CrossMoDA 2021 challenge: Benchmark of cross-modality domain adaptation techniques for vestibular schwannoma and cochlea segmentation.
CrossMoDA 2021 challenge: Benchmark of cross-modality domain adaptation techniques for vestibular schwannoma and cochlea segmentation.
复制标题
DOI：
10.1016/j.media.2022.102628
复制
发表时间：
2023-01
影响因子：
10.9
通讯作者：
Vercauteren, Tom
中科院分区：
文献类型：
--
作者：
Dorent, Reuben;Kujawa, Aaron;Ivory, Marina;Bakas, Spyridon;Rieke, Nicola;Joutard, Samuel;Glocker, Ben;Cardoso, Jorge;Modat, Marc;Batmanghelich, Kayhan;Belkov, Arseniy;Calisto, Maria Baldeon;Choi, Jae Won;Dawant, Benoit M.;Dong, Hexin;Escalera, Sergio;Fan, Yubo;Hansen, Lasse;Heinrich, Mattias P.;Joshi, Smriti;Kashtanova, Victoriya;Kim, Hyeon Gyu;Kondo, Satoshi;Kruse, Christian N.;Lai-Yuen, Susana K.;Li, Hao;Liu, Han;Ly, Buntheng;Oguz, Ipek;Shin, Hyungseob;Shirokikh, Boris;Su, Zixian;Wang, Guotai;Wu, Jianghao;Xu, Yanwu;Yao, Kai;Zhang, Li;Ourselin, Sebastien;Shapey, Jonathan;Vercauteren, Tom
领域自适应（DA）近年来在医学影像界引起了极大的兴趣。虽然已经提出了大量用于图像分割的数据挖掘技术，但这些技术中的大多数已经在私有数据集或小型公开数据集上进行了验证。此外，这些数据集主要解决单类问题。为了解决这些限制，跨模态域适应（crossMoDA）挑战与第24届医学图像计算和计算机辅助干预国际会议（MICCAI 2021）一起组织。crosssmoda是第一个用于无监督跨模态域自适应的大型多类基准。挑战的目标是分割涉及前庭神经鞘瘤（VS）随访和治疗计划的两个关键脑结构：VS和耳蜗。目前，VS患者的诊断和监测通常采用对比增强T1 (ceT1) MR成像。然而，人们对使用非对比成像序列（如高分辨率T2 （hrT2）成像）越来越感兴趣。为此，我们建立了一个无监督的跨模态分割基准。训练数据集提供带注释的ceT1扫描（N=105）和未配对的未注释的hrT2扫描（N=105）。目的是根据测试集（N=137）提供的hrT2扫描自动执行单侧VS和双侧耳蜗分割。考虑到模态之间的大强度分布差距和结构的小体积，这个问题尤其具有挑战性。共有来自16个国家的55个团队向验证排行榜提交了预测。其中，来自9个不同国家的16个团队提交了他们的算法进入评估阶段。表现最好的团队所达到的表现水平非常高（最佳Dice得分中位数- VS: 88.4%; Cochleas: 85.7%），并且接近于完全监督（Dice得分中位数- VS: 92.5%; Cochleas: 87.7%）。所有性能最好的方法都使用图像到图像的转换方法，将源域图像转换为伪目标域图像。然后使用这些生成的图像和为源图像提供的手动注释来训练分割网络。无监督跨模态域自适应的第一个大型多类基准。前庭神经鞘瘤的两种结构分割随访及治疗方案。源域（对比后T1）和目标域（T2）之间存在较大的域间隙。对来自9个国家的16支队伍提出的技术进行了广泛的比较。表现最好的团队使用图像到图像的转换来弥合领域差距。
Domain Adaptation (DA) has recently been of strong interest in the medical imaging community. While a large variety of DA techniques have been proposed for image segmentation, most of these techniques have been validated either on private datasets or on small publicly available datasets. Moreover, these datasets mostly addressed single-class problems. To tackle these limitations, the Cross-Modality Domain Adaptation (crossMoDA) challenge was organised in conjunction with the 24th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2021). CrossMoDA is the first large and multi-class benchmark for unsupervised cross-modality Domain Adaptation. The goal of the challenge is to segment two key brain structures involved in the follow-up and treatment planning of vestibular schwannoma (VS): the VS and the cochleas. Currently, the diagnosis and surveillance in patients with VS are commonly performed using contrast-enhanced T1 (ceT1) MR imaging. However, there is growing interest in using non-contrast imaging sequences such as high-resolution T2 (hrT2) imaging. For this reason, we established an unsupervised cross-modality segmentation benchmark. The training dataset provides annotated ceT1 scans (N=105) and unpaired non-annotated hrT2 scans (N=105). The aim was to automatically perform unilateral VS and bilateral cochlea segmentation on hrT2 scans as provided in the testing set (N=137). This problem is particularly challenging given the large intensity distribution gap across the modalities and the small volume of the structures. A total of 55 teams from 16 countries submitted predictions to the validation leaderboard. Among them, 16 teams from 9 different countries submitted their algorithm for the evaluation phase. The level of performance reached by the top-performing teams is strikingly high (best median Dice score — VS: 88.4%; Cochleas: 85.7%) and close to full supervision (median Dice score — VS: 92.5%; Cochleas: 87.7%). All top-performing methods made use of an image-to-image translation approach to transform the source-domain images into pseudo-target-domain images. A segmentation network was then trained using these generated images and the manual annotations provided for the source image. The first large and multi-class benchmark for unsupervised cross-modality Domain Adaptation. Segmentation of two structures for the follow-up and treatment planning of vestibular schwannoma. Large domain gap between the source (post-contrast T1) and target (T2) domains. An extensive comparison of the techniques proposed by 16 teams from 9 countries. Top-performing teams used image-to-image translation to bridge the domain gap.
登录
查看更多内容
影响因子：
4.4
作者：
Clark, Kenneth;Vendt, Bruce;Prior, Fred
通讯作者：
Prior, Fred
使用生成对抗网络进行心脏图像分割的双向跨模态无监督域适应
影响因子：
7.7
作者：
Cui, Hengfei;Chang Yuwen;Zhang, Yanning
通讯作者：
Zhang, Yanning
DOI：
10.1016/j.isprsjprs.2020.01.013
发表时间：
2020-04-01
影响因子：
12.7
作者：
Diakogiannis, Foivos, I;Waldner, Francois;Wu, Chen
通讯作者：
Wu, Chen
影响因子：
48
作者：
Isensee, Fabian;Jaeger, Paul F.;Maier-Hein, Klaus H.
通讯作者：
Maier-Hein, Klaus H.
影响因子：
10.9
作者：
Kavur, A. Emre;Gezer, N. Sinem;Selver, M. Alper
通讯作者：
Selver, M. Alper