Converting an MNI atlas to orig space

[jdp71]

When trying to use 3dfractionize (with the warp option) to take an MNI atlas from standard space to orig space, I get an error stating “** Warp dataset doesn’t contain a warp transformation!”

My command is:

3dfractionize -template <SS_warper_ouput+orig> -input <MNI_atlas+tlrc> -warp <SS_warper_output+tlrc> -prefix <MNI_atlas+orig> -clip 0.05 -vote

I get an error on that ‘-warp’ option. When I try to use the command without it, the resulting +orig space atlas is far too large to fit properly on the template anatomical image I used.

I’ve used 3dfractionize before with standardized anatomical images that were made with @auto_tlrc and they ran successfully since that command also creates a WarpDrive. However after using @SS_Warper for standardizing, I believe 3dfractionize is failing since there isn’t a WarpDrive log that gets created from @SS_Warper. Is there a current workaround for this?

[dglen]

3dfractionize only works on affinely transformed or 12-piece manually transformed datasets. @SS_warper uses both affine and nonlinear transformations. Follow the examples in this presentation, around slide 38:

https://afni.nimh.nih.gov/pub/dist/edu/latest/afni_handouts/afni11_roi.pdf

[jdp71]

Hi Daniel,

Thank you for your quick response!

I tried the 3dNwarpApply command mentioned in slide 38 and used the SS_warper relevant outputs:

3dNwarpApply -prefix origspace_mniatlas
-nwarp ‘anatQQ.s001.aff12.1D INV(anatQQ.s001_WARP.nii)’
-source mniatlas+tlrc
-master anatQQ.s001+orig

It resulted in the attached image. Am I using the correct “WARP” and “affine” SS_warper outputs?

[dglen]

For @SSwarper, it looks like you have invert the affine transformation. You can do this inline, try something like this instead:

3dNwarpApply -nwarp ‘INV(anatQQ.s001.aff12.1D) INV(anatQQ.s001_WARP.nii)’ -prefix mpm_orig -master anatSS.s001+orig
-source MNI_caez_mpm_22+tlrc. -ainterp NN
3drefit -cmap INT_CMAP -copytables MNI_caez_mpm_22+tlrc. mpm_orig+orig.