CT系列图像的三维层次化子块生长分割

CT image segmentation based on three-dimensional hierarchical sub-block region growing

准确地从CT系列图像提取感兴趣的组织是手术规划的基础,针对肝脏轮廓分割存在分割不全的问题,提出了基于三维区域生长算法的腹部CT图像分割方法。算法首先由用户选择若干个生长点,然后充分利用CT系列图像层间的相似性,提出基于子块的改进区域生长算法,实现三维的层次化子块区域生长,以更准确提取肝脏区域,其中生长准则由系统分析用户选择的生长点的邻域子块属性获得,以减少用户的干预。实验结果表明,算法能在较少的干预下快速分割出来CT系列图像中的肝脏轮廓。

Segmentation of region of interest based on abdominal Computed Tomography（CT） sequences images is a crucial step in surgical planning.However,precisely carrying out this step remains a challenge.By considering the situation of keeping the wrong boundary as the liver surface,an approach for CT image segmentation based on 3D hierarchical region growing algorithm is proposed.In the approach,several seed points are selected firstly.Considering the similarity of the neighbor slices in the medical volume data,a modified 3D sub-block region growing algorithm is developed.In order to reduce the user interaction,the growing criteria are estimated automatically through investigation of the statistical characteristics in the local regions of the seed points.Experiments results show the proposed method can efficiently segment the live region from serial abdominal CT images with little user interaction.