Büchel C, Turner R, Friston K
Wellcome Department of Cognitive Neurology, Institute of Neurology, United Kingdom.
Magn Reson Med. 1997 Nov;38(5):691-4. doi: 10.1002/mrm.1910380502.
Applications of fMRI in functional brain imaging are mainly confined to single subject designs, prohibiting the assessment of subject or group by condition interactions (i.e., differential activations) or areas of conjoint activation. In this paper a framework for fMRI group designs, using statistical parametric mapping, is introduced. It is generally believed that intersubject averaging, which requires spatial normalization and smoothing, will decrease the effective spatial resolution of fMRI or its sensitivity. A subcortical activation of the lateral geniculate nucleus (LGN) was therefore chosen to demonstrate the feasibility and power of intersubject averaging in the context of fMRI. Seven volunteers were studied, while looking at a blank screen or radially moving dots. LGN activation was demonstrated in all single subject analyses and in the group analysis.
功能磁共振成像（fMRI）在脑功能成像中的应用主要局限于单受试者设计，这使得无法评估受试者或组间因条件交互作用（即差异激活）或联合激活区域而产生的情况。本文介绍了一种使用统计参数映射的fMRI组设计框架。人们普遍认为，需要进行空间归一化和平滑处理的受试者间平均法会降低fMRI的有效空间分辨率或其灵敏度。因此，选择外侧膝状体（LGN）的皮质下激活来证明在fMRI背景下受试者间平均法的可行性和效能。对7名志愿者进行了研究，他们分别观看空白屏幕或径向移动的点。在所有单受试者分析和组分析中均显示出LGN激活。
