基于Voronoi图的血管中心线提取方法

为解决血管分割及中心线提取技术在提取血管分叉及细小血管时往往存在较大误差的问题，提出一种新的基于Voronoi图的中心线提取方法. 该方法利用血管几何特性确定其中心线，有效抑制了图像灰度分布不均匀以及噪声的干扰. 通过优化抽样方法有效利用血管的曲率信息，根据分叉结构与血管边界曲率差异提出不同的采样方式，在降低采样点数目的同时确保中心线提取的准确性与连续性. 实验结果证明该方法具有良好的鲁棒性，获得的中心线提取误差小于0.42像素，能够快速并准确地在造影图像中提取出血管中心线，同时有效解决了分割血管分叉点时采样不连续的问题. 

A Fast Vascular Centerline Extraction Method Based on Voronoi Diagram

Through the technology of vascular segmentation, the vascular structure in the angiograms can be extracted from the complex background, which has important clinical value for the diagnosis and treatment of vascular diseases. For the uneven gray scale distribution of the X-ray accumulation in the angiographic images, most traditional methods present large extraction errors for vessels with multiple intersections or for small segments. In this paper, a novel method is developed for the extraction of the vasculature from coronary angiographic images based on Voronoi diagram. The developed method employs geometric properties to determine the centerline. Hence, it can minimize the effect of uneven gray scale distribution and the bifurcation interferences. Accurate vascular extraction can be guaranteed by optimized sampling method. Experimental results show that the developed method is very effective and robust, which can obtain accurate centerlines from the angiographic images quickly.