摘要
目的研究独立分量分析(ICA)方法在区别肝纤维化疾病中的应用。方法采用FastICA算法,在基于正常肝组织与纤维化肝组织可以看作是独立信源的前提下,对同一病人不同纤维化时期和正常人的超声图像选取局部区域后分别进行独立分量分离。结果FastICA算法可以较快的分离出各组的独立分量,且异常组的独立分量数明显多于正常组。结论对肝纤维化超声图像的独立成分进行分析,并与相应部位的正常肝组织超声图像的独立成分进行比较,是一种值得尝试的新的超声图像分析方法。
[Objective] To study the appliance of the Independent Component Analysis (ICA) method in hepatic fibrosis. [Method] Fast ICA algorithm is used to distinguish the hepatic fibrosis from the normal 2D ultrasonic images, which was based on the statistical independence. [Result] Each group's independent components could be extracted quickly by FastlCA, and the abnormal's components are more than the normal's. [Conclusion] It is a new feasible analyzing tool to apply ICA in the ultrasonographic study for hepatic fibrosis.
出处
《中国医学工程》
2005年第6期567-570,574,共5页
China Medical Engineering
基金
浙江省自然科学基金项目(302684)
关键词
独立分量分析(ICA)
肝纤维化
超声图像
independent component analysis (ICA)
hepatic fibrosis
ultrasonic images