【作者】 张朝霞； 陈晓冬； 李宏霄； 汪毅； 郁道银； 贾忠伟； 浦奎； 王星；

【Author】 ZHANG Zhaoxia 1,YU Daoyin 1,LI Hongxiao 1,WANG Yi 1,CHEN Xiaodong 1 JIA Zhongwei 2,PU Kui 2,WANG Xing 2 (1.College of Precision Instrument & Opto-electronic Engineering,Tianjin University,Key Laboratory of Opto-electronics Information and Technical Science (Tianjin University),Ministry of Education,Tianjin Optical Society,Tianjin 300072,China) (2.PLA 254 Hospital,Tianjin,300142,China)

【机构】 天津大学精密仪器与光电子工程学院,光电信息技术科学教育部重点实验室； 中国人民解放军第二五四医院；

【摘要】 针对多层螺旋CT(MSCT)中冠状动脉分割时受周围静脉血管等组织的影响而容易发生的泄漏问题,提出了一种基于最佳方向性梯度通量血管增强的分割方法。首先,得到原始图像的梯度向量场,选择合适的半径,计算球面上特定方向上投影梯度的通量,寻找使得流向球体内部的投影通量最小的最佳方向。之后求解最佳方向上的梯度通量矩阵的特征值,利用特征值构造血管相似度响应函数,对冠状动脉进行增强。采用自适应阈值的区域生长方法将冠脉血管分割出来。实验结果表明,该算法受冠脉周围组织的影响较小,避免了泄漏的发生,而且能提取到较多的细小分支。更多还原

【Abstract】 A segmentation algorithm based on optimal oriented flux (OOF) vessel enhancement is proposed to segment coronary arteries in cardiac multislice CT (MSCT) images correctly by eliminating influence of nearby structure such as vein.First,the gradient vector field is calculated,a series of radius are chosen to compute the OOF on the sphere surface.The optimal direction where the inward projected flux is minimal coincides with where the vessel lie in,so the eigenvalues can be calculated with which vessel similarity function will be constructed to enhance the original data.Hereafter,an adaptive threshold region growing is executed to extract the coronary arteries.Experiment results illustrate the algorithm is not easily influenced by the structure around the arteries and thus over-segmentation can be avoid,simultaneously,more thin branches can be obtained.更多还原