作　　者： ; ; ;

机构地区： 南方医科大学生物医学工程学院医学工程系

出　　处： 《中国医学物理学杂志》 2005年第5期644-647,共4页

摘　　要： 本文提出了对Hany Farid医学图像弹性配准算法的一个改进算法.利用正交的Hermite多项式在最佳平方逼近的误差标准下设计了一种离散微分算子,通过对不同特点的医学图像作配准将传统微分算子、原算法微分算子和微分算子进行了比较,验证了使用Hermite微分算子在保持了原算法配准效果的同时极大地提高了配准速度. In this paper, a novel method is proposed to improve Hany Farid＇s elastic registration scheme. Based on Hermite orthogonal polynomial, a discrete derivative filter is designed according to the principle of least square. Using the registration results of a set of medical images with different features, the effect of Hermite derivative filter is compared with ones of difference filter and Hany Farid＇s optimal derivative filter, which proving that Hermite derivative filter can significantly reduce the cost-time and obtain the same registration precision simultaneously.

关 键 词： 医学图像 弹性配准 离散微分算子

领　　域： [自动化与计算机技术] [自动化与计算机技术]