用于三维医学图像配准的“粗精”混合算法研究  

A‘coarse and fine'hybrid algorithm for three-dimensional medical image registration

作　　者：翁飞[1] 侯文广[2] 曾明平[1] 

机构地区：[1]武汉大学中南医院设备处,武汉430071 [2]华中科技大学生命科学与技术学院,武汉430074

出　　处：《北京生物医学工程》2016年第1期12-17,共6页Beijing Biomedical Engineering

摘　　要：目的三维医学图像配准能够为临床诊断提供更多更丰富的信息,是医学图像处理领域的研究热点。本文针对配准中传统方法很难兼顾到配准的准确度和速度,提出一种基于＂粗精＂混合配准的算法。方法首先,采用主成分分析方法对图像进行粗配准,减小浮动图像和参考图像之间的差异,得到精配准良好的初始参数;然后,采用改进的Powell算法在初始参数的基础上进行精配准;最后,以三维MRI图像为例设计了两组实验进行验证。结果该方法配准精度高,旋转参数误差低至0001,得到的图像灰度误差可限制在2以内。此外与全局优化方法相比,该方法保证配准精度的同时,在速度上可提高2~3倍。结论实验证明该方法可同时兼顾配准的准确度和速度。ObjectiveThree-dimensional medical image registration can provide more comprehensiveinformation for clinical diagnosis and is a hot field of medical image processing. Traditional registration methodsare difficult to take into account the accuracy and speed of registration at the same time. To solve the problem,this paper proposes an method based on ‘ coarse and fine＇ hybrid registration algorithm.MethodsFirst,principal component analysis was used for coarse image registration in order to reduce the difference betweenfloating images and the reference images and get good initial fine alignment parameters. Then,improved Powellalgorithm was used for refined registration based on initial parameters. Finally,two groups of experiments basedon 3D MRI images were designed to verify the method.ResultsThe registration accuracy of this method washigh,in which the rotation parameter error was about 0 001 while the image gray level error was limited to 2Additionally,this method could be 2 to 3 times faster than the global method while the registration accuracy notbe reduced.ConclusionsExperiments illustrated the registration accuracy and speed could be taken intoaccount at the same time in this method.

关 键 词：图像配准 主成分分析 Powell优化算法 

分 类 号：R318.04[医药卫生—生物医学工程] TP391.9[医药卫生—基础医学]