Masking out background voxels or obvious non-brain noise is useful in most fMRI analyses. A brain mask restricts analysis to voxels that plausibly contain brain tissue rather than air, skull, scanner background, or other irrelevant regions.
This can be achieved within BrainVoyager using a manual intensity threshold when computing GLMs. However, in some cases it might be necessary to mask out background noise in an FMR/STC dataset before running a GLM or when performing other analyses. But please note that a mask that excludes genuine brain voxels can remove signal from areas with naturally low intensity or susceptibility-related signal dropout, such as orbitofrontal cortex or anterior temporal regions. This matters especially if those regions are scientifically important.
In practice, a good mask is usually slightly conservative: include essentially all plausible brain voxels and exclude obvious air/background.
A mask file should not be generated too early in the fMRI preprocessing workflow. The reason not to rely on a mask generated directly from a single raw functional volume is that motion and EPI distortions can make the brain boundary shift across time.
After motion correction, the volumes are in a common functional space, so a mask is much more stable.
The BV Intensity Mask application, available for Windows and macOS, creates a mean-intensity background mask for BrainVoyager FMR/STC data (File Version 7), exports the masked FMR/STC data and saves the binary mask as a BrainVoyager MAP file.
Download:
BV Intensity Mask for Windows:
BV Intensity Mask for macOS arm