具有全数据或部分数据的多模态医学图像配准  

Multimodal medical image registration with full or partial data

作　　者：李正伟 LI Zhengwei(The Engineering&Technical College of Chengdu University of Technology,Leshan 614007,China)

机构地区：[1]成都理工大学工程技术学院,四川乐山614007

出　　处：《电子设计工程》2021年第14期172-179,共8页Electronic Design Engineering

摘　　要：多模态图像配准是整合两幅或多幅图像信息的主要步骤。图像之间除了强度变化和结构差异外,可能有部分或全部重叠,这给配准过程的成功增加了额外的障碍。该文提出了一种多模态到单模态的变换方法,以便在全数据和部分数据重叠的情况下获得多模态图像的精确配准。使用全数据的模拟和临床人脑图像检验了该方法的有效性,并与广泛使用的基于互信息(MI)的技术进行了比较。利用RIRE数据集,对CT图像分别进行PD、T1和T2-MRI配准,得到了1.37 mm、1.00 mm和1.41 mm的平均绝对误差。对所提出的变换在多模态部分重叠图像配准中的效果进行了实证研究。Multimodal image registration is the main step of synthesizing two or more image information.In addition to the intensity changes and structural differences between images,they may overlap partially or completely,which adds additional obstacles to the success of the registration process.In this paper,a multi-modal to single-mode transformation method is proposed to achieve the accurate registration of multi-modal images when the whole data and part of the data overlap.The validity of this method is tested by using the full data simulation and clinical human brain image,and compared with the widely used technology based on information theory.Using the RIRE data set,PD,T1 and T2-MRI images were registered respectively,and the mean absolute errors of 1.37 mm,1.00 mm and 1.41 mm were obtained.The effect of the proposed transformation in the registration of multimodal partially overlapped images is studied.

关 键 词：图像配准 多模态 部分重叠图像 流形学习 

分 类 号：TN919.8[电子电信—通信与信息系统]