【摘要】 直径的差别和造影剂的厚度不同造成血管在MRA成像中灰度不均匀,加上图像噪声的影响,使得血管的分割非常困难。而本文从血管在MRA图像中的特性出发,提出了逆浓度扩散模型,实现了血管的增强;另外本文还提出了基于直方图的Mumford-Shah模型计算方法,实现了图像的快速分割。实验表明,本文提出的方法能有效应用于二维及三维MRA图像中的血管分割。更多还原

【Abstract】 The grayscales of blood vessels in MRA images suffer a great range of variety because of different diameters of the ves- sels and different thickness of the contrast medium,while the negative effects of noise even make the segmentation of blood vessels diffi- cult.In the paper,based on vessel’s characteristics in MRA,we designed a new“anti-concentration diffusion method”to intensify them,and also realized a fast segmentation arithmetic based on Mumford-Shah model relating to grayscale histogram.Experimental re- sults show that our method works effectively well on the segmentation of vessels in two-dimensioned or three-dimensioned MRA imaging.更多还原