神经外科导航系统基于迭代局部最近点法技术的形态学三维数据场的快速配准与融合  

Auto and rapid medical image registration for image guided neurosurgery system based on iterative local closest point

作　　者：穆晓兰[1] 王满宁[1] 宋志坚[1] 

机构地区：[1]复旦大学数字医学研究中心,上海200032

出　　处：《解剖学杂志》2005年第4期395-396,399,F0002,共4页Chinese Journal of Anatomy

基　　金：国家自然科学基金(30370417)

摘　　要：目的:拟开发一种全自动快速配准技术,用于临床形态学三维数据场的配准与融合。方法:用本研究提出的基于迭代局部最近点法(ILCP)技术的形态学三维数据场的快速配准技术,用临床实际病例的CT、MRI图像进行配准实验。结果:实验得到了满意的全自动快速配准效果,结果显示ILCP法比普通ICP法快10-30倍,比最大互信息法快8-10倍;耗时可以为临床医生所接受。结论:所提出的方法是切实可行的,为进一步应用于临床打下了坚实的基础。To develop an auto and rapid registration algorithm for registration and fusion of clinical morphological 3-dimensional data volume. Methods： An auto and rapid medical image registration system for morphologic 3-dimensional data algorithm was established based on the ILCP, and examined with CT, MRI images in clinical patients. Results： The experiments showed that CT and MRI images were registered within a clinically acceptable time and results had a fairly good accuracy. ILCP was 10 to 30 times faster than that of global search ICP, and 8 to 10 times fast than that of MMI. Conclusion： The method is successfully established and can be accepted by doctors in clinic. And it should be studied further for clinical application.

关 键 词：神经外科导航系统 迭代局部最近点法技术 形态学 三维数据场 配准方法 融合方法 

分 类 号：R-39[医药卫生]