肝脏分割的种子点自动获取方法  

Automatic Seed Point Selection for Liver Segmentation

作　　者：宋晓[1] 贾松浩[1] 黄晓阳[2] 程明[2] 黄绍辉[2] 王博亮[2] 

机构地区：[1]南阳师范学院计算机与信息技术学院,南阳473061 [2]厦门大学信息科学与技术学院,厦门361005

出　　处：《系统仿真学报》2013年第9期2075-2078,共4页Journal of System Simulation

基　　金：国家自然科学基金(61001144;61102137;61202011);河南省基础与前沿技术研究项目(122300410258);南阳师范学院专项项目(ZX2013012)

摘　　要：为了解决传统手工选取种子点影响肝脏自动分割算法稳定性的问题,提出了一种基于灰度统计模型的自动种子点获取方法。该方法根据肝脏的结构特性和灰度特性,将种子点获取问题转化为肝脏的识别和定位问题,通过对肝脏CT图像序列集的灰度统计和分析,进而快速准确的定位到肝内种子点。整个过程无需用户交互,保证了种子点不会因操作人员不同而发生位置变化。In order to solve the problem of traditional manual method for seed point selection in liver segmentation process, an automatic seed point selection method was proposed based on statistic model of intensity. This method is according to the characteristics of structure and grayscale of the liver. It transfers the problem of obtaining seed point to the identification and location of the liver. By means of statistics and analysis of liver intensity on the sequence of CT images, the seed point can be obtained quickly and accurately. The whole process has no user interaction. And it ensures the position of seed point will not change by different operators.

关 键 词：肝脏分割 种子点 灰度统计模型 CT图像 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]