Coordinate based random effect size meta-analysis of neuroimaging studies

C.R. Tench, Radu Tanasescu, C.S. Constantinescu, D.P. Auer, W.J. Cottam
2017 NeuroImage  
Meta-analysis, neuroimaging, voxel based morphometry, functional MRI not peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was . http://dx.doi.org/10.1101/089565 doi: bioRxiv preprint first posted online Nov. 25, 2016; not peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was . http://dx.doi.org/10.1101/089565 doi:
more » ... ioRxiv preprint first posted online Nov. 25, 2016; coordinates from different experiments meaning that coordinates forming a small cluster are more significant than if they formed a larger cluster. Yet it is not clear, for example, that studies reporting thalamic coordinates producing a cluster over the thalamic volume should be less significant than the same studies reporting coordinates producing a smaller cluster in the smaller putamen structure. Finally, the uncorrected p-value threshold employed by both SDM and ES-SDM does not control the type 1 error rate in a principled way (Bennett 2009), and without estimated error rates there is no way to assess the significance of the results given the~210 5 voxel-wise statistical tests; this may propagate the very problems the MA was employed to mitigate (see (Tanasescu 2016) for example).
doi:10.1016/j.neuroimage.2017.04.002 pmid:28389386 fatcat:ozwkzjxbrjbarbb4bgnocvfwgy