节点文献
冠状动脉造影图像序列中血管运动特征的提取
Dynamic Feature Extraction of Coronary Artery Motion from Angiogram Sequences
【摘要】 采用基于特征的光流法,由单面冠状动脉造影图像序列估计二维血管的运动.从连续的两帧图像中提取出动脉主要分支的中心线,沿中心线采用光流法计算出血管以及相关心外膜的运动参数.通过分析整个冠状动脉树的运动,得到对心脏运动的定量估计.该算法是全自动的,不需要操作者的参与.实验结果证明了该算法的可行性.
【Abstract】 The estimation of the 2-dimensional motion of the coronary vessel from a sequence of single plane angiographic images, using the feature-based optical flow method, is presented. The main artery centerlines are first extracted from two successive images and then the motion parameters of artery and related epicardium are computed along these centerlines. The quantitative assessment of human heart motion is obtained from the analysis of the entire coronary artery tree. The algorithm is automatic without operator′s interactive procedures.
【关键词】 冠状动脉造影;
图像序列分析;
二维运动特征提取;
光流法;
【Key words】 coronary artery angiography; image sequence analysis; 2-D motion feature extraction; optical flow;
【Key words】 coronary artery angiography; image sequence analysis; 2-D motion feature extraction; optical flow;
- 【文献出处】 天津大学学报 ,Journal of Tianjin University , 编辑部邮箱 ,2003年06期
- 【分类号】R310
- 【被引频次】5
- 【下载频次】97