ETAC combines the results from using multiple sets of parameters to give experimental results that are not strongly dependent on arbitrary parameter choices. The method presented here, called “Equitable Thresholding and Clustering” (ETAC), was developed to reduce the influence of the selection of these arbitrary analysis parameters. Second, among the surviving voxels, accept only those that form neighborhoods with other surviving voxels in a cluster of some threshold size or larger (referred to as “ C” or clustering threshold).īoth the “ S” and “ C” steps involve the selection of arbitrary parameters for voxel-wise significance and cluster threshold. First, examine each voxel and reject any whose test statistic likelihood ( p-value) is larger than some user-selected p-threshold (referred to herein as “ S” or statistical threshold). Leveraging contiguity is commonly carried out using dual thresholding of statistical parametric maps, performed in two successive steps. One strategy for multiple comparison correction is to leverage the observation that plausible results are likely to span multiple contiguous voxel locations, leading to the rejection of statistical findings that are spatially too small or too insignificant. One of the challenges in analyzing a functional magnetic resonance imaging (FMRI) experiment is an appropriate correction for the large number of multiple comparisons, resulting from separate statistical tests at each spatial location (voxel). In addition, a task FMRI data collection was used to compare ETACs true positive detection power versus a standard cluster detection method, demonstrating that ETAC is able to detect true results and control false positives while reducing reliance on arbitrary analysis parameters. The approach was validated with pseudotask timings in resting-state brain data.
ETAC utilizes resampling methods to estimate the FPR and thus does not rely on parametric assumptions about the spatial correlation of FMRI noise.
The approach adjusts the cluster-thresholding parameter of each subtest in an equitable way, so that the individual false-positive rates (FPRs) are balanced across subtests to achieve a desired final FPR (e.g., 5%). The union of these subtest results decides which voxels are accepted. The proposed “Equitable Thresholding and Clustering” (ETAC) approach seeks to reduce the dependence of clustering results on arbitrary parameter values by using multiple subtests, each equivalent to a standard FMRI clustering analysis, to make decisions about which groups of voxels are potentially significant. This article describes a hybrid method to threshold functional magnetic resonance imaging (FMRI) group statistical maps derived from voxel-wise second-level statistical analyses.
0 kommentar(er)
