计算机与现代化 ›› 2011, Vol. 1 ›› Issue (1): 113-4.doi: 10.3969/j.issn.1006-2475.2011.01.033

• 图像处理 • 上一篇    下一篇

基于Hessian矩阵和区域生长的肝血管树的分割算法研究

刘晏丽1,2,赵卫东1,2,陈宇飞1,2,王志成1,2,孟齐文1,2   

  1. 1.同济大学企业数字化技术教育部工程研究中心,上海 200092; 2.同济大学CAD研究中心,上海 201804
  • 收稿日期:2010-09-01 修回日期:1900-01-01 出版日期:2011-01-20 发布日期:2011-01-20

Research on Liver Vessel Tree Segmentation Algorithm Using Hessian Matrix and Region Growing

LIU Yan-li1,2, ZHAO Wei-dong1,2, CHEN Yu-fei1,2, WANG Zhi-cheng1,2, MENG Qi-wen1,2   

  1. 1.Engineering Research Center for Enterprise Digital Technology, Ministry of Education, Tongji University, Shanghai 200092, China;2.Research Center of CAD, Tongji University, Shanghai 201804, China
  • Received:2010-09-01 Revised:1900-01-01 Online:2011-01-20 Published:2011-01-20

摘要:

由于血管成像设备的限制和血管本身形态结构的复杂性,使得血管分割成为一个难题。针对CT肝脏血管图像的特征,设计一种将Hessian矩阵多尺度滤波、形态学和三维区域生长相结合的方法,实现血管分割。首先,使用灰度变换突出目标区域,利用基于Hessian矩阵的多尺度滤波方法对血管进行增强;其次,对增强后的血管图像进行闭运算,并进行三维区域生长;最后,利用膨胀运算将血管的弱连接进行连接。实验表明该方法具有一定的可行性和有效性,能够很好地实现血管的分割,所产生的血管连续光滑,很适合作为三维重建的数据集。

关键词: 血管分割, Hessian矩阵, 闭运算, 三维区域生长

Abstract:

Due to the limitation of vascular imaging equipment and complexity of vascular morphology, vessel segmentation has been challenging. A method is proposed to realize 3D vessel segmentation in view of the character of CT liver vasculature, which combines Hessian matrixbased multiscale filter, morphological method and 3D region growing. Firstly, gray transformation is used to enhance region of interest, and the vasculature is enhanced using Hessianmatrixbased multiscale filter. Secondly closing operation and 3D region growing algorithm are utilized on the enhanced images. Finally, the disconnection is connected via dilation operation. Experimental results show that the method can effectively extract vasculature that is suitable for 3D reconstruction owing to their continuity and smoothing.

Key words: vessel segmentation, Hessian matrix, closing operation, 3D region growing