基于步态视频的小脑共济失调严重程度深度神经网络预测

IF 7.4 1区医学 Q1 CLINICAL NEUROLOGY

Movement Disorders Pub Date : 2025-01-22 DOI:10.1002/mds.30113

Katsuki Eguchi MD, PhD, Hiroaki Yaguchi MD, PhD, Hisashi Uwatoko MD, PhD, Yuki Iida MD, PhD, Shinsuke Hamada MD, PhD, Sanae Honma MD, PhD, Asako Takei MD, PhD, Fumio Moriwaka MD, PhD, Ichiro Yabe MD, PhD

{"title":"基于步态视频的小脑共济失调严重程度深度神经网络预测","authors":"Katsuki Eguchi MD, PhD, Hiroaki Yaguchi MD, PhD, Hisashi Uwatoko MD, PhD, Yuki Iida MD, PhD, Shinsuke Hamada MD, PhD, Sanae Honma MD, PhD, Asako Takei MD, PhD, Fumio Moriwaka MD, PhD, Ichiro Yabe MD, PhD","doi":"10.1002/mds.30113","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Background</h3>\n \n <p>Pose estimation algorithms applied to two-dimensional videos evaluate gait disturbances; however, a few studies have used this method to evaluate ataxic gait.</p>\n </section>\n \n <section>\n \n <h3> Objective</h3>\n \n <p>The aim was to assess whether a pose estimation algorithm can predict the severity of cerebellar ataxia by applying it to gait videos.</p>\n </section>\n \n <section>\n \n <h3> Methods</h3>\n \n <p>We video-recorded 66 patients with degenerative cerebellar diseases performing the timed up-and-go test. Key points from the gait videos extracted by a pose estimation algorithm were input into a deep learning model to predict the Scale for the Assessment and Rating of Ataxia (SARA) score. We also evaluated video segments that the model focused on to predict ataxia severity.</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>The model achieved a root-mean-square error of 2.30 and a coefficient of determination of 0.79 in predicting the SARA score. It primarily focused on standing, turning, and body sway to assess severity.</p>\n </section>\n \n <section>\n \n <h3> Conclusions</h3>\n \n <p>This study demonstrated that the model may capture gait characteristics from key-point data and has the potential to predict SARA scores. © 2025 International Parkinson and Movement Disorder Society.</p>\n </section>\n </div>","PeriodicalId":213,"journal":{"name":"Movement Disorders","volume":"40 4","pages":"752-758"},"PeriodicalIF":7.4000,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Gait Video–Based Prediction of Severity of Cerebellar Ataxia Using Deep Neural Networks\",\"authors\":\"Katsuki Eguchi MD, PhD, Hiroaki Yaguchi MD, PhD, Hisashi Uwatoko MD, PhD, Yuki Iida MD, PhD, Shinsuke Hamada MD, PhD, Sanae Honma MD, PhD, Asako Takei MD, PhD, Fumio Moriwaka MD, PhD, Ichiro Yabe MD, PhD\",\"doi\":\"10.1002/mds.30113\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div>\\n \\n \\n <section>\\n \\n <h3> Background</h3>\\n \\n <p>Pose estimation algorithms applied to two-dimensional videos evaluate gait disturbances; however, a few studies have used this method to evaluate ataxic gait.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Objective</h3>\\n \\n <p>The aim was to assess whether a pose estimation algorithm can predict the severity of cerebellar ataxia by applying it to gait videos.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Methods</h3>\\n \\n <p>We video-recorded 66 patients with degenerative cerebellar diseases performing the timed up-and-go test. Key points from the gait videos extracted by a pose estimation algorithm were input into a deep learning model to predict the Scale for the Assessment and Rating of Ataxia (SARA) score. We also evaluated video segments that the model focused on to predict ataxia severity.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Results</h3>\\n \\n <p>The model achieved a root-mean-square error of 2.30 and a coefficient of determination of 0.79 in predicting the SARA score. It primarily focused on standing, turning, and body sway to assess severity.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Conclusions</h3>\\n \\n <p>This study demonstrated that the model may capture gait characteristics from key-point data and has the potential to predict SARA scores. © 2025 International Parkinson and Movement Disorder Society.</p>\\n </section>\\n </div>\",\"PeriodicalId\":213,\"journal\":{\"name\":\"Movement Disorders\",\"volume\":\"40 4\",\"pages\":\"752-758\"},\"PeriodicalIF\":7.4000,\"publicationDate\":\"2025-01-22\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Movement Disorders\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/mds.30113\",\"RegionNum\":1,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"CLINICAL NEUROLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Movement Disorders","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/mds.30113","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CLINICAL NEUROLOGY","Score":null,"Total":0}

引用次数: 0

摘要

背景姿态估计算法应用于二维视频评估步态干扰；然而，很少有研究使用这种方法来评估共济失调步态。目的将姿态估计算法应用于步态视频，评估姿态估计算法能否预测小脑性共济失调的严重程度。方法对66例退行性小脑疾病患者进行定时up - and - go测试。将姿态估计算法提取的步态视频中的关键点输入到深度学习模型中，以预测共济失调（SARA）评分的评估和评级量表。我们还评估了模型关注的视频片段，以预测共济失调的严重程度。结果该模型预测SARA评分的均方根误差为2.30，决定系数为0.79。它主要集中在站立，转身和身体摇摆来评估严重程度。本研究表明，该模型可以从关键点数据中捕获步态特征，并具有预测SARA评分的潜力。©2025国际帕金森和运动障碍学会。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Gait Video–Based Prediction of Severity of Cerebellar Ataxia Using Deep Neural Networks

Background

Pose estimation algorithms applied to two-dimensional videos evaluate gait disturbances; however, a few studies have used this method to evaluate ataxic gait.

Objective

The aim was to assess whether a pose estimation algorithm can predict the severity of cerebellar ataxia by applying it to gait videos.

Methods

We video-recorded 66 patients with degenerative cerebellar diseases performing the timed up-and-go test. Key points from the gait videos extracted by a pose estimation algorithm were input into a deep learning model to predict the Scale for the Assessment and Rating of Ataxia (SARA) score. We also evaluated video segments that the model focused on to predict ataxia severity.

Results

The model achieved a root-mean-square error of 2.30 and a coefficient of determination of 0.79 in predicting the SARA score. It primarily focused on standing, turning, and body sway to assess severity.

Conclusions

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Movement Disorders 医学-临床神经学

CiteScore

13.30

自引率

8.10%

发文量

371

审稿时长

12 months

期刊介绍： Movement Disorders publishes a variety of content types including Reviews, Viewpoints, Full Length Articles, Historical Reports, Brief Reports, and Letters. The journal considers original manuscripts on topics related to the diagnosis, therapeutics, pharmacology, biochemistry, physiology, etiology, genetics, and epidemiology of movement disorders. Appropriate topics include Parkinsonism, Chorea, Tremors, Dystonia, Myoclonus, Tics, Tardive Dyskinesia, Spasticity, and Ataxia.