医学图像配准中的一种互信息快速计算方法  

Fast calculation method of mutual information in medical image registration

作　　者：杜俊俐[1] 邱道尹[1] 朱齐亮[2] 黄心汉[1] 

机构地区：[1]华中科技大学控制科学与工程系 [2]华北水利水电学院信息工程学院,郑州450011

出　　处：《计算机工程与应用》2008年第22期180-182,188,共4页Computer Engineering and Applications

基　　金：河南省科技公关计划No.0624220100~~

摘　　要：基于互信息的医学图像配准,可以达到亚像素级精度且无需提取图像的解剖特征,是一种高精稳健的配准方法。但其中频繁的互信息计算使配准速度很慢,不能满足临床的实时要求。提出一种基于统计直方图灰度压缩的互信息加速计算方法,能显著减少灰度级,加快互信息计算,从而加快配准速度。实验表明,该方法能在不影响配准精度的前提下显著缩短配准时间。The medical image registration based on mutual information can reach sub-pixel accuracy and don＇t need to extract anatomical features from images,is a high-precision sound registration way.But frequent involving a large number of floating-point calculation of the mutual information,the registration rate is slow,cart not meet the demands of real-time.This paper presents a statistical gray histogram mutual information compression method can significantly reduce the amount of computation mutual information,thus speeding up the registration rate.The experiments show that the method can reduce the registration time under the premise not affect the accuracy of registration.

关 键 词：医学图像配准 互信息 灰度压缩 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]