【摘要】 目的 :探讨多重线性回归分析和一般线性 (GLT)检验在fMRI数据处理中的应用。材料和方法 :采用多重线性回归和GLT技术分析了一组复杂运动与简单运动的功能磁共振成像数据 ,并与传统互相关分析结果进行比较。结果 :采用互相关分析 ,选用不同的阈值得出了完全不同的结论 ;而采用多重线性回归和GLT则不需要主观设定域值 ,并可对复杂与简单两种运动任务的激活强度进行定量比较 ,表明各脑区不同程度地参与两种运动任务 ,可被看作有一定等级的平行加工系统。结论 :与传统互相关分析相比 ,本研究的方法更为客观 ,并适合于定量分析两种不同任务所激活的脑活动的差异。更多还原

【Abstract】 Purpose: To evaluate validity of multiple linear regression and general linear test (GLT) in analysis of fMRI data. Materials and Methods: fMRI data acquired during complex and simple movements were andalyzed using multiple linear regression and GLT; the outcomes were compared with those based on conventional cross-correlatin analysis. Results: Using cross-correlation analysis, substantially different conclusions were drawn with two different thresholds. On the contrary, the new method introduced in this report did not need to subjectively set threshold and could directly compare brain activation during performance of two tasks. With this new method, we provided strong evidence that most individual motor-related brain regions were involved in simple and complex movements, suggesting that all these regions could be regarded as a series of hierarchical and parallel systems. Conclusion: Compared with the conventional cross-correlation analysis, the method mentioned in this paper appears to be more objective and reliable, and can be used to quantitatively compare brain activation between two different tasks.更多还原