Myo-Guide:基于机器学习的神经肌肉疾病MRI诊断Web应用

IF 9.4 1区 医学 Q1 GERIATRICS & GERONTOLOGY
Jose Verdu-Diaz, Carla Bolano-Díaz, Alejandro Gonzalez-Chamorro, Sam Fitzsimmons, Jodi Warman-Chardon, Goknur Selen Kocak, Debora Mucida-Alvim, Ian C. Smith, John Vissing, Nanna Scharff Poulsen, Sushan Luo, Cristina Domínguez-González, Laura Bermejo-Guerrero, David Gomez-Andres, Javier Sotoca, Anna Pichiecchio, Silvia Nicolosi, Mauro Monforte, Claudia Brogna, Eugenio Mercuri, Jorge Alfredo Bevilacqua, Jorge Díaz-Jara, Benjamín Pizarro-Galleguillos, Peter Krkoska, Jorge Alonso-Pérez, Montse Olivé, Erik H. Niks, Hermien E. Kan, James Lilleker, Mark Roberts, Bianca Buchignani, Jinhong Shin, Florence Esselin, Emmanuelle Le Bars, Anne Marie Childs, Edoardo Malfatti, Anna Sarkozy, Luke Perry, Sniya Sudhakar, Edmar Zanoteli, Filipe Tupinamba Di Pace, Emma Matthews, Shahram Attarian, David Bendahan, Matteo Garibaldi, Laura Fionda, Alicia Alonso-Jiménez, Robert Carlier, Ali Asghar Okhovat, Shahriar Nafissi, Atchayaram Nalini, Seena Vengalil, Kieren Hollingsworth, Chiara Marini-Bettolo, Volker Straub, Giorgio Tasca, Jaume Bacardit, Jordi Díaz-Manera, the Myo-Guide Consortium
{"title":"Myo-Guide:基于机器学习的神经肌肉疾病MRI诊断Web应用","authors":"Jose Verdu-Diaz,&nbsp;Carla Bolano-Díaz,&nbsp;Alejandro Gonzalez-Chamorro,&nbsp;Sam Fitzsimmons,&nbsp;Jodi Warman-Chardon,&nbsp;Goknur Selen Kocak,&nbsp;Debora Mucida-Alvim,&nbsp;Ian C. Smith,&nbsp;John Vissing,&nbsp;Nanna Scharff Poulsen,&nbsp;Sushan Luo,&nbsp;Cristina Domínguez-González,&nbsp;Laura Bermejo-Guerrero,&nbsp;David Gomez-Andres,&nbsp;Javier Sotoca,&nbsp;Anna Pichiecchio,&nbsp;Silvia Nicolosi,&nbsp;Mauro Monforte,&nbsp;Claudia Brogna,&nbsp;Eugenio Mercuri,&nbsp;Jorge Alfredo Bevilacqua,&nbsp;Jorge Díaz-Jara,&nbsp;Benjamín Pizarro-Galleguillos,&nbsp;Peter Krkoska,&nbsp;Jorge Alonso-Pérez,&nbsp;Montse Olivé,&nbsp;Erik H. Niks,&nbsp;Hermien E. Kan,&nbsp;James Lilleker,&nbsp;Mark Roberts,&nbsp;Bianca Buchignani,&nbsp;Jinhong Shin,&nbsp;Florence Esselin,&nbsp;Emmanuelle Le Bars,&nbsp;Anne Marie Childs,&nbsp;Edoardo Malfatti,&nbsp;Anna Sarkozy,&nbsp;Luke Perry,&nbsp;Sniya Sudhakar,&nbsp;Edmar Zanoteli,&nbsp;Filipe Tupinamba Di Pace,&nbsp;Emma Matthews,&nbsp;Shahram Attarian,&nbsp;David Bendahan,&nbsp;Matteo Garibaldi,&nbsp;Laura Fionda,&nbsp;Alicia Alonso-Jiménez,&nbsp;Robert Carlier,&nbsp;Ali Asghar Okhovat,&nbsp;Shahriar Nafissi,&nbsp;Atchayaram Nalini,&nbsp;Seena Vengalil,&nbsp;Kieren Hollingsworth,&nbsp;Chiara Marini-Bettolo,&nbsp;Volker Straub,&nbsp;Giorgio Tasca,&nbsp;Jaume Bacardit,&nbsp;Jordi Díaz-Manera,&nbsp;the Myo-Guide Consortium","doi":"10.1002/jcsm.13815","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Background</h3>\n \n <p>Neuromuscular diseases (NMDs) are rare disorders characterized by progressive muscle fibre loss, leading to replacement by fibrotic and fatty tissue, muscle weakness and disability. Early diagnosis is critical for therapeutic decisions, care planning and genetic counselling. Muscle magnetic resonance imaging (MRI) has emerged as a valuable diagnostic tool by identifying characteristic patterns of muscle involvement. However, the increasing complexity of these patterns complicates their interpretation, limiting their clinical utility. Additionally, multi-study data aggregation introduces heterogeneity challenges. This study presents a novel multi-study harmonization pipeline for muscle MRI and an AI-driven diagnostic tool to assist clinicians in identifying disease-specific muscle involvement patterns.</p>\n </section>\n \n <section>\n \n <h3> Methods</h3>\n \n <p>We developed a preprocessing pipeline to standardize MRI fat content across datasets, minimizing source bias. An ensemble of XGBoost models was trained to classify patients based on intramuscular fat replacement, age at MRI and sex. The SHapley Additive exPlanations (SHAP) framework was adapted to analyse model predictions and identify disease-specific muscle involvement patterns. To address class imbalance, training and evaluation were conducted using class-balanced metrics. The model's performance was compared against four expert clinicians using 14 previously unseen MRI scans.</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>Using our harmonization approach, we curated a dataset of 2961 MRI samples from genetically confirmed cases of 20 paediatric and adult NMDs. The model achieved a balanced accuracy of 64.8% ± 3.4%, with a weighted top-3 accuracy of 84.7% ± 1.8% and top-5 accuracy of 90.2% ± 2.4%. It also identified key features relevant for differential diagnosis, aiding clinical decision-making. Compared to four expert clinicians, the model obtained the highest top-3 accuracy (75.0% ± 4.8%). The diagnostic tool has been implemented as a free web platform, providing global access to the medical community.</p>\n </section>\n \n <section>\n \n <h3> Conclusions</h3>\n \n <p>The application of AI in muscle MRI for NMD diagnosis remains underexplored due to data scarcity. This study introduces a framework for dataset harmonization, enabling advanced computational techniques. Our findings demonstrate the potential of AI-based approaches to enhance differential diagnosis by identifying disease-specific muscle involvement patterns. The developed tool surpasses expert performance in diagnostic ranking and is accessible to clinicians worldwide via the Myo-Guide online platform.</p>\n </section>\n </div>","PeriodicalId":48911,"journal":{"name":"Journal of Cachexia Sarcopenia and Muscle","volume":"16 3","pages":""},"PeriodicalIF":9.4000,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/jcsm.13815","citationCount":"0","resultStr":"{\"title\":\"Myo-Guide: A Machine Learning-Based Web Application for Neuromuscular Disease Diagnosis With MRI\",\"authors\":\"Jose Verdu-Diaz,&nbsp;Carla Bolano-Díaz,&nbsp;Alejandro Gonzalez-Chamorro,&nbsp;Sam Fitzsimmons,&nbsp;Jodi Warman-Chardon,&nbsp;Goknur Selen Kocak,&nbsp;Debora Mucida-Alvim,&nbsp;Ian C. Smith,&nbsp;John Vissing,&nbsp;Nanna Scharff Poulsen,&nbsp;Sushan Luo,&nbsp;Cristina Domínguez-González,&nbsp;Laura Bermejo-Guerrero,&nbsp;David Gomez-Andres,&nbsp;Javier Sotoca,&nbsp;Anna Pichiecchio,&nbsp;Silvia Nicolosi,&nbsp;Mauro Monforte,&nbsp;Claudia Brogna,&nbsp;Eugenio Mercuri,&nbsp;Jorge Alfredo Bevilacqua,&nbsp;Jorge Díaz-Jara,&nbsp;Benjamín Pizarro-Galleguillos,&nbsp;Peter Krkoska,&nbsp;Jorge Alonso-Pérez,&nbsp;Montse Olivé,&nbsp;Erik H. Niks,&nbsp;Hermien E. Kan,&nbsp;James Lilleker,&nbsp;Mark Roberts,&nbsp;Bianca Buchignani,&nbsp;Jinhong Shin,&nbsp;Florence Esselin,&nbsp;Emmanuelle Le Bars,&nbsp;Anne Marie Childs,&nbsp;Edoardo Malfatti,&nbsp;Anna Sarkozy,&nbsp;Luke Perry,&nbsp;Sniya Sudhakar,&nbsp;Edmar Zanoteli,&nbsp;Filipe Tupinamba Di Pace,&nbsp;Emma Matthews,&nbsp;Shahram Attarian,&nbsp;David Bendahan,&nbsp;Matteo Garibaldi,&nbsp;Laura Fionda,&nbsp;Alicia Alonso-Jiménez,&nbsp;Robert Carlier,&nbsp;Ali Asghar Okhovat,&nbsp;Shahriar Nafissi,&nbsp;Atchayaram Nalini,&nbsp;Seena Vengalil,&nbsp;Kieren Hollingsworth,&nbsp;Chiara Marini-Bettolo,&nbsp;Volker Straub,&nbsp;Giorgio Tasca,&nbsp;Jaume Bacardit,&nbsp;Jordi Díaz-Manera,&nbsp;the Myo-Guide Consortium\",\"doi\":\"10.1002/jcsm.13815\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div>\\n \\n \\n <section>\\n \\n <h3> Background</h3>\\n \\n <p>Neuromuscular diseases (NMDs) are rare disorders characterized by progressive muscle fibre loss, leading to replacement by fibrotic and fatty tissue, muscle weakness and disability. Early diagnosis is critical for therapeutic decisions, care planning and genetic counselling. Muscle magnetic resonance imaging (MRI) has emerged as a valuable diagnostic tool by identifying characteristic patterns of muscle involvement. However, the increasing complexity of these patterns complicates their interpretation, limiting their clinical utility. Additionally, multi-study data aggregation introduces heterogeneity challenges. This study presents a novel multi-study harmonization pipeline for muscle MRI and an AI-driven diagnostic tool to assist clinicians in identifying disease-specific muscle involvement patterns.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Methods</h3>\\n \\n <p>We developed a preprocessing pipeline to standardize MRI fat content across datasets, minimizing source bias. An ensemble of XGBoost models was trained to classify patients based on intramuscular fat replacement, age at MRI and sex. The SHapley Additive exPlanations (SHAP) framework was adapted to analyse model predictions and identify disease-specific muscle involvement patterns. To address class imbalance, training and evaluation were conducted using class-balanced metrics. The model's performance was compared against four expert clinicians using 14 previously unseen MRI scans.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Results</h3>\\n \\n <p>Using our harmonization approach, we curated a dataset of 2961 MRI samples from genetically confirmed cases of 20 paediatric and adult NMDs. The model achieved a balanced accuracy of 64.8% ± 3.4%, with a weighted top-3 accuracy of 84.7% ± 1.8% and top-5 accuracy of 90.2% ± 2.4%. It also identified key features relevant for differential diagnosis, aiding clinical decision-making. Compared to four expert clinicians, the model obtained the highest top-3 accuracy (75.0% ± 4.8%). The diagnostic tool has been implemented as a free web platform, providing global access to the medical community.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Conclusions</h3>\\n \\n <p>The application of AI in muscle MRI for NMD diagnosis remains underexplored due to data scarcity. This study introduces a framework for dataset harmonization, enabling advanced computational techniques. Our findings demonstrate the potential of AI-based approaches to enhance differential diagnosis by identifying disease-specific muscle involvement patterns. The developed tool surpasses expert performance in diagnostic ranking and is accessible to clinicians worldwide via the Myo-Guide online platform.</p>\\n </section>\\n </div>\",\"PeriodicalId\":48911,\"journal\":{\"name\":\"Journal of Cachexia Sarcopenia and Muscle\",\"volume\":\"16 3\",\"pages\":\"\"},\"PeriodicalIF\":9.4000,\"publicationDate\":\"2025-04-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://onlinelibrary.wiley.com/doi/epdf/10.1002/jcsm.13815\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Cachexia Sarcopenia and Muscle\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13815\",\"RegionNum\":1,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"GERIATRICS & GERONTOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Cachexia Sarcopenia and Muscle","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13815","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GERIATRICS & GERONTOLOGY","Score":null,"Total":0}
引用次数: 0

摘要

神经肌肉疾病(NMDs)是一种罕见的疾病,其特征是进行性肌纤维损失,导致纤维化和脂肪组织替代,肌肉无力和残疾。早期诊断对于治疗决策、护理计划和遗传咨询至关重要。肌肉磁共振成像(MRI)已成为一种有价值的诊断工具,通过识别肌肉受累的特征模式。然而,这些模式的复杂性增加了其解释的复杂性,限制了其临床应用。此外,多研究数据聚合带来了异质性挑战。本研究提出了一种新的肌肉MRI多研究协调管道和人工智能驱动的诊断工具,以帮助临床医生识别疾病特异性肌肉受累模式。方法:我们开发了一种预处理管道,以标准化各数据集的MRI脂肪含量,最大限度地减少源偏差。训练了一组XGBoost模型,根据肌内脂肪替代、MRI年龄和性别对患者进行分类。SHapley加性解释(SHAP)框架被用于分析模型预测和识别疾病特异性肌肉受累模式。为了解决班级失衡问题,培训和评估采用班级平衡指标进行。该模型的性能与四位专家临床医生使用14个以前未见过的MRI扫描进行了比较。使用我们的协调方法,我们整理了来自20例儿科和成人nmd遗传确诊病例的2961个MRI样本的数据集。该模型的平衡精度为64.8%±3.4%,加权前3精度为84.7%±1.8%,加权前5精度为90.2%±2.4%。它还确定了与鉴别诊断相关的关键特征,帮助临床决策。与4位专家临床医生相比,该模型获得了最高的前3名准确率(75.0%±4.8%)。该诊断工具已作为一个免费的网络平台实施,为全球医疗界提供访问。结论由于数据缺乏,人工智能在肌肉MRI诊断NMD中的应用尚待探索。本研究引入了一个数据集协调框架,使先进的计算技术成为可能。我们的研究结果证明了基于人工智能的方法通过识别疾病特异性肌肉受累模式来增强鉴别诊断的潜力。开发的工具在诊断排名方面超过了专家的表现,并且通过Myo-Guide在线平台可供全球临床医生使用。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Myo-Guide: A Machine Learning-Based Web Application for Neuromuscular Disease Diagnosis With MRI

Myo-Guide: A Machine Learning-Based Web Application for Neuromuscular Disease Diagnosis With MRI

Background

Neuromuscular diseases (NMDs) are rare disorders characterized by progressive muscle fibre loss, leading to replacement by fibrotic and fatty tissue, muscle weakness and disability. Early diagnosis is critical for therapeutic decisions, care planning and genetic counselling. Muscle magnetic resonance imaging (MRI) has emerged as a valuable diagnostic tool by identifying characteristic patterns of muscle involvement. However, the increasing complexity of these patterns complicates their interpretation, limiting their clinical utility. Additionally, multi-study data aggregation introduces heterogeneity challenges. This study presents a novel multi-study harmonization pipeline for muscle MRI and an AI-driven diagnostic tool to assist clinicians in identifying disease-specific muscle involvement patterns.

Methods

We developed a preprocessing pipeline to standardize MRI fat content across datasets, minimizing source bias. An ensemble of XGBoost models was trained to classify patients based on intramuscular fat replacement, age at MRI and sex. The SHapley Additive exPlanations (SHAP) framework was adapted to analyse model predictions and identify disease-specific muscle involvement patterns. To address class imbalance, training and evaluation were conducted using class-balanced metrics. The model's performance was compared against four expert clinicians using 14 previously unseen MRI scans.

Results

Using our harmonization approach, we curated a dataset of 2961 MRI samples from genetically confirmed cases of 20 paediatric and adult NMDs. The model achieved a balanced accuracy of 64.8% ± 3.4%, with a weighted top-3 accuracy of 84.7% ± 1.8% and top-5 accuracy of 90.2% ± 2.4%. It also identified key features relevant for differential diagnosis, aiding clinical decision-making. Compared to four expert clinicians, the model obtained the highest top-3 accuracy (75.0% ± 4.8%). The diagnostic tool has been implemented as a free web platform, providing global access to the medical community.

Conclusions

The application of AI in muscle MRI for NMD diagnosis remains underexplored due to data scarcity. This study introduces a framework for dataset harmonization, enabling advanced computational techniques. Our findings demonstrate the potential of AI-based approaches to enhance differential diagnosis by identifying disease-specific muscle involvement patterns. The developed tool surpasses expert performance in diagnostic ranking and is accessible to clinicians worldwide via the Myo-Guide online platform.

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
Journal of Cachexia Sarcopenia and Muscle
Journal of Cachexia Sarcopenia and Muscle MEDICINE, GENERAL & INTERNAL-
CiteScore
13.30
自引率
12.40%
发文量
234
审稿时长
16 weeks
期刊介绍: The Journal of Cachexia, Sarcopenia and Muscle is a peer-reviewed international journal dedicated to publishing materials related to cachexia and sarcopenia, as well as body composition and its physiological and pathophysiological changes across the lifespan and in response to various illnesses from all fields of life sciences. The journal aims to provide a reliable resource for professionals interested in related research or involved in the clinical care of affected patients, such as those suffering from AIDS, cancer, chronic heart failure, chronic lung disease, liver cirrhosis, chronic kidney failure, rheumatoid arthritis, or sepsis.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信