基于区域生长法的医学图像分割  

Medical image segmentation based on region growing arithmetic

作　　者：舒添慧[1] 胥布工[1] 胡战虎[2] 

机构地区：[1]华南理工大学自动化科学与工程学院,广东广州510640 [2]广东省科学院自动化工程研制中心,广东广州510070

出　　处：《微计算机信息》2008年第18期284-285,195,共3页Control & Automation

基　　金：广东省自然科学(05100514)课题名称:基于ROI特征提取和SVM多类分类的医学图像检索;2005年粤港关键领域重点突破招标项目

摘　　要：医学图像分割是CT图像的三维重建等后续操作的基础,分割的准确性对医生判断至关重要。本文深入研究区域生长算法,并用VC++实现了该算法。区域生长法可满足图像分割的多种要求,既提取CT图的边缘信息,又针对某一特定区域的病灶单独实行分割,缩短了手术计划的时间,实验结果为临床诊断提供有利的帮助信息,提高了手术导航的效率。The medical image segmentation is the foundation of the follow-up operation, just like 3D reconstruction...etc, of CT image. Accuracy of segmentation is very important to the doctor＇s judging. This paper mainly narrated the concept of region growing arithmetic, and has realized this algorithm with VC＋＋＋ The arithmetic is more suitable than before for complicate and multiplicity image. It not only can obtain the edge information, but also can make segmentation for special region. The time of surgical planning is be shortened. Experiment tests show that this function improved efficiency in navigation surgical.

关 键 词：图像分割 医学图像 区域生长 手术导航 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]