Moderated t-tests for group-level fMRI analysis

Abstract

In recent years, there has been significant criticism of functional magnetic resonance imaging (fMRI) studies with small sample sizes. The argument is that such studies have low statistical power, as well as reduced likelihood for statistically significant results to be true effects. The prevalence of these studies has led to a situation where a large number of published results are not replicable and likely false. Despite this growing body of evidence, small sample fMRI studies continue to be regularly performed; likely due to the high cost of scanning. In this report we investigate the use of a moderated t -statistic for performing group-level fMRI analysis to help alleviate problems related to small sample sizes. The proposed approach, implemented in the popular R-package LIMMA (linear models for microarray data), has found wide usage in the genomics literature for dealing with similar issues. Utilizing task-based fMRI data from the Human Connectome Project (HCP), we compare the performance of the moderated t -statistic with the standard t -statistic, as well as the pseudo t -statistic commonly used in non-parametric fMRI analysis. We find that the moderated t -test significantly outperforms both alternative approaches for studies with sample sizes less than 40 subjects. Further, we find that the results were consistent both when using voxel-based and cluster-based thresholding. We also introduce an R-package, LIMMI (linear models for medical images), that provides a quick and convenient way to apply the method to fMRI data.

SAN: mitigating spatial covariance heterogeneity in cortical thickness data collected from multiple scanners or sites

“…Considering such smoothness in the estimation step would reduce the variance of the parameter estimates, which is expected to improve the performance of the Stage 2 of SAN especially when the number of subjects is small. To achieve this, SAN pools information within a series of prespecified local neighbors [32]. Specifically, we define local neighbors 𝒩 r ( v ) for each vertex as the set of vertices whose geodesic distances from vertex v are less than or equal to r .…”

“…When r = ∞, we estimate θ iv and by using all vertices in the data, which would result in ‘too smooth’ estimates shrunk towards the brain-level means and variances. An appropriate level of r needs to be chosen by considering the bias-variance tradeoff (Wang et al, 2021). In this paper, we choose ComBat model with r = 5mm as a default to borrow information across neighbors without inducing too much bias.…”

“…To achieve this, SAN pools information within a series of prespecified local neighbors (Wang et al, 2021). Specifically, we define a local neighbor 𝒩 r ( v ) for each vertex as the set of vertices whose geodesic distances from vertex v are less than or equal to r , and apply ComBat (Johnson et al, 2007; Fortin et al, 2018).…”

CLEAN: Leveraging spatial autocorrelation in neuroimaging data in clusterwise inference

“…Examples of a nonparametric approach are the methods that locally smooth the variance of the test statistics. FSL provides a variance smoothing option [10] to generate 'pseudo' t statistics, and Wang et al [22] recently proposed an approach based on empirical Bayes, termed 'locally moderated' t statistic, by pooling variance information across a predefined set of neighbors. An example of a parametric approach includes modelling the spatial covariance using a Gaussian process (GP).…”

Integrative Systems Biology Analysis Elucidates Mastitis Disease Underlying Functional Modules in Dairy Cattle

“…The meta-analysis included the following steps: after quantile normalization, labeling samples as "Infected" or "Healthy"; the "Gene Symbol" was matched to multiple probe IDs using the interquartile range for probe selection (Wang et al, 2012). Merging genes is an approach used to determine which genes should be studied further (Wang et al, 2021). Because the number of genes in the research varied, the multiple gene expression datasets may not have been adequately matched by genes.…”