Hello,
I am trying to do skull stripe on cat’s brain. But there is no cats’ template, do you have some suggestions on by doing it on afni. I tried to modify the parameters on monkey’s template, but it didn’t work well.
Thanks.
Meizhen
I’ve never seen this in MRI images, but I was able to download a high-resolution cat MRI from the link below and script this to work with that dataset.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4060963/
There is also this recent paper that produced a group template, but I don’t see where that is publicly available.
https://www.frontiersin.org/articles/10.3389/fvets.2018.00172/full
#!/bin/tcsh
3dCatSkin
skullstripping for cat MRI
example with MPRage dataset downloaded from here:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4060963/
Heather L. Gray-Edwards, et al, High resolution MRI anatomy of the cat brain at 3 Tesla,
J. Neuroscience Methods. 2014 Apr 30; 227: 10–17.
mv 3D\ MPRage.nii.gz 3DMPrage.nii.gz
3dcopy 3DMPrage.nii.gz cat_MP
fix the orientation of this data (it’s wrong in the distributed header)
edited here to add missing orientation code
3drefit -orient RPI cat_MP+orig
also make this cardinal
3drefit -deoblique cat_MP+orig
make temporary copy
3dcopy cat_MP+orig. cat_MP_temp_resize
change dimensions to scale this to more human size
3drefit -xdel 1.0 -ydel 1.0 -zdel 1.0 -keepcen cat_MP_temp_resize+orig.
strip skull as if it were human
3dSkullStrip -prefix cat_ns -input cat_MP_temp_resize+orig. -shrink_fac_bot_lim 2.0
restore original dimensions
3drefit -xdel 0.417 -ydel 0.417 -zdel 0.4 -keepcen cat_ns+orig
Hi Daniel,
Thanks for your information. That’s very nice of you.
Actually I tried doing skull stripe on monkey template, with fc from 0.1-0.2, it could run with single 3dSkullStripe, but it failed in align_epi_anat.py -anat2epi, for it could not do skullstripe in align’s default option. My purpose is to do resting connectivity analysis, now got stuck in this step.
Thanks.
Meizhen
Did you try the steps I suggested for the cat? 3dSkullStrip has defaults for humans. The other animal options there do a similar resizing trick inside the program.
You forgot to include the orientation in your 3drefit command. I believe it should either be RPI or LPI.
Hi Daniel,
Thanks very much for your information. It’s helpful. I tried and it turned out to be much better!
Thanks,
Meizhen
Hi Phil,
Thanks very much, yes, I set it RPI and it worked!
Meizhen
Hi Daniel,
I used the resize tricky to do the skull-stripe, it looks nice, but when I check the alignment, I found that the brain size in epi and structure images were not the same. I resized the 0.5mm isotropic structure images to 1mm isotropic and 1mm isotropic original epi images to 2mm isotropic. What went wrong?
Thanks for your patience and kind help.
Meizhen
I’m not clear what is wrong here. Can you post the exact commands you used and what you think is wrong?