基于因子分析法的超声肝灌注定量分析初步研究  被引量:4

A Preliminary Study on Quantification of Ultrasound Hepatic Perfusion Based on Factor Analysis

在线阅读下载全文

作  者:张冀[1] 丁明跃[1] 孟燔[1,2] 张旭明[1] 尉迟明[1] 

机构地区:[1]华中科技大学生命科学与技术学院图像处理和智能控制国家教育部重点实验室,武汉430074 [2]武汉市第一医院设备科,武汉430070

出  处:《中国生物医学工程学报》2011年第3期346-351,共6页Chinese Journal of Biomedical Engineering

基  金:国家重点基础研究发展(973)计划(2011CB933103);中英国际科技合作项目(2009DFA12290)

摘  要:局灶性肝结节(FLLs)通常可从超声造影图像序列中提取的时间强度曲线进行量化评价。为了克服人工选择的主观性并自动提取曲线,提出一种用于因子分析的基于顶点搜寻的替代-近似(RA)算法。该算法将高维图像序列映射到一维空间,先找到与生理结构对应的两个顶点。然后,根据这两个已知点的信息,在二维空间中可以寻找到第3个顶点。这3个顶点对应着3条不同的目标曲线。实验采用6个在患者自由呼吸运动下采集的肝细胞癌病例对RA算法进行验证。实验结果表明,此方法能准确提取到具有生理意义的因子曲线和相应的因子图,提取的因子曲线与感兴趣区域(ROI)测定曲线的平均相关系数值为0.91±0.03。初步证明了RA算法在FLLs灌注的量化分析具有可行性。Focal liver lesions(FLLs) are usually quantitatively assessed by blood time-intensity curves(TICs) extracted from dynamic contrast-enhanced ultrasound(CEUS) image sequences.To overcome the subjectivity of manual operation and extract TICs automatically,a novel replace-approximation(RA) method based on apex-seeking for factor analysis was proposed.By mapping the high dimensional image series into one-dimensional space,the first two apexes corresponding to the physiological structures were determined.According to the information of two apexes,the third one was determined in two-dimensional space.The obtained apexes are corresponding to three different targeted TICs.This method was tested on six hepatocellular carcinomas cases which were collected during patients breathed freely.Experimental results showed that the RA method could extract physiological factor curves(targeted curves) and the corresponding factor images accurately.The mean correlation coefficients between the extracted factor curves and the region of interest(ROI) measurements were 0.91±0.03.Our work preliminarily demonstrated the feasibility of the RA method in the quantitative analysis of perfusion for FLLs.

关 键 词:替代-近似 因子分析 超声造影 时间-强度曲线 肝灌注 

分 类 号:R318.08[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