HI

I just wanted to share my problem and my fix, in case someone else has it. It was really quite simple to fix but it may help someone.

One of my PESTICA results came out really bad when I checked the template coregistration, the coregistered image had the brain all twisted for some reason, so the final template was completely wrong for respiration and cardiac.

What I did to fix it was to play with the 3dAllineate line in the run_pestica.sh script. Specifically what I did was to change this line:

# PESTICA coreg to template
if [ ! -f $epi_mask.crg2mni.nii ] ; then
echo "Coregistration to EPI template"
3dAllineate -automask -prefix ./$epi_mask.crg2mni.nii -source $epi_mask+orig -base $PESTICA_DIR/meanepi_mni.nii -1Dmatrix_save $epi_mask.coreg.mni.1D



3dAllineate -cmass -cost ls -automask -prefix ./$epi_mask.crg2mni.nii -source $epi_mask+orig -base $PESTICA_DIR/meanepi_mni.nii -1Dmatrix_save $epi_mask.coreg.mni.1D

So, just added a search for the center of mass and the cost function to Least Squares and that worked for me. There can be any type of customizing here so if you keep having troubles, just keep tweaking.

That's one of the great features of PESTICA, the customization.

Hope this helps.

Kind regards

Eduardo

Thank you Eduardo, this is a very helpful tip. 3dAllineate can be finicky if there is a big head movement, we should add this in the next version to accept 3dAllineate strings as an option.