节点文献

冠脉造影图像血管增强及减影的优化方法

Vessel Enhancement and Subtraction Optimization Method for Coronary Angiography Image

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 宗娇娇张天序

【Author】 ZONG Jiaojiao;ZHANG Tianxu;School of Electrical and Information Engineering,Wuhan Institute of Technology;Institute for Pattern Recognition and Artificial Intelligence,Huazhong University of Science and Technology;

【通讯作者】 张天序;

【机构】 武汉工程大学电气信息学院华中科技大学图像处理与人工智能研究所

【摘要】 针对于传统X射线冠状动脉造影序列图像的血管的减影包含严重伪影以及减影结果不完整等问题,提出了冠脉造影图像血管增强以及减影优化的方法。基于冠脉造影序列图像的时空特性,将相同时相的图像匹配进行减影,得到初步减影结果。针对初步减影部分缺失,对有导管遮挡没有被识别的单根血管进行二次增强处理。增强方法为:运用区域生长确定被导管遮挡的血管部分,以此部分为基础按规定模板扩展,形成覆盖导管区域,区分血管和导管,改变两部分的灰度值,增大对比度。最后采用灰度线性变换、中值滤波等步骤,得到优化的冠脉血管减影。实验表明,本减影及优化方法可以从冠脉造影序列图中提取较为完整清晰的冠脉血管并且具有很好的可控性和鲁棒性。

【Abstract】 Aiming at the problems that the subtraction results of traditional X-ray coronary angiography sequence images contain serious artifacts and incompleteness, methods of vessel enhancement and subtraction optimization were proposed. Preliminary subtraction results were obtained by analyzing the spatiotemporal characteristics of coronary angiography sequence images and matching the images of the same time phase. For the partial missing of the preliminary subtraction, secondary enhancement was performed on single unrecognized vessels with occlusion by the catheter. The enhancement method is using the region growth to determine the vascular blocked by the catheter. This area was selected as the basis and expanded according to the prescribed template. Therefore, the covered catheter area can be formed. Then, the contrast between the vessel and the catheter was increased by changing the gray value of the two areas. Finally, steps such as grayscale linear transformation and median filtering were used to obtain the optimized coronary angiography. The result suggests that the vessel enhancement and subtraction optimization method can extract more complete and clear coronary vessels from coronary angiography sequence diagram, and has good controllability and robustness.

  • 【文献出处】 武汉工程大学学报 ,Journal of Wuhan Institute of Technology , 编辑部邮箱 ,2021年01期
  • 【分类号】TP391.41;R816.2
  • 【下载频次】123
节点文献中: 

本文链接的文献网络图示:

本文的引文网络