摘要
CT脑血管的三维重构依赖于图像的二维断层扫描,脑血管感兴趣区的提取是图像处理的关键。为了保证CT脑血管三维重建的图像质量,本文提出了基于曲线聚类的CT脑血管感兴趣区提取算法,首先移除头部CT图像中的头部骨骼,随后采用曲线聚类提取脑血管感兴趣区。该方法关注显影剂的时间放射性曲线的时间序列特征,能够有效的提取脑血管边缘,实验表明该方法具有良好的精度与鲁棒性。
The source data of 3-dimension CT cerebrovascular medical image reconstruction relies on 2D tomography slices.The key part of image processing is cerebrovascular region of interest(ROI)extraction.To insure the quality of 3D image reconstruction,CT cerebrovascular ROI extraction algorithm based on curvilinear clustering is proposed in this paper.Firstly,the head bone of CT image is removed.And then,cerebrovascular ROI is extracted using curvilinear clustering algorithm.This method focuses on the temporal radiation profile of the developer and can effectively extract cerebrovascular ROI.The experiments on real images show that the method proposed in this paper is more accurate and robust.
作者
杨文博
黄华
YANG Wen-bo;HUANG Hua
出处
《信息技术与信息化》
2018年第4期87-90,共4页
Information Technology and Informatization
关键词
脑血管
曲线聚类
显影剂
cerebrovascular
curvilinear clustering
developer
