Dear all,

 

I have run a VBM analysis that generated results outside the GM and was wondering how this is possible. I have played with absolute threshold values up to 0.2 and still getting this. Of note, I did not use the SPM segmentation algorithm (it wasn’t particularly successful in our particular dataset), but another tool  and then took the segmented GM images, smoothed them and entered them in the analysis. Do I have to use a higher absolute threshold? Other ideas? Thanks.