利用深度学习检测胸部CT上的闭塞性细支气管炎综合征。

IF 5.4 Q1 MEDICINE, RESEARCH & EXPERIMENTAL
Mateusz Koziński, Doruk Oner, Jakub Gwizdała, Catherine Beigelman-Aubry, Pascal Fua, Angela Koutsokera, Alessio Casutt, Argyro Vraka, Michele De Palma, John-David Aubert, Horst Bischof, Christophe von Garnier, Sahand Jamal Rahi, Martin Urschler, Nahal Mansouri
{"title":"利用深度学习检测胸部CT上的闭塞性细支气管炎综合征。","authors":"Mateusz Koziński, Doruk Oner, Jakub Gwizdała, Catherine Beigelman-Aubry, Pascal Fua, Angela Koutsokera, Alessio Casutt, Argyro Vraka, Michele De Palma, John-David Aubert, Horst Bischof, Christophe von Garnier, Sahand Jamal Rahi, Martin Urschler, Nahal Mansouri","doi":"10.1038/s43856-025-00732-x","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Bronchiolitis Obliterans Syndrome (BOS), a fibrotic airway disease that may develop after lung transplantation, conventionally relies on pulmonary function tests (PFTs) for diagnosis due to limitations of CT imaging. Deep neural networks (DNNs) have not previously been used for BOS detection. This study aims to train a DNN to detect BOS in CT scans using an approach tailored for low-data scenarios.</p><p><strong>Methods: </strong>We trained a DNN to detect BOS in CT scans using a co-training method designed to enhance performance in low-data environments. Our method employs an auxiliary task that makes the DNN more sensitive to disease manifestations and less sensitive to the patient's anatomical features. The DNN was tasked with predicting the sequence of two CT scans taken from the same BOS patient at least six months apart. We evaluated this approach on CT scans from 75 post-transplant patients, including 26 with BOS, and used a ROC-AUC metric to assess performance.</p><p><strong>Results: </strong>We show that our DNN method achieves a ROC-AUC of 0.90 (95% CI: 0.840-0.953) in distinguishing BOS from non-BOS in CT scans. Performance correlates with BOS progression, with ROC-AUC values of 0.88 for stage I, 0.91 for stage II, and 0.94 for stage III BOS. Notably, the DNN shows comparable performance on standard- and high-resolution CT scans. It also demonstrates the ability to predict BOS in at-risk patients (FEV1 between 80% and 90% of best FEV1) with a ROC-AUC of 0.87 (95% CI: 0.735-0.974). Using visual interpretation techniques for DNNs, we reveal sensitivity to hyperlucent/hypoattenuated areas indicative of air-trapping or bronchiectasis.</p><p><strong>Conclusions: </strong>Our approach shows potential for improving BOS diagnosis by enabling early detection and management. The ability to detect BOS from standard-resolution scans at any stage of respiration makes this method more accessible than previous approaches. Additionally, our findings highlight that techniques to limit overfitting are crucial for unlocking the potential of DNNs in low-data settings, which could assist clinicians in BOS studies with limited patient data.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"18"},"PeriodicalIF":5.4000,"publicationDate":"2025-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11739420/pdf/","citationCount":"0","resultStr":"{\"title\":\"Harnessing deep learning to detect bronchiolitis obliterans syndrome from chest CT.\",\"authors\":\"Mateusz Koziński, Doruk Oner, Jakub Gwizdała, Catherine Beigelman-Aubry, Pascal Fua, Angela Koutsokera, Alessio Casutt, Argyro Vraka, Michele De Palma, John-David Aubert, Horst Bischof, Christophe von Garnier, Sahand Jamal Rahi, Martin Urschler, Nahal Mansouri\",\"doi\":\"10.1038/s43856-025-00732-x\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>Bronchiolitis Obliterans Syndrome (BOS), a fibrotic airway disease that may develop after lung transplantation, conventionally relies on pulmonary function tests (PFTs) for diagnosis due to limitations of CT imaging. Deep neural networks (DNNs) have not previously been used for BOS detection. This study aims to train a DNN to detect BOS in CT scans using an approach tailored for low-data scenarios.</p><p><strong>Methods: </strong>We trained a DNN to detect BOS in CT scans using a co-training method designed to enhance performance in low-data environments. Our method employs an auxiliary task that makes the DNN more sensitive to disease manifestations and less sensitive to the patient's anatomical features. The DNN was tasked with predicting the sequence of two CT scans taken from the same BOS patient at least six months apart. We evaluated this approach on CT scans from 75 post-transplant patients, including 26 with BOS, and used a ROC-AUC metric to assess performance.</p><p><strong>Results: </strong>We show that our DNN method achieves a ROC-AUC of 0.90 (95% CI: 0.840-0.953) in distinguishing BOS from non-BOS in CT scans. Performance correlates with BOS progression, with ROC-AUC values of 0.88 for stage I, 0.91 for stage II, and 0.94 for stage III BOS. Notably, the DNN shows comparable performance on standard- and high-resolution CT scans. It also demonstrates the ability to predict BOS in at-risk patients (FEV1 between 80% and 90% of best FEV1) with a ROC-AUC of 0.87 (95% CI: 0.735-0.974). Using visual interpretation techniques for DNNs, we reveal sensitivity to hyperlucent/hypoattenuated areas indicative of air-trapping or bronchiectasis.</p><p><strong>Conclusions: </strong>Our approach shows potential for improving BOS diagnosis by enabling early detection and management. The ability to detect BOS from standard-resolution scans at any stage of respiration makes this method more accessible than previous approaches. Additionally, our findings highlight that techniques to limit overfitting are crucial for unlocking the potential of DNNs in low-data settings, which could assist clinicians in BOS studies with limited patient data.</p>\",\"PeriodicalId\":72646,\"journal\":{\"name\":\"Communications medicine\",\"volume\":\"5 1\",\"pages\":\"18\"},\"PeriodicalIF\":5.4000,\"publicationDate\":\"2025-01-16\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11739420/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Communications medicine\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1038/s43856-025-00732-x\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"MEDICINE, RESEARCH & EXPERIMENTAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Communications medicine","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1038/s43856-025-00732-x","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MEDICINE, RESEARCH & EXPERIMENTAL","Score":null,"Total":0}
引用次数: 0

摘要

背景:闭塞性毛细支气管炎综合征(BOS)是一种可能在肺移植后发生的纤维化气道疾病,由于CT成像的限制,传统上依赖于肺功能检查(PFTs)进行诊断。深度神经网络(dnn)以前没有用于BOS检测。本研究旨在训练DNN使用针对低数据场景定制的方法来检测CT扫描中的BOS。方法:我们使用一种旨在提高低数据环境下性能的协同训练方法训练DNN来检测CT扫描中的BOS。我们的方法采用了一种辅助任务,使DNN对疾病表现更加敏感,而对患者的解剖特征不那么敏感。DNN的任务是预测同一名BOS患者相隔至少6个月的两次CT扫描的顺序。我们对75例移植后患者的CT扫描进行了评估,其中包括26例BOS患者,并使用ROC-AUC指标评估其表现。结果:我们发现我们的DNN方法在CT扫描中区分BOS和非BOS的ROC-AUC为0.90 (95% CI: 0.840-0.953)。表现与BOS进展相关,第一阶段的ROC-AUC值为0.88,第二阶段为0.91,第三阶段为0.94。值得注意的是,DNN在标准和高分辨率CT扫描中表现出相当的性能。它还证明了预测高危患者(FEV1在最佳FEV1的80% - 90%之间)BOS的能力,ROC-AUC为0.87 (95% CI: 0.735-0.974)。使用dnn的视觉解释技术,我们揭示了对指示空气捕获或支气管扩张的高光/低衰减区域的敏感性。结论:我们的方法显示了通过早期发现和管理来改善BOS诊断的潜力。在呼吸的任何阶段从标准分辨率扫描中检测BOS的能力使该方法比以前的方法更容易使用。此外,我们的研究结果强调,限制过拟合的技术对于在低数据环境中释放dnn的潜力至关重要,这可以帮助临床医生在患者数据有限的情况下进行BOS研究。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Harnessing deep learning to detect bronchiolitis obliterans syndrome from chest CT.

Background: Bronchiolitis Obliterans Syndrome (BOS), a fibrotic airway disease that may develop after lung transplantation, conventionally relies on pulmonary function tests (PFTs) for diagnosis due to limitations of CT imaging. Deep neural networks (DNNs) have not previously been used for BOS detection. This study aims to train a DNN to detect BOS in CT scans using an approach tailored for low-data scenarios.

Methods: We trained a DNN to detect BOS in CT scans using a co-training method designed to enhance performance in low-data environments. Our method employs an auxiliary task that makes the DNN more sensitive to disease manifestations and less sensitive to the patient's anatomical features. The DNN was tasked with predicting the sequence of two CT scans taken from the same BOS patient at least six months apart. We evaluated this approach on CT scans from 75 post-transplant patients, including 26 with BOS, and used a ROC-AUC metric to assess performance.

Results: We show that our DNN method achieves a ROC-AUC of 0.90 (95% CI: 0.840-0.953) in distinguishing BOS from non-BOS in CT scans. Performance correlates with BOS progression, with ROC-AUC values of 0.88 for stage I, 0.91 for stage II, and 0.94 for stage III BOS. Notably, the DNN shows comparable performance on standard- and high-resolution CT scans. It also demonstrates the ability to predict BOS in at-risk patients (FEV1 between 80% and 90% of best FEV1) with a ROC-AUC of 0.87 (95% CI: 0.735-0.974). Using visual interpretation techniques for DNNs, we reveal sensitivity to hyperlucent/hypoattenuated areas indicative of air-trapping or bronchiectasis.

Conclusions: Our approach shows potential for improving BOS diagnosis by enabling early detection and management. The ability to detect BOS from standard-resolution scans at any stage of respiration makes this method more accessible than previous approaches. Additionally, our findings highlight that techniques to limit overfitting are crucial for unlocking the potential of DNNs in low-data settings, which could assist clinicians in BOS studies with limited patient data.

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
自引率
0.00%
发文量
0
×
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