Hi Antonella, the right way to do it is to figure out what is going wrong with mri_glmfit-sim. Did it obviously fail in some way? Can you send the log file? Should be permcsd.mri_glmfit-sim.log. If you need to run it for debugging purposes, you can run it with 10 iterations (change permcsd to permcsdtmp so you don't overwrite what is there). This should produce all the files, so if something is not there, it should be easy to track down why. doug

Antonella Kis wrote:

Dear Doug,

When I run the: mri_glmfit-sim \ --glmdir GroupAnalysis_pos.glmdir \ --sim perm 10000 3 permcsd \ --sim-sign pos \ --cwpvalthresh .05

I am not getting my clusters file permcsd.sig.cluster.mgh with the clusters summary so as you recommended I am running the

mri_volcluster --in Group_Analysis.glmdir/contrast/sig.mgh --mask Group_Analysis.glmdir/mask.mgh --no-fixmni --cwsig Group_Analysis.glmdir/contrast/permcsd.sig.cluster.mgh --sum Group_Analysis.glmdir/contrast/permcsd.sig.cluster.summary --ocn Group_Analysis.glmdir/contrast/permcsd.sig.ocn.mgh --cwpvalthresh .05 --seg cvs_avg35 norm.mgz --csdpdf Group_Analysis.glmdir/contrast/permcsd.pdf.dat --csd Group_Analysis.glmdir/csd/permcsd.j001-contrast.csd --vwsig Group_Analysis.glmdir/contrast/permcsd.sig.voxel.mgh

My question is: it is OK if I am not mentioning any registration or what should I input under --reg while I running the mri_volcluster if my for DTI preprocessing I used the dt_recon which computes for each subject an fa.nii and register.dat using the Talairach registration and later I registered all my subjects on the CVS space?

Thank you. AK