Hi SDM team,

We are conducting a VBM meta-analysis. We have full t-maps from all studies. The t-maps we have are for the cerebellum only and were generated using the SPM SUIT toolbox. While the results we are getting from SDM are reasonable, we just wanted to confirm how exactly the voxel-wise permutation-based FWE correction is being calculated. We noticed that the sdm_mask.nii.gz file includes whole-brain grey matter, rather then just the cerebellum.

1) Does this mean that the FWE correction is based on the number of voxels across the whole brain, rather than the the inputted t-maps which were just of the cerebellum?

2) if so, is there a way to ensure that the correction is only based on voxels were present in the t-maps? 

Thank you!

Dear Dr Chopra,

Apologies for the late reply. The default and recommended FWER correction in SDM-PSI is the "threshold-free cluster enhancement" (TFCE), and TFCE does depend on the mask. Therefore, we would strongly suggest restrict all SDM analyses to the cerebellum.

The use of a new mask is not straightforward, because SDM-PSI also needs a special binary file that contains information about the neighboring voxels of each voxel within the mask. That said, if you send us the mask of the cerebellum that you would like to use, we can create and send you the appropriate files to run SDM restricted to the cerebellum.

With best regards,

Joaquim