一种改进的3D血管增强算法

发布时间：2018-07-07 16:17

  本文选题：血管增强 + Hessian矩阵　； 参考：《计算机工程与应用》2016年15期

【摘要】：Sato的滤波器可以有效地滤除3D数据中的面状结构和球状结构,但是对于背景信息滤除效果不佳。基于Sato的血管增强原理,提出了一种改进算法。新算法将背景像素的Hessian矩阵特征值也考虑到滤波器的设计中,增强了滤波器滤除背景像素的功能。基于临床颈动脉的CTA数据和脑部血管MRA数据,做了实验分析。实验结果表明,在保留原算法优点的情况下,背景内容被大幅度移除,图像对比度得到进一步提高,血管结构更加清晰可辨。
[Abstract]:Sato filter can effectively filter out the surface structure and spherical structure in 3D data, but the filtering effect is not good for background information. Based on the principle of Sato, an improved algorithm is proposed. The new algorithm takes the Hessian matrix eigenvalue of background pixels into account in the design of the filter and enhances the filter's function of filtering background pixels. Based on CTA data of clinical carotid artery and MRA data of cerebral vessels, experimental analysis was made. The experimental results show that, with the advantages of the original algorithm, the background content is removed greatly, the image contrast is further improved, and the vascular structure is more clearly discernible.
【作者单位】： 湖南大学电气与信息工程学院;中国科学院深圳先进技术研究院;
【基金】：国家自然科学基金重点项目(No.61433016) 湖南省科技计划项目(No.2014FJ3075)
【分类号】：R816.2;TP391.41
，

本文编号：2105518