A unified variational segmentation framework with a level-set based sparse composite shape prior.
A unified variational segmentation framework with a level-set based sparse composite shape prior.
复制标题
DOI：
10.1088/0031-9155/60/5/1865
复制
发表时间：
2015-03-07
影响因子：
3.5
通讯作者：
Ruan D
中科院分区：
文献类型：
--
作者：
Liu W;Ruan D
图像分割在许多医学应用中起着至关重要的作用。低SNR条件和各种伪影使其自动化具有挑战性。为了实现鲁棒和准确的分割结果，一个很好的方法是引入适当的形状先验。在这项研究中，我们提出了一个统一的变分分割框架，正则化的目标形状与水平集为基础的稀疏复合先验。当用块最小化/decent方案解决变分问题时，稀疏复合先验的正则化影响可以被观察到以调整到最近的形状估计，并且可以被解释为“动态”形状先验，但由于统一的能量框架而不会损害收敛。所提出的方法被应用于分割胼胝体从二维MR图像和肝脏从三维CT卷。采用Dice相似系数和Hausdorff距离对该方法进行了性能评价，并与两种基于水平集的分割方法进行了比较。所提出的方法在两个实验中都取得了统计上显著的更高精度，并且与基准方法相比，避免了具有相似强度的周围结构的错误包含/排除。
Image segmentation plays an essential role in many medical applications. Low SNR conditions and various artifacts makes its automation challenging. To achieve robust and accurate segmentation results, a good approach is to introduce proper shape priors. In this study, we present a unified variational segmentation framework that regularizes the target shape with a level-set based sparse composite prior. When the variational problem is solved with a block minimization/decent scheme, the regularizing impact of the sparse composite prior can be observed to adjust to the most recent shape estimate, and may be interpreted as a “dynamic” shape prior, yet without compromising convergence thanks to the unified energy framework. The proposed method was applied to segment corpus callosum from 2D MR images and liver from 3D CT volumes. Its performance was evaluated using Dice Similarity Coefficient and Hausdorff distance, and compared with two benchmark level-set based segmentation methods. The proposed method has achieved statistically significant higher accuracy in both experiments and avoided faulty inclusion/exclusion of surrounding structures with similar intensities, as opposed to the benchmark methods.
登录
查看更多内容
影响因子：
10.9
作者：
Zhang, Shaoting;Zhan, Yiqiang;Zhou, Xiang Sean
通讯作者：
Zhou, Xiang Sean
影响因子：
10.6
作者：
Chan, TF;Vese, LA
通讯作者：
Vese, LA
影响因子：
19.5
作者：
Cremers, Daniel;Osher, Stanley J.;Soatto, Stefano
通讯作者：
Soatto, Stefano
影响因子：
19.5
作者：
KASS, M;WITKIN, A;TERZOPOULOS, D
通讯作者：
TERZOPOULOS, D
影响因子：
4.7
作者：
Cootes, TF;Taylor, CJ
通讯作者：
Taylor, CJ