Learning deep similarity metric for 3D MR-TRUS image registration.
Learning deep similarity metric for 3D MR-TRUS image registration.
复制标题
DOI：
10.1007/s11548-018-1875-7
复制
发表时间：
2019-03
影响因子：
3
通讯作者：
Yan P
中科院分区：
文献类型：
--
作者：
Haskins G;Kruecker J;Kruger U;Xu S;Pinto PA;Wood BJ;Yan P
经直肠超声 (TRUS) 和磁共振 (MR) 图像的融合用于指导靶向前列腺活检，显着提高了侵袭性癌症的活检产量。 MR-TRUS 融合的一个关键组成部分是图像配准。然而，由于两种成像方式之间存在较大的外观差异，获得强大的自动 MR-TRUS 配准非常具有挑战性。本文提出的工作旨在通过解决两个挑战来解决这个问题：（i）合适的相似性度量的定义和（ii）合适的优化策略的确定。这项工作提出使用深度卷积神经网络来学习 MR-TRUS 配准的相似性度量。我们还使用复合优化策略来探索解空间，以便为学习指标的二阶优化寻找合适的初始化。此外，使用多遍方法来平滑度量以进行优化。学习到的相似性度量优于经典的互信息以及最先进的基于 MIND 特征的方法。结果表明，整体配准框架具有较大的捕获范围。对于这个具有挑战性的问题，所提出的基于深度相似度度量的方法获得了 3.86mm 的平均 TRE（初始 TRE 为 16mm）。使用深度神经网络学习的相似性度量可用于评估任何给定图像配准的质量，并且可与上述优化框架结合使用以执行对不良初始化具有鲁棒性的自动配准。
The fusion of transrectal ultrasound (TRUS) and magnetic resonance (MR) images for guiding targeted prostate biopsy has significantly improved the biopsy yield of aggressive cancers. A key component of MR–TRUS fusion is image registration. However, it is very challenging to obtain a robust automatic MR–TRUS registration due to the large appearance difference between the two imaging modalities. The work presented in this paper aims to tackle this problem by addressing two challenges: (i) the definition of a suitable similarity metric and (ii) the determination of a suitable optimization strategy. This work proposes the use of a deep convolutional neural network to learn a similarity metric for MR–TRUS registration. We also use a composite optimization strategy that explores the solution space in order to search for a suitable initialization for the second-order optimization of the learned metric. Further, a multi-pass approach is used in order to smooth the metric for optimization. The learned similarity metric outperforms the classical mutual information and also the state-of-the-art MIND feature-based methods. The results indicate that the overall registration framework has a large capture range. The proposed deep similarity metric-based approach obtained a mean TRE of 3.86mm (with an initial TRE of 16mm) for this challenging problem. A similarity metric that is learned using a deep neural network can be used to assess the quality of any given image registration and can be used in conjunction with the aforementioned optimization framework to perform automatic registration that is robust to poor initialization.
登录
查看更多内容
MR/超声融合引导的活检与超声引导活检的比较，以诊断前列腺癌。
DOI：
10.1001/jama.2014.17942
发表时间：
2015-01-27
期刊：
JAMA
影响因子：
--
作者：
Siddiqui MM;Rais-Bahrami S;Turkbey B;George AK;Rothwax J;Shakir N;Okoro C;Raskolnikov D;Parnes HL;Linehan WM;Merino MJ;Simon RM;Choyke PL;Wood BJ;Pinto PA
通讯作者：
Pinto PA
影响因子：
4.6
作者：
Cheng X;Boza-Serrano A;Turesson MF;Deierborg T;Ekblad E;Voss U
通讯作者：
Voss U
影响因子：
10.6
作者：
Khallaghi, Siavash;Sanchez, C. Antonio;Abolmaesumi, Purang
通讯作者：
Abolmaesumi, Purang
影响因子：
4.8
作者：
Calio, B.;Sidana, A.;Pinto, P.
通讯作者：
Pinto, P.
DOI：
10.1016/j.juro.2011.05.078
发表时间：
2011-10
期刊：
The Journal of urology
影响因子：
--
作者：
Pinto PA;Chung PH;Rastinehad AR;Baccala AA Jr;Kruecker J;Benjamin CJ;Xu S;Yan P;Kadoury S;Chua C;Locklin JK;Turkbey B;Shih JH;Gates SP;Buckner C;Bratslavsky G;Linehan WM;Glossop ND;Choyke PL;Wood BJ
通讯作者：
Wood BJ