摘要
本文提出了有效的、能被临床应用所接受的磁共振 (MR)和CT医学图像配准方法。在基于体素灰度的医学图像配准领域 ,本文采用了全新的相关比相似性测度作为配准的测度准则。具体设计时 ,采用了加速的多分辨率配准方案 ,对方案中涉及的几何变换选取、重采样、多分辨率体数据表达及最优化方法进行了设计分析。最后 ,利用本文提出的多分辨率配准方法 ,对MR和CT临床医学图像进行配准 ,给出了令人满意的效果。
An effective multi-resolution MR and CT images registration method designed for clinical use is presented. A new correlation ratio similarity measure in voxel intensity based medical image registration field was adopted. The rigid transformation, resampling, multi-resolution medical image representation and optimization methods for accelerated multi-resolution registration scheme were discussed in detail. Experimental results based on clinical MR and CT images show that correlation ratio based multi-resolution registration method works well during clinical application.
出处
《中国生物医学工程学报》
EI
CAS
CSCD
北大核心
2003年第1期1-5,11,共6页
Chinese Journal of Biomedical Engineering
基金
上海市科学发展基金资助项目 ( 985 10 70 16 )
关键词
医学图像配准
相关比
多分辨率
相似性测度
系统设计
Computerized tomography
Correlation methods
Magnetic resonance imaging
Medical imaging