【摘要】 目的验证可否利用独立成分分析(ICA)技术和静息fMRI数据对脑功能区进行定位。方法利用ICA方法,通过研究静息状态的脑功能联结来获取功能区的定位。静息数据的采集采用短TR,在低通滤波(截止频率0.08Hz)后可以去除生理噪声的主要影响。在数据分析中,对ICA结果进行了可复制性分析,只保留可复制性较高的成分,之后将ICA结果与传统的"种子像素"方法获得的结果进行定量的一致性分析。结果ICA能够在不设定"种子像素"的情况下从静息fMRI数据中分解出运动系统和初级视觉系统的功能联结图,并在所有被试上都与"种子像素"方法有较高一致性。ICA在同一数据中可以同时分解出上述两个系统的功能联结图。结论ICA克服了"种子像素"方法的主观性,稳定、准确地从静息fMRI数据中分解出了脑功能联结图。本研究支持初级功能系统内的联系要明显强于系统间的联系的假设,显示ICA方法具有良好的临床应用潜力。更多还原

【Abstract】 Objective To examine the feasibility of functional localization in the human brain with resting-state (task-free) fMRI data using independent component analysis (ICA). Methods ICA was used to study the functional connectivity in resting-state in order to locate the functional regions. The resting-state fMRI data were collected using short TR,and the major impact of various physiological noises was eliminated after the data were low-pass filtered (cutoff frequency=0.08 Hz). ICA components were verified through reproducibility analysis,and only highly reproducible components were retained in the analysis of data. The results of ICA and the "seed voxel" method were then quantitatively compared for consistency. Results ICA was able to separate the functional connectivity maps for motor and primary visual systems without selecting the "seed voxel". The results of ICA had high consistency with those of traditional "seed voxel" method. Furthermore,ICA simultaneously obtained the functional connectivity maps for the two systems within one dataset. Conclusion ICA overcame the subjectiveness in the "seed voxel" method,and was capable to obtain functional connectivity from resting-state fMRI data. This study supports the hypothesis that there is stronger functional connectivity within primary systems than between them. Moreover,the current study has demonstrated potential capability of ICA in clinical applications.更多还原