Brainnetome Atlas


I am trying to load the brainnetome atlas as an overlay (,157540,157568).

I have the atlas in AFNI_data5, so I changed to that directory and launched AFNI from there.

The output of “ls” is as follows:

@demo.Artifact           MyData                  s00.ap                     zz.errts.s620.fanaticor+orig.BRIK.gz
BN_Atlas_246_1mm.nii     index.html              s620_rest_r1+orig.BRIK.gz  zz.errts.s620.fanaticor+orig.HEAD
BN_Atlas_246_1mm.nii.gz  out.ss_review.s620.txt  s620_rest_r1+orig.HEAD

The BN_Atlas_246_1mm.nii.gz file is highlighted in red, along with the zz.errts.620.fanaticor+orig.BRIK.gz file.

When I click the overlay button in the GUI, only “s620_resr_r1” and “zz.errts.s620.fanaticor” appear – there is no sign of the BN atlas. The same applies for the underlay button. After clicking the overlay button, I see this in the terminal:

** ERROR (nifti_image_read): failed to find header file for '/home/marcus/AFNI_data5/BN_Atlas_246_1mm.nii'
** ERROR (nifti_image_read): failed to find header file for '/home/marcus/AFNI_data5/BN_Atlas_246_1mm.nii.gz'

I thought I could perhaps solve this using the “3dcopy” function, changing the file to a more AFNI-compatible file type. Here is the result of that attempt:

3dcopy BN_Atlas_246_1mm.nii.gz BN_Atlas
++ 3dcopy: AFNI version=AFNI_22.0.11 (Feb 18 2022) [64-bit]
** ERROR (nifti_image_read): failed to find header file for '/home/marcus/AFNI_data5/BN_Atlas_246_1mm.nii.gz'
** FATAL ERROR: no datasets found!
** Program compile date = Feb 18 2022

Does anyone know what the issue is here? This is my first time attempting to use AFNI so I may have missed something basic.
It could be something to do with adding the BN atlas to my ‘.afnirc’ file?

Thank you for your time,

Hi, Marcus-

Datasets have “view” and “space” properties—see slides 33-36 here:
afni_open -aw afni30_vol_dset_basics.pdf

These record/mark when a dataset is in a standard space of some kind (MNI, IBT, TLRC, NMT, etc.) vs in “original”/subject space. If you see the BRIK/HEAD file have “+orig” in it, it is in ‘orig’ space. If it has ‘+tlrc’, it is in some kind of standard space (confusingly, not necessarily Talairach-Tournoux specifically). You can also use:

3dinfo -av_space -space -prefix DSET1 DSET2 ...

to see the “view space” and specific space names on any dset (NIFTI, BRIK/HEAD, etc.)

The GUI does not allow “orig” and “tlrc” dsets to be overlaid. By looking at the name, I can see that “zz.errts.s620.fanaticor+orig.HEAD” is in orig space, and I am guessing that BN_Atlas_246_1mm.nii is in standard space. You can verify this mismatch of views with:

3dinfo -av_space -space -prefix zz.errts.s620.fanaticor+orig.HEAD BN_Atlas_246_1mm.nii

Is that errts file from Was a “tlrc” block used there, and are you sure that dataset should be overlayable on a template (i.e., did it get aligned to that space during processing)?

If you want to view that dset on a standard space template, you can use 3drefit to alter its header properties. Because 3drefit does not create a new dset, you should copy it first:

3dcopy zz.errts.s620.fanaticor+orig.HEAD  NEW_zz.errts.s620.fanaticor
3drefit -space TLRC -view tlrc NEW_zz.errts.s620.fanaticor+orig

… after which, you will see the NEW* file listed as “NEW_zz.errts.s620.fanaticor+tlrc.*”, and that should be overlayable on the other standard space data. (NB: any dataset with a standard space should be overlayable, they don’t need to all be specifically MNI or TLRC, say. If the brainnetome space is MNI specifically, you could put “3drefit -space MNI -view tlrc NEW_zz.errts.s620.fanaticor+orig”, if you wanted.)


Just to add to Paul’s advice, the BN_Atlas_246_1mm.nii.gz should be in your binary directory (probably ~/abin). I don’t know where the other uncompressed one comes from (without the .nii.gz). You may want to fetch that atlas again or even the whole binary package.