Demonstration and validation of Kernel Density Estimation for spatial meta-analyses in cognitive neuroscience using simulated data.
Demonstration and validation of Kernel Density Estimation for spatial meta-analyses in cognitive neuroscience using simulated data.
复制标题
DOI：
10.1016/j.dib.2017.06.003
复制
发表时间：
2017-08
期刊：
影响因子：
1.2
通讯作者：
Kotz SA
中科院分区：
文献类型：
--
作者：
Belyk M;Brown S;Kotz SA
本文提供的数据与题为《IFG眼眶内语义和情感表达的融合》(Belyk等人，2017年)的研究文章有关。这篇研究文章报道了对大脑成像实验的空间荟萃分析，该实验比较了人类对语义和情感交流信号的感知。本文简要介绍了核密度估计(KDE)作为神经成像数据的一种新的统计方法的应用。首先，我们对KDE和之前发表的应用激活似然估计的元分析进行了并列比较，激活似然估计是认知神经科学中元分析的主要方法。其次，我们对已知空间属性的模拟数据进行分析，以测试KDE对不同空间分离程度的敏感度。KDE成功地在模拟数据中检测到真实的空间差异，并且在没有真实差异的情况下显示了很少的假阳性。用于模拟和分析这些数据的R代码被公开提供，以促进对神经成像数据的KDE的进一步评估，并将其传播给认知神经科学家。
The data presented in this article are related to the research article entitled “Convergence of semantics and emotional expression within the IFG pars orbitalis” (Belyk et al., 2017). The research article reports a spatial meta-analysis of brain imaging experiments on the perception of semantic compared to emotional communicative signals in humans. This Data in Brief article demonstrates and validates the use of Kernel Density Estimation (KDE) as a novel statistical approach to neuroimaging data. First, we performed a side-by-side comparison of KDE with a previously published meta-analysis that applied activation likelihood estimation, which is the predominant approach to meta-analyses in cognitive neuroscience. Second, we analyzed data simulated with known spatial properties to test the sensitivity of KDE to varying degrees of spatial separation. KDE successfully detected true spatial differences in simulated data and displayed few false positives when no true differences were present. R code to simulate and analyze these data is made publicly available to facilitate the further evaluation of KDE for neuroimaging data and its dissemination to cognitive neuroscientists.