从CT体数据场提取人体器官的方法  

Human Organ Extraction Method from CT Volumetric Data

作　　者：田沄[1] 张莉华[2] 周明全[1] 武仲科[1] 

机构地区：[1]北京师范大学信息科学与技术学院,北京100875 [2]黄淮学院计算机科学系,驻马店463000

出　　处：《中国生物医学工程学报》2009年第1期37-41,共5页Chinese Journal of Biomedical Engineering

基　　金：国家自然科学基金重点项目(60736008);国家自然科学基金资助项目(60673100/F020106)

摘　　要：改进传统窗口调节函数,结合模糊集理论和三维区域生长技术,提出并实现一种采用CT体数据场的人体器官提取新方法。首先在DICOM图像的原始数据空间域中,将窗口调节至使对应组织器官的灰度信息和形状信息的显示最为充分;然后,运用模糊集理论对其进行模糊增强,最后对增强后的结果运用区域生长法将器官提取出来。实验表明本方法能够有效提取出多数的人体器官。不仅提取结果细节清晰,而且分割成功率达75%以上,同时该方法具有较强的鲁棒性。A novel window adjustment function was developed for organ extraction from CT volumetric data. Based on the adjustment function, combining to fuzzy sets theory and 3D region growing technology, an approach for organ extraction from the CT human volumetric data was developed. Firstly, to expose the organs of interest fully, the window was adjusted appropriately in the raw data space of DICOM images; and then the volumetric data contrast was enhanced using fuzzy sets; Lastly, the organs were extracted using 3D region growing algorithms. Experimental results showed that the proposed method could effectively extract most human organs and was also robust. The extraeted organs were in details with the accuracy rate no less than 75 %.

关 键 词：体数据场 组织器官 窗口调节 对比度增强 三维区域生长 

分 类 号：TN911.73[电子电信—通信与信息系统]