基于拉伸空间变换算法在DSA配准中的应用

提出一种基于拉伸的空间变换算法, 首次将其应用于数字减影血管造影术(DSA) 图像配准中以消除运动伪影. 实验结果表明, 与传统线性空间变换算法相比, 采用基于拉伸的空间变换算法配准后的减影图像质量显著改善. 为提高计算速度, 仅对选定的控制点进行块匹配操作, 图像其余像素点基于相邻像素运动位移的连续性, 均由找到的运动位移点经空间变换和灰度插值计算得到. 基于拉伸的空间变换算法可避免采用传统线性空间变换算法所产生的病态解, 属于非线性有理变换. 提出的基于拉伸的空间变换算法简单实用, 可准确有效地确定需估算的运动位移点位置, 且配准后的减影图像质量优于采用线性空间变换算法配准的减影图像质量.

Application of a spatial transformation method based on stretching to digital subtraction angiography registration

A spatial transformation based on stretching gives superior initial registration of digital subtraction angiography (DSA) images. It reduces artifacts caused by motion displacement. The spatial transformation method is more effective than traditional linear transformation methods. Results show that the quality of DSA images after registration is improved significantly. Block matching was used to calculate the motion displacement of each pixel in a given DSA image. Computational overhead was reduced by calculating only reference points as control points for block matching. The remaining pixels in the DSA image were mapped by spatial transformation and gray-level interpolation of the calculated reference points. This spatial transformation method based on stretching solves unexpected ill-behaved solutions common to traditional linear space transformations. Mathematical analysis of this stretching based space transformation likens it to a nonlinear but rational transformation. Investigation confirms this spatial transformation method to be simpler and more efficient than linear transformation methods. Motion displacement is better compensated and the quality of DSA images is significantly improved with this method.