadiffusionmodelframework