文章基本信息

标题：Detecting Brain Activations in Functional Magnetic Resonance Imaging (fMRI) Experiments with a Maximum Cross-Correlation Statistic
本地全文：下载
作者：Kinfemichael Gedif ; William R. Schucany ; Wayne A. Woodward 等
期刊名称：Journal of Data Science
印刷版ISSN：1680-743X
电子版ISSN：1683-8602
出版年度：2012
卷号：10
期号：3
页码：403-418
出版社：Tingmao Publish Company
摘要：Various statistical models have been proposed to analyze fMRIdata. The usual goal is to make inferences about the eects that are relatedto an external stimulus. The primary focus of this paper is on those statisticalmethods that enable one to detect `signicantly activated' regions of thebrain due to event-related stimuli. Most of these methods share a commonproperty, requiring estimation of the hemodynamic response function (HRF)as part of the deterministic component of the statistical model.We propose and investigate a new approach that does not require HRFts to detect `activated' voxels. We argue that the method not only avoidstting a specic HRF, but still takes into account that the unknown responseis delayed and smeared in time. This method also adapts to dierential responsesof the BOLD response across dierent brain regions and experimentalsessions. The maximum cross-correlation between the kernel-smoothedstimulus sequence and shifted (lagged) values of the observed response is theproposed test statistic.Using our recommended approach we show through realistic simulationsand with real data that we obtain better sensitivity than simple correlationmethods using default values of SPM2. The simulation experiment incorporatesdierent HRFs empirically determined from real data. The noisemodels are also dierent AR(3) ts and fractional Gaussians estimated fromreal data. We conclude that our proposed method is more powerful thansimple correlation procedures, because of its robustness to variation in theHRF.
关键词：HRF; kernel; nonparametric.