一种可变系数的对数脑血管减影算法研究  被引量:1

Logarithmic subtraction algorithm of variable coefficient on cerebral blood vessels

在线阅读下载全文

作  者:欧阳丽蓉[1] 邓振生[1] 严昂[2] 

机构地区:[1]中南大学地球科学与信息物理学院,长沙410083 [2]中南大学湘雅医院,长沙410008

出  处:《北京生物医学工程》2012年第1期36-39,共4页Beijing Biomedical Engineering

摘  要:目的针对数字减影血管造影(digital subtraction angiography,DSA)血管图像中非血管结构难以完全去除,严重干扰对血管分析与诊断的问题,本文寻求一种适用于脑血管DSA图像的减影方法。方法根据X射线在人体组织中的吸收特性,提出一种可变系数的对数减影算法。该算法首先对造影前后的图像分别进行对数变换再相减,然后设计一个自适应的扩大系数,将对数减影后的图像进行灰度动态范围的扩大。结果对源自临床的一个实际造影图像(蒙片和盈片)分别采用直接减影、一般的对数减影和本文的对数减影算法进行处理,三种方法中本文算法获得的图像非血管结构较少,对比度强。结论实验结果证明,可变系数对数减影算法能有效消除或降低背景噪声,获得血管灰度一致、对比度较强的血管图像。Objective Since the non-vascular structures can hardly be fully expelled from the blood vessels in the image of digital subtraction angiography ( DSA ) , which seriously interferes the analysis and diagnosis of blood vessels, this study looks for a subtraction method suitable for cerebral blood vessels from DSA. Methods According to the absorbing features of X-ray in human bodies, a logarithmic subtraction algorithm of variable coefficient was proposed. Logarithmic transformation was carried on the images of before- and after-angiography respectively,which were then subtracted. Then,a self-adaptive expanded coefficient was used to expand the range of the gray level of the subtraction image. Results Direct subtraction, ordinary logarithmic subtraction and the proposed method were performed on the clinical angiography images respectively in this paper. It showed that the proposed method could obtain images with less non-vascular structures and better contrast than the other two methods. Conclusions The logarithmic subtraction algorithm of variable coefficient might effectively eliminate or decrease background noises, which was helpful to obtain blood vessel images with uniform gray levels and good contrast.

关 键 词:数字减影血管造影 可变系数 对数变换 

分 类 号:R318.04[医药卫生—生物医学工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象