13 肺气管树的细分割 - 中国图象图形学报.DOC

中图法分类号：TP391 文献标识码：A 文章编号： 论文引用格式：  结合区域生长与灰度重建的CT图像 肺气管树分割 彭双，肖昌炎 湖南大学电气与信息工程学院, 长沙 410082 摘 要：目的：CT图像中肺气管树分割对肺部疾病的精确定位与量化评估具有重要意义。针对不同气管分枝在大小、形状和密度分布方面的差异，本文提出一种区域生长与形态学灰度重建相结合的三维肺气管树分割流程，重点解决气管提取过程中的局部泄漏问题。方法：首先，采用阈值分割和形态学闭运算提取肺实质以定义感兴趣区域；然后，通过改进迟滞阈值区域生长法分割较粗气管，结合局部体积突变指标抑制侧向泄漏；接着，利用三维形态学灰度重建算法分割较细气管，并采用形状约束连接元分析和管形描述子剔除伪气管区域；最后，将上述两步分割结果融合成完整气管树。结果：算法验证采用EXACT’09竞赛提供的公开数据，本文选取20例测试图像及手工标记结果作为参考，分别从分杈点、分支数量和分支数比率等方面进行量化评估。实验结果表明，本文方法能在较低泄漏前提下成功检测出参考标准中一半以上的分支、平均分支数比率达到55.5%。结论：与竞赛网站公布的其它方法相比，本文方法结果评价指标处于中上水平，但算法简单、计算复杂度低、易于实现，在泄漏检测方面有一定优势。 关键词： 气管树分割；区域生长；形态学重建；CT图像 Segmentation of Pulmonary airway tree in CT images by combining region growing and grayscale reconstruction algorithms Peng Shuang, Xiao Changyan College of Electrical and Information Engineering, Hunan University, Changsha, 410082, China Abstract: Objective: Segmentation of pulmonary airway tree plays an important role in accurate localization and quantitative evaluation of lung diseases. Considering the size, shape and density variety among different bronchus branches, we propose a method for 3D pulmonary airway tree segmentation by combing a region growing and a morphological grayscale reconstruction algorithms, where the main effort focuses on solving the local leakage problems. Method: First, the region of interest or mask is defined by extracting the lung parenchyma with a global threshold and a morphological closing operation; Then, the trachea and big bronchi are segmented using an improved region growing method on basis of an iterative hysteresis threshold, and a local volume explosion index is adopted to suppress the lateral leakage; The smaller bronchi are segmented using a 3D morphological grayscale reconstruction algorithm, and a connected component based shape descriptor is defined to remove the pseudo-tracheal regions. Finally, a complete airway tree is obtained by integrating segmentation results from the above two steps. Result: The algorithms are tested with the publicly av