Yes, it will. You will need to create a "frame mask" volume in which has a frame for each subject. If a voxel for a subject should be included in the analysis, then it should be a 1, otherwise a 0. If you want to use your input data to provide the mask, then just mri_binarize --i y.mgh --min .0001 --o framemask.mgh where .0001 is just a small number meant to capture any non-zero value. You may need to make it closer to 0 if your input data has small values. The way it works is that a different GLM is run at each voxel. While this sounds like a good idea, there can be problems in that you can have an effect in one voxel that goes away in an adjacent voxel because the model changes, so you can get funny looking maps that appear to be cut off. If it is just a matter of some voxels that were not scanned, then maybe it is not a big deal.
On 01/18/2016 12:31 PM, Victor Montal Blancafort wrote:
Dear FS Experts,
I have seen few posts about this mri_glmfit flag, and I would like to know how it works. In an older post, (https://www.mail-archive.com/freesurfer%40nmr.mgh.harvard.edu/msg32412.html), Doug suggested to use this flag when only some subjects should be exclude for each vertex (i, e. differnet number of subjects to exclude per vertex). Could anyone give me an insight on how this works?
My intention is to perform a GLM analysis of some DTI metric volumes projected into the surface. However, some of the DWI volumes are cutted, so I would like exclude those vertexs where there is no signal (i,e value 0). Will --frame-mask work for this? Any other suggestion?
Any help will be very appreciated! Thank you in advance,
Victor Montal
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer