Imagine this situation: you’ve just finished running an fMRI experiment, and after hours of willing your subjects to stay still and painstakingly pre-processing their data, you’re finally ready to run some analyses. The first step in this analysis is typically to consider which brain regions to investigate further. Different researchers navigate this process in different ways, depending in part upon whether they’re interested in characterizing a particular region of the brain (region-of-interest analyses) or a larger swathe of cortex. Based on some of the ideas that researchers have used to select large swathes, we developed a new method to select voxels that respond consistently over the course of the experiment, and reliably differentiate among conditions. We call this method Reliability-Based Voxel Selection.
You can find more information about how to use this method and a comparison between using Reliability-Based Voxel Selection and Activity-Based Voxel Selection (another common method) in our article in Neuroimage or at the preprint linked above. I also have an FAQ page on this website.
