基于模糊聚类和C-V模型的水平集方法图像分割

摘要

To propose a coronary artery segmentation model based on level set method for MSCT.First,the CT data were dealt to gain clustering information and membership matrix using fuzzy clustering method.Secondly,the initial contour of level set method was defined using the clustering information.Finally the coronary artery was extracted using C-V model and the image segmentation was completed.Experimental results showed that,compared to conventional methods,the proposed method which had higher precision increased overlap rate by 11.82 ％,therefore can be used for assist clinical diagnosis of coronary heart disease.%提出一种基于水平集方法的MSCT冠状动脉分割模型.首先在CT数据中进行模糊聚类的预处理,提取聚类信息和隶属度矩阵；然后应用聚类信息指导水平集方法初始轮廓的选取,用C-V模型进行冠状动脉的提取,完成图像分割.实验结果表明,上述方法具有较高的分割精度,分割重叠率相比传统方法提高了11.82％,可用于临床辅助诊断冠状动脉性心脏病.