Anything with 0 thickness is assumed to not be a valid cortical vertex. As such, we don't smooth it so that it does not corrupt other vertices that are valid. When you stack the images together and use this masked smoothing, any invalid vertex will become 0 for all subjects and will be ignored in the GLM analysis.
On 4/6/2021 9:38 AM, Daniel Gutstein wrote:
External Email - Use Caution
Hello,
We are looking at the vertex-wise thickness measures for participants within the fsaverage space. There are a number of indices within the cortex that have thickness values of 0 in the thickness.fsaverage.mgh surface files. In FreeSurfer version 5.1, those same indices no longer have thickness 0 in the corresponding rh and lh fwhm smoothed thickness surface files due to the smoothing effect. However, in versions 6 and 7.1.1, the indices with 0 values remain 0 even in the fwhm smoothed files. Do you know why it was chosen that smoothing should have no impact on zeros in the later versions of FreeSurfer? Also, if we are interested in impacting the zeros from the later versions, is there a methodology which you would recommend?
Thanks,
Daniel
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer