Weakly-supervised convolutional neural networks for multimodal image registration.
Weakly-supervised convolutional neural networks for multimodal image registration.
复制标题
DOI：
10.1016/j.media.2018.07.002
复制
发表时间：
2018-10
影响因子：
10.9
通讯作者：
Vercauteren T
中科院分区：
文献类型：
--
作者：
Hu Y;Modat M;Gibson E;Li W;Ghavami N;Bonmati E;Wang G;Bandula S;Moore CM;Emberton M;Ourselin S;Noble JA;Barratt DC;Vercauteren T
多模态图像配准的监督学习的基本挑战之一是缺乏体素级空间对应的基础真值。这项工作描述了一种从解剖标签中包含的更高级别对应信息推断体素级转换的方法。我们认为，对于图像对的参考集，这种标签比体素级对应更可靠和实用。典型的解剖标记可能包括实体器官、血管、导管、结构边界和其他特定主题的特殊标记。提出的端到端卷积神经网络方法旨在预测位移场，以便在训练过程中对齐单个图像对的多个标记对应结构，而仅使用未标记的图像对作为网络输入进行推理。我们强调了所提出策略的多功能性，用于训练，利用不同类型的解剖标签，这些标签不需要在所有训练图像对上被识别。在推理中，生成的3D可变形图像配准算法实时运行，并且完全自动化，无需任何解剖标签或初始化。比较了几种不同的网络结构，用于登记前列腺癌患者的t2加权磁共振图像和三维经直肠超声图像。通过交叉验证实验，对76例患者的108对多模态图像进行了高质量解剖标记测试，获得了里程碑质心的中位目标配准误差为3.6 mm，前列腺的中位Dice为0.87。
One of the fundamental challenges in supervised learning for multimodal image registration is the lack of ground-truth for voxel-level spatial correspondence. This work describes a method to infer voxel-level transformation from higher-level correspondence information contained in anatomical labels. We argue that such labels are more reliable and practical to obtain for reference sets of image pairs than voxel-level correspondence. Typical anatomical labels of interest may include solid organs, vessels, ducts, structure boundaries and other subject-specific ad hoc landmarks. The proposed end-to-end convolutional neural network approach aims to predict displacement fields to align multiple labelled corresponding structures for individual image pairs during the training, while only unlabelled image pairs are used as the network input for inference. We highlight the versatility of the proposed strategy, for training, utilising diverse types of anatomical labels, which need not to be identifiable over all training image pairs. At inference, the resulting 3D deformable image registration algorithm runs in real-time and is fully-automated without requiring any anatomical labels or initialisation. Several network architecture variants are compared for registering T2-weighted magnetic resonance images and 3D transrectal ultrasound images from prostate cancer patients. A median target registration error of 3.6 mm on landmark centroids and a median Dice of 0.87 on prostate glands are achieved from cross-validation experiments, in which 108 pairs of multimodal images from 76 patients were tested with high-quality anatomical labels.
登录
查看更多内容
影响因子：
10.6
作者：
De Silva, Tharindu;Cool, Derek W.;Ward, Aaron D.
通讯作者：
Ward, Aaron D.
影响因子：
10.9
作者：
Dou, Qi;Yu, Lequan;Heng, Pheng-Ann
通讯作者：
Heng, Pheng-Ann
DOI：
10.1007/978-3-319-66182-7_35
发表时间：
2017-09
期刊：
Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
影响因子：
--
作者：
Cao X;Yang J;Zhang J;Nie D;Kim MJ;Wang Q;Shen D
通讯作者：
Shen D
影响因子：
23.4
作者：
Dickinson, Louise;Ahmed, Hashim U.;Emberton, Mark
通讯作者：
Emberton, Mark
DOI：
10.1109/tbme.2016.2582734
发表时间：
2017-04
期刊：
IEEE transactions on bio-medical engineering
影响因子：
--
作者：
Hu Y;Kasivisvanathan V;Simmons LA;Clarkson MJ;Thompson SA;Shah TT;Ahmed HU;Punwani S;Hawkes DJ;Emberton M;Moore CM;Barratt DC
通讯作者：
Barratt DC