功能磁共振图像的独立成分分析与提取  

Independent Component Extraction and Analysis of Functional Magnetic Resonance Image

作　　者：武杰[1] 余莹 杨叶 WU Jie;YU Ying;YANG Ye(School of Medical Instrument and Food Engineering,Universityof Shanghai for Science and Technology Shanghai ,200093;EST Yishitong( Shanghai) Medical Equipment Co. LTD Shanghai,200093)

机构地区：[1]上海理工大学医疗器械与食品学院,上海200093 [2]伊士通(上海)医疗器械有限公司,上海200093

出　　处：《生物医学工程学进展》2018年第4期192-195,共4页Progress in Biomedical Engineering

基　　金：国家自然科学基金(61101174);上海理工大学微创基金项目(YS30810175)

摘　　要：目的用独立成分分析方法(Independent Component Analysis,ICA)处理视觉任务态f MRI数据,并从f MRI信号中分离出任务相关和非相关的独立成分。方法运用快速不动点算法(Fast ICA)处理-功能磁共振数据,得到各独立成分的时间多元回归系数和时间进程图,结合对实验任务的分析,选取识别出各类独立成分。结果分别识别出视觉任务相关独立成分、类似周期信号独立成分、头动信号独立成分。结论把独立成分分析方法应用到对f MRI数据的处理当中,不仅能够找到真正与任务相关的独立成分,而且能够识别出其他相关因素引起的独立成分,从而为科研实验或图像的分析诊断提供科学依据。Objective Using the independent component analysis(ICA)method to process fMRI data and separate the task-related and non-task-related independent components from fMRI signals.Methods The Fast ICA algorithm was used to process the functional magnetic resonance data of visual-task modality and obtain the time multivariate regression coefficients and the time-course diagrams of each independent component.Then,the independent components were selected and identified in combination with the analysis of the experimental task.Results Independent components of visual tasks,similar period signals and head movement signals were respectively identified.Conclusion The application of the independent component analysis to the processing of fMRI data can not only help to find out the true independent components related to the tasks,but also identify the independent components caused by other related factors.Thus,it can provide scientific basis for research experiments or analysis and diagnosis of images.

关 键 词：独立成分分析 视觉任务 核磁共振成像 

分 类 号：R737.25[医药卫生—肿瘤]