评估用于弥漫性低级别胶质瘤随访的 nnU-Net 型自动临床肿瘤体积分割工具

IF 3 3区 医学 Q2 CLINICAL NEUROLOGY
Margaux Verdier , Jeremy Deverdun , Nicolas Menjot de Champfleur , Hugues Duffau , Philippe Lam , Thomas Dos Santos , Thomas Troalen , Bénédicte Maréchal , Till Huelnhagen , Emmanuelle Le Bars
{"title":"评估用于弥漫性低级别胶质瘤随访的 nnU-Net 型自动临床肿瘤体积分割工具","authors":"Margaux Verdier ,&nbsp;Jeremy Deverdun ,&nbsp;Nicolas Menjot de Champfleur ,&nbsp;Hugues Duffau ,&nbsp;Philippe Lam ,&nbsp;Thomas Dos Santos ,&nbsp;Thomas Troalen ,&nbsp;Bénédicte Maréchal ,&nbsp;Till Huelnhagen ,&nbsp;Emmanuelle Le Bars","doi":"10.1016/j.neurad.2023.05.008","DOIUrl":null,"url":null,"abstract":"<div><h3>Background and purpose</h3><p><span>Diffuse low-grade gliomas (DLGG) are characterized by a slow and continuous growth and always evolve towards an aggressive grade. Accurate prediction of the malignant transformation<span> is essential as it requires immediate therapeutic intervention. One of its most precise predictors is the velocity of diameter expansion (VDE). Currently, the VDE is estimated either by </span></span>linear measurements<span> or by manual delineation of the DLGG on T2 FLAIR acquisitions. However, because of the DLGG's infiltrative nature and its blurred contours, manual measures are challenging and variable, even for experts. Therefore we propose an automated segmentation algorithm using a 2D nnU-Net, to 1) gain time and 2) standardize VDE assessment.</span></p></div><div><h3>Materials and Methods</h3><p>The 2D nnU-Net was trained on 318 acquisitions (T2 FLAIR &amp; 3DT1 longitudinal follow-up of 30 patients, including pre- &amp; post-surgery acquisitions, different scanners, vendors, imaging parameters…). Automated vs. manual segmentation performance was evaluated on 167 acquisitions, and its clinical interest was validated by quantifying the amount of manual correction required after automated segmentation of 98 novel acquisitions.</p></div><div><h3>Results</h3><p>Automated segmentation showed a good performance with a mean Dice Similarity Coefficient (DSC) of 0.82±0.13 with manual segmentation and a substantial concordance between VDE calculations. Major manual corrections (i.e., DSC&lt;0.7) were necessary only in 3/98 cases and 81% of the cases had a DSC&gt;0.9.</p></div><div><h3>Conclusion</h3><p>The proposed automated segmentation algorithm can successfully segment DLGG on highly variable MRI data. Although manual corrections are sometimes necessary, it provides a reliable, standardized and time-winning support for VDE extraction to asses DLGG growth.</p></div>","PeriodicalId":50115,"journal":{"name":"Journal of Neuroradiology","volume":null,"pages":null},"PeriodicalIF":3.0000,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Evaluation of a nnU-Net type automated clinical volumetric tumor segmentation tool for diffuse low-grade glioma follow-up\",\"authors\":\"Margaux Verdier ,&nbsp;Jeremy Deverdun ,&nbsp;Nicolas Menjot de Champfleur ,&nbsp;Hugues Duffau ,&nbsp;Philippe Lam ,&nbsp;Thomas Dos Santos ,&nbsp;Thomas Troalen ,&nbsp;Bénédicte Maréchal ,&nbsp;Till Huelnhagen ,&nbsp;Emmanuelle Le Bars\",\"doi\":\"10.1016/j.neurad.2023.05.008\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><h3>Background and purpose</h3><p><span>Diffuse low-grade gliomas (DLGG) are characterized by a slow and continuous growth and always evolve towards an aggressive grade. Accurate prediction of the malignant transformation<span> is essential as it requires immediate therapeutic intervention. One of its most precise predictors is the velocity of diameter expansion (VDE). Currently, the VDE is estimated either by </span></span>linear measurements<span> or by manual delineation of the DLGG on T2 FLAIR acquisitions. However, because of the DLGG's infiltrative nature and its blurred contours, manual measures are challenging and variable, even for experts. Therefore we propose an automated segmentation algorithm using a 2D nnU-Net, to 1) gain time and 2) standardize VDE assessment.</span></p></div><div><h3>Materials and Methods</h3><p>The 2D nnU-Net was trained on 318 acquisitions (T2 FLAIR &amp; 3DT1 longitudinal follow-up of 30 patients, including pre- &amp; post-surgery acquisitions, different scanners, vendors, imaging parameters…). Automated vs. manual segmentation performance was evaluated on 167 acquisitions, and its clinical interest was validated by quantifying the amount of manual correction required after automated segmentation of 98 novel acquisitions.</p></div><div><h3>Results</h3><p>Automated segmentation showed a good performance with a mean Dice Similarity Coefficient (DSC) of 0.82±0.13 with manual segmentation and a substantial concordance between VDE calculations. Major manual corrections (i.e., DSC&lt;0.7) were necessary only in 3/98 cases and 81% of the cases had a DSC&gt;0.9.</p></div><div><h3>Conclusion</h3><p>The proposed automated segmentation algorithm can successfully segment DLGG on highly variable MRI data. Although manual corrections are sometimes necessary, it provides a reliable, standardized and time-winning support for VDE extraction to asses DLGG growth.</p></div>\",\"PeriodicalId\":50115,\"journal\":{\"name\":\"Journal of Neuroradiology\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.0000,\"publicationDate\":\"2024-02-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Neuroradiology\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0150986123002134\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"CLINICAL NEUROLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Neuroradiology","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0150986123002134","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CLINICAL NEUROLOGY","Score":null,"Total":0}
引用次数: 0

摘要

背景和目的弥漫性低级别胶质瘤(DLGG)的特点是生长缓慢而持续,并总是向侵袭性级别演变。准确预测恶性转化至关重要,因为这需要立即进行治疗干预。直径扩张速度(VDE)是最精确的预测指标之一。目前,VDE 是通过线性测量或在 T2 FLAIR 采集中手动划定 DLGG 来估算的。然而,由于 DLGG 的浸润性及其模糊的轮廓,即使对专家来说,人工测量也具有挑战性和可变性。因此,我们提出了一种使用 2D nnU-Net 的自动分割算法,目的是:1)节省时间;2)使 VDE 评估标准化。材料与方法在 318 次采集(30 位患者的 T2 FLAIR 和 3DT1 纵向随访,包括手术前和手术后采集、不同的扫描仪、供应商、成像参数......)上训练 2D nnU-Net。结果 自动分割显示出良好的性能,与手动分割相比,平均骰子相似系数(DSC)为(0.82±0.13),VDE计算结果也非常一致。只有 3/98 的病例需要进行主要的人工校正(即 DSC<0.7),81% 的病例的 DSC>0.9.结论所提出的自动分割算法可以在高度多变的 MRI 数据上成功分割 DLGG。虽然有时需要人工校正,但它为评估 DLGG 生长的 VDE 提取提供了可靠、标准化和省时的支持。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Evaluation of a nnU-Net type automated clinical volumetric tumor segmentation tool for diffuse low-grade glioma follow-up

Evaluation of a nnU-Net type automated clinical volumetric tumor segmentation tool for diffuse low-grade glioma follow-up

Evaluation of a nnU-Net type automated clinical volumetric tumor segmentation tool for diffuse low-grade glioma follow-up

Background and purpose

Diffuse low-grade gliomas (DLGG) are characterized by a slow and continuous growth and always evolve towards an aggressive grade. Accurate prediction of the malignant transformation is essential as it requires immediate therapeutic intervention. One of its most precise predictors is the velocity of diameter expansion (VDE). Currently, the VDE is estimated either by linear measurements or by manual delineation of the DLGG on T2 FLAIR acquisitions. However, because of the DLGG's infiltrative nature and its blurred contours, manual measures are challenging and variable, even for experts. Therefore we propose an automated segmentation algorithm using a 2D nnU-Net, to 1) gain time and 2) standardize VDE assessment.

Materials and Methods

The 2D nnU-Net was trained on 318 acquisitions (T2 FLAIR & 3DT1 longitudinal follow-up of 30 patients, including pre- & post-surgery acquisitions, different scanners, vendors, imaging parameters…). Automated vs. manual segmentation performance was evaluated on 167 acquisitions, and its clinical interest was validated by quantifying the amount of manual correction required after automated segmentation of 98 novel acquisitions.

Results

Automated segmentation showed a good performance with a mean Dice Similarity Coefficient (DSC) of 0.82±0.13 with manual segmentation and a substantial concordance between VDE calculations. Major manual corrections (i.e., DSC<0.7) were necessary only in 3/98 cases and 81% of the cases had a DSC>0.9.

Conclusion

The proposed automated segmentation algorithm can successfully segment DLGG on highly variable MRI data. Although manual corrections are sometimes necessary, it provides a reliable, standardized and time-winning support for VDE extraction to asses DLGG growth.

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 Neuroradiology
Journal of Neuroradiology 医学-核医学
CiteScore
6.10
自引率
5.70%
发文量
142
审稿时长
6-12 weeks
期刊介绍: The Journal of Neuroradiology is a peer-reviewed journal, publishing worldwide clinical and basic research in the field of diagnostic and Interventional neuroradiology, translational and molecular neuroimaging, and artificial intelligence in neuroradiology. The Journal of Neuroradiology considers for publication articles, reviews, technical notes and letters to the editors (correspondence section), provided that the methodology and scientific content are of high quality, and that the results will have substantial clinical impact and/or physiological importance.
×
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