有源平均场：在水平集框架中求解平均场近似
Active mean fields: solving the mean field approximation in the level set framework.
作者信息
Pohl Kilian M, Kikinis Ron, Wells William M
机构信息
Surgical Planning Laboratory, Harvard Medical School and Brigham and Women's Hospital, Boston, MA, USA.
出版信息
Inf Process Med Imaging. 2007;20:26-37. doi: 10.1007/978-3-540-73273-0_3.
Abstract
We describe a new approach for estimating the posterior probability of tissue labels. Conventional likelihood models are combined with a curve length prior on boundaries, and an approximate posterior distribution on labels is sought via the Mean Field approach. Optimizing the resulting estimator by gradient descent leads to a level set style algorithm where the level set functions are the logarithm-of-odds encoding of the posterior label probabilities in an unconstrained linear vector space. Applications with more than two labels are easily accommodated. The label assignment is accomplished by the Maximum A Posteriori rule, so there are no problems of "overlap" or "vacuum". We test the method on synthetic images with additive noise. In addition, we segment a magnetic resonance scan into the major brain compartments and subcortical structures.
摘要
我们描述了一种估计组织标签后验概率的新方法。传统的似然模型与边界上的曲线长度先验相结合，并通过平均场方法寻求标签上的近似后验分布。通过梯度下降优化所得估计器会得到一种水平集风格的算法，其中水平集函数是无约束线性向量空间中后验标签概率的对数优势编码。该方法可轻松应用于具有两个以上标签的情况。标签分配通过最大后验规则完成，因此不存在“重叠”或“空白”问题。我们在带有加性噪声的合成图像上测试了该方法。此外，我们将磁共振扫描分割成主要的脑区隔和皮质下结构。
相似文献
1
Active mean fields: solving the mean field approximation in the level set framework.有源平均场：在水平集框架中求解平均场近似
Inf Process Med Imaging. 2007;20:26-37. doi: 10.1007/978-3-540-73273-0_3.
2
A marginalized MAP approach and EM optimization for pair-wise registration.一种用于成对配准的边缘化最大后验概率方法和期望最大化优化
Inf Process Med Imaging. 2007;20:662-74. doi: 10.1007/978-3-540-73273-0_55.
3
A log-Euclidean framework for statistics on diffeomorphisms.一种用于微分同胚统计的对数欧几里得框架。
Med Image Comput Comput Assist Interv. 2006;9(Pt 1):924-31. doi: 10.1007/11866565_113.
4
Geometry driven volumetric registration.几何驱动的体积配准
Inf Process Med Imaging. 2007;20:675-86. doi: 10.1007/978-3-540-73273-0_56.
5
A unifying approach to registration, segmentation, and intensity correction.一种用于配准、分割和强度校正的统一方法。
Med Image Comput Comput Assist Interv. 2005;8(Pt 1):310-8. doi: 10.1007/11566465_39.
6
A general framework for image segmentation using ordered spatial dependency.一种使用有序空间依赖性进行图像分割的通用框架。
Med Image Comput Comput Assist Interv. 2006;9(Pt 2):848-55. doi: 10.1007/11866763_104.
7
Many heads are better than one: jointly removing bias from multiple MRIs using nonparametric maximum likelihood.三个臭皮匠，顶个诸葛亮：使用非参数最大似然法联合去除多个磁共振成像中的偏差。
Inf Process Med Imaging. 2005;19:615-26. doi: 10.1007/11505730_51.
8
Using the logarithm of odds to define a vector space on probabilistic atlases.使用对数优势在概率图谱上定义向量空间。
Med Image Anal. 2007 Oct;11(5):465-77. doi: 10.1016/j.media.2007.06.003. Epub 2007 Jun 22.
9
Prior knowledge driven multiscale segmentation of brain MRI.基于先验知识的脑磁共振成像多尺度分割
Med Image Comput Comput Assist Interv. 2007;10(Pt 2):118-26. doi: 10.1007/978-3-540-75759-7_15.
10
Coupled shape distribution-based segmentation of multiple objects.基于耦合形状分布的多目标分割
Inf Process Med Imaging. 2005;19:345-56. doi: 10.1007/11505730_29.
引用本文的文献
1
An Optimal, Generative Model for Estimating Multi-Label Probabilistic Maps.用于估计多标签概率图谱的最优生成模型。
IEEE Trans Med Imaging. 2020 Jul;39(7):2316-2326. doi: 10.1109/TMI.2020.2968917. Epub 2020 Jan 23.
2
ACTIVE MEAN FIELDS FOR PROBABILISTIC IMAGE SEGMENTATION: CONNECTIONS WITH CHAN-VESE AND RUDIN-OSHER-FATEMI MODELS.用于概率图像分割的主动平均场：与Chan-Vese模型和Rudin-Osher-Fatemi模型的联系
SIAM J Imaging Sci. 2017;10(3):1069-1103. doi: 10.1137/16M1058601. Epub 2017 Jul 27.
3
OPTIMAL PARAMETER MAP ESTIMATION FOR SHAPE REPRESENTATION: A GENERATIVE APPROACH.用于形状表示的最优参数映射估计：一种生成式方法。
Proc IEEE Int Symp Biomed Imaging. 2016 Apr;2016:660-663. doi: 10.1109/ISBI.2016.7493353. Epub 2016 Jun 16.
4
A Multiple Object Geometric Deformable Model for Image Segmentation.一种用于图像分割的多目标几何可变形模型
Comput Vis Image Underst. 2013 Feb 1;117(2):145-157. doi: 10.1016/j.cviu.2012.10.006.
5
A Multi-Compartment Segmentation Framework With Homeomorphic Level Sets.一种具有同胚水平集的多隔室分割框架。
Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2008:1-6. doi: 10.1109/CVPR.2008.4587475.
6
A Multiple Geometric Deformable Model Framework for Homeomorphic 3D Medical Image Segmentation.一种用于同胚三维医学图像分割的多重几何可变形模型框架。
Conf Comput Vis Pattern Recognit Workshops. 2008 Jul 15;2008:1-7. doi: 10.1109/CVPRW.2008.4563013.
7
Segmentation of image ensembles via latent atlases.通过潜在图谱对图像集进行分割。
Med Image Anal. 2010 Oct;14(5):654-65. doi: 10.1016/j.media.2010.05.004. Epub 2010 Jun 4.
8
A hybrid geometric-statistical deformable model for automated 3-D segmentation in brain MRI.一种用于脑磁共振成像中自动三维分割的混合几何-统计可变形模型。
IEEE Trans Biomed Eng. 2009 Jul;56(7):1838-48. doi: 10.1109/TBME.2009.2017509. Epub 2009 Mar 27.
9
Validation of a fully automated 3D hippocampal segmentation method using subjects with Alzheimer's disease mild cognitive impairment, and elderly controls.使用患有阿尔茨海默病轻度认知障碍的受试者和老年对照对一种全自动三维海马体分割方法进行验证。
Neuroimage. 2008 Oct 15;43(1):59-68. doi: 10.1016/j.neuroimage.2008.07.003. Epub 2008 Jul 16.
本文引用的文献
1
Logarithm odds maps for shape representation.用于形状表示的对数几率图。
Med Image Comput Comput Assist Interv. 2006;9(Pt 2):955-63. doi: 10.1007/11866763_117.
2
User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability.用户引导的解剖结构三维主动轮廓分割：显著提高效率和可靠性。
Neuroimage. 2006 Jul 1;31(3):1116-28. doi: 10.1016/j.neuroimage.2006.01.015. Epub 2006 Mar 20.
3
A Bayesian model for joint segmentation and registration.一种用于联合分割与配准的贝叶斯模型。
Neuroimage. 2006 May 15;31(1):228-39. doi: 10.1016/j.neuroimage.2005.11.044. Epub 2006 Feb 7.
4
Neighbor-constrained segmentation with level set based 3-D deformable models.基于水平集的三维可变形模型的邻域约束分割
IEEE Trans Med Imaging. 2004 Aug;23(8):940-8. doi: 10.1109/TMI.2004.830802.
5
An adaptive level set segmentation on a triangulated mesh.三角网格上的自适应水平集分割。
IEEE Trans Med Imaging. 2004 Feb;23(2):191-201. doi: 10.1109/TMI.2003.822823.
6
Automatically parcellating the human cerebral cortex.自动分割人类大脑皮层。
Cereb Cortex. 2004 Jan;14(1):11-22. doi: 10.1093/cercor/bhg087.
7
A shape-based approach to the segmentation of medical imagery using level sets.一种基于水平集的医学图像分割的形状基方法。
IEEE Trans Med Imaging. 2003 Feb;22(2):137-54. doi: 10.1109/TMI.2002.808355.
8
Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain.全脑分割：人脑神经解剖结构的自动标记
Neuron. 2002 Jan 31;33(3):341-55. doi: 10.1016/s0896-6273(02)00569-x.
9
Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm.通过隐马尔可夫随机场模型和期望最大化算法对脑部磁共振图像进行分割。
IEEE Trans Med Imaging. 2001 Jan;20(1):45-57. doi: 10.1109/42.906424.
10
Automated model-based tissue classification of MR images of the brain.基于模型的脑磁共振图像自动组织分类
IEEE Trans Med Imaging. 1999 Oct;18(10):897-908. doi: 10.1109/42.811270.
Zotero 插件