基于ITK的Mattes互信息医学图像配准研究  

Research of Mattes Mutual Information Medical Image Registration Based on ITK

作　　者：张密[1] 吴效明[1] 周凌宏[2] 

机构地区：[1]华南理工大学生物医学工程系,广州510006 [2]南方医科大学生物医学工程学院,广州510515

出　　处：《微计算机信息》2010年第8期189-191,共3页Control & Automation

基　　金：基金申请人:吴效明;项目名称:基于无线穿戴式检测技术的社区数字医疗健康服系统;基金颁发部门:广东省科技计划基金委(2007B031302003)

摘　　要：近年来非刚性医学图像配准在临床上的需求越来越明显,而使用互信息的方法已经受到许多学者的关注。本文较为详细的介绍了医学图像分割与配准算法工具包在VS2008中的安装和使用,基于该工具包完成了2D-2DMRI脑部图像的Mat-tes互信息配准,对互信息算法及其实现做了研究,并提出了使用图像金字塔算法策略联合互信息测度进行配准的方案。实验表明,与没有使用联合策略的配准方法相比,该方案在配准鲁棒性没有受到明显影响的同时,提高了配准速度和精度。The requirement of non-rigid medical image registration is more and more obvious recently, and there are many researchers pay attention to the area.This paper employed ITK（the Insight Segmentation and Registration Toolkit） to finish the registration of 2D-2D MRI brain image based on Mutual information in Visual Studio 2008, studied Mutual information algorithms and its realization, and introduced the basic processes and key elements of medical image registration in ITK.ITK which is based on visual studio in this paper acts excellently,the solution can satisfy the study demand of related researchers.This paper combined image pyramid algorithm and Mattes＇ methods to realize the mutual information registration,compared to other methods,it can achieve better precision and higher efficiency.

关 键 词：互信息 医学图像分割与配准算法工具包 图像金字塔算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]