数据增强对基于深度学习的长轴 Cine-MRI 分段的影响

IF 1.8 Q3 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Algorithms Pub Date : 2023-12-25 DOI:10.3390/a17010010
François Legrand, Richard Macwan, Alain Lalande, Lisa Métairie, Thomas Decourselle
{"title":"数据增强对基于深度学习的长轴 Cine-MRI 分段的影响","authors":"François Legrand, Richard Macwan, Alain Lalande, Lisa Métairie, Thomas Decourselle","doi":"10.3390/a17010010","DOIUrl":null,"url":null,"abstract":"Automated Cardiac Magnetic Resonance segmentation serves as a crucial tool for the evaluation of cardiac function, facilitating faster clinical assessments that prove advantageous for both practitioners and patients alike. Recent studies have predominantly concentrated on delineating structures on short-axis orientation, placing less emphasis on long-axis representations due to the intricate nature of structures in the latter. Taking these consideration into account, we present a robust hierarchy-based augmentation strategy coupled with the compact and fast Efficient-Net (ENet) architecture for the automated segmentation of two-chamber and four-chamber Cine-MRI images. We observed an average Dice improvement of 0.99% on the two-chamber images and of 2.15% on the four-chamber images, and an average Hausdorff distance improvement of 21.3% on the two-chamber images and of 29.6% on the four-chamber images. The practical viability of our approach was validated by computing clinical metrics such as the Left Ventricular Ejection Fraction (LVEF) and left ventricular volume (LVC). We observed acceptable biases, with a +2.81% deviation on the LVEF for the two-chamber images and a +0.11% deviation for the four-chamber images.","PeriodicalId":7636,"journal":{"name":"Algorithms","volume":"23 S2","pages":""},"PeriodicalIF":1.8000,"publicationDate":"2023-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Effect of Data Augmentation on Deep-Learning-Based Segmentation of Long-Axis Cine-MRI\",\"authors\":\"François Legrand, Richard Macwan, Alain Lalande, Lisa Métairie, Thomas Decourselle\",\"doi\":\"10.3390/a17010010\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Automated Cardiac Magnetic Resonance segmentation serves as a crucial tool for the evaluation of cardiac function, facilitating faster clinical assessments that prove advantageous for both practitioners and patients alike. Recent studies have predominantly concentrated on delineating structures on short-axis orientation, placing less emphasis on long-axis representations due to the intricate nature of structures in the latter. Taking these consideration into account, we present a robust hierarchy-based augmentation strategy coupled with the compact and fast Efficient-Net (ENet) architecture for the automated segmentation of two-chamber and four-chamber Cine-MRI images. We observed an average Dice improvement of 0.99% on the two-chamber images and of 2.15% on the four-chamber images, and an average Hausdorff distance improvement of 21.3% on the two-chamber images and of 29.6% on the four-chamber images. The practical viability of our approach was validated by computing clinical metrics such as the Left Ventricular Ejection Fraction (LVEF) and left ventricular volume (LVC). We observed acceptable biases, with a +2.81% deviation on the LVEF for the two-chamber images and a +0.11% deviation for the four-chamber images.\",\"PeriodicalId\":7636,\"journal\":{\"name\":\"Algorithms\",\"volume\":\"23 S2\",\"pages\":\"\"},\"PeriodicalIF\":1.8000,\"publicationDate\":\"2023-12-25\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Algorithms\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.3390/a17010010\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Algorithms","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3390/a17010010","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0

摘要

自动心脏磁共振分割是评估心脏功能的重要工具,有助于更快地进行临床评估,对医生和患者都很有利。近期的研究主要集中在短轴方向上的结构划分,而对长轴方向上的结构划分重视不够,因为长轴方向上的结构错综复杂。考虑到这些因素,我们提出了一种基于层次结构的稳健增强策略,并结合紧凑快速的 Efficient-Net (ENet) 架构,用于两腔和四腔 Cine-MRI 图像的自动分割。我们观察到两腔图像的 Dice 平均改善率为 0.99%,四腔图像的 Dice 平均改善率为 2.15%,两腔图像的 Hausdorff 距离平均改善率为 21.3%,四腔图像的 Hausdorff 距离平均改善率为 29.6%。通过计算左室射血分数(LVEF)和左室容积(LVC)等临床指标,验证了我们方法的实际可行性。我们观察到了可接受的偏差,两腔图像的 LVEF 偏差为 +2.81%,四腔图像的 LVEF 偏差为 +0.11%。
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
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Effect of Data Augmentation on Deep-Learning-Based Segmentation of Long-Axis Cine-MRI
Automated Cardiac Magnetic Resonance segmentation serves as a crucial tool for the evaluation of cardiac function, facilitating faster clinical assessments that prove advantageous for both practitioners and patients alike. Recent studies have predominantly concentrated on delineating structures on short-axis orientation, placing less emphasis on long-axis representations due to the intricate nature of structures in the latter. Taking these consideration into account, we present a robust hierarchy-based augmentation strategy coupled with the compact and fast Efficient-Net (ENet) architecture for the automated segmentation of two-chamber and four-chamber Cine-MRI images. We observed an average Dice improvement of 0.99% on the two-chamber images and of 2.15% on the four-chamber images, and an average Hausdorff distance improvement of 21.3% on the two-chamber images and of 29.6% on the four-chamber images. The practical viability of our approach was validated by computing clinical metrics such as the Left Ventricular Ejection Fraction (LVEF) and left ventricular volume (LVC). We observed acceptable biases, with a +2.81% deviation on the LVEF for the two-chamber images and a +0.11% deviation for the four-chamber images.
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
Algorithms
Algorithms Mathematics-Numerical Analysis
CiteScore
4.10
自引率
4.30%
发文量
394
审稿时长
11 weeks
期刊最新文献
Specification Mining Based on the Ordering Points to Identify the Clustering Structure Clustering Algorithm and Model Checking Personalized Advertising in E-Commerce: Using Clickstream Data to Target High-Value Customers Navigating the Maps: Euclidean vs. Road Network Distances in Spatial Queries Hybrid Sparrow Search-Exponential Distribution Optimization with Differential Evolution for Parameter Prediction of Solar Photovoltaic Models Particle Swarm Optimization-Based Unconstrained Polygonal Fitting of 2D Shapes
×
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
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 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 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1