Desc-aparcaseg_dseg.nii.gz and MNI files missing after running fMRIprep

[Shahar_Ben_Noun]

Hi everyone, I’d appreciate your help regrading trouble I’m having with fMRIprep

Summary of what happened:

I ran the minimal level version of fMRIprep and was missing the files of the templates such as “sub-01_desc-aparcaseg_dseg.nii” and “sub-01_desc-aseg_dseg.nii”

I thought it might has to do with me running it in its minimal version so I tried running it again at full version but the problem recurred and even more concerning, I realized I don’t have the MNI transformed files either - no “sub-01_space-MNI152NLin2009cAsym_desc-brain_mask.nii”

Command used

singularity run -B /ems/elsc-labs/benyakov-a/Lab-Shared/BIDS_data/06_05_24_pilot/ -B /ems/elsc-labs/benyakov-a/Lab-Shared/collab_projects/2nd_run/imaging/ -B /usr/local/freesurfer/7.3.2/ --cleanenv /opt/ct/fmriprep_23.2.1.sif /ems/elsc-labs/benyakov-a/Lab-Shared/BIDS_data/06_05_24_pilot/ /ems/elsc-labs/benyakov-a/Lab-Shared/collab_projects/2nd_run/fmriprep_output participant -w /ems/elsc-labs/benyakov-a/Lab-Shared/collab_projects/2nd_run/imaging/fmriprep_work --fs-license-file /usr/local/freesurfer/7.3.2/license.txt --output-space T1w --nthreads 9 --participant-label 01 02 03

I thought this is rather unexpected as I got a message saying “fMRIPrep finished successfully!” and checking the html report I see no errors occured.

Thanks in advance!
Shahar

---

[Steven]

Hi @Shahar_Ben_Noun,

You did not include MNI as an output space in your fmriprep command.

Best,
Steven

[Shahar_Ben_Noun]

If I understood correctly the documentation at fmriprep version documentation fMRIPrep uses MNI152NLin2009cAsym as standardization reference by default, why wasn’t it the case now?

Also, will putting MNI in output space create the desc-aparcaseg_dseg.nii as well?

Thanks for the swift reply!

[Steven]

Hi @Shahar_Ben_Noun,

It is the default, but by specifying your own output space you override the default.

Best,
Steven

[Shahar_Ben_Noun]

@Steven

That makes perfect sense, thanks!

Will putting MNI in output space create the desc-aparcaseg_dseg.nii files as well?

[Steven]

I would think that would come just from using FreeSurfer, but it’s worth a shot.

[Shahar_Ben_Noun]

@Steven Alright :))
Thank you so much!

[effigies]

No (aparc+)aesg output from fmriprep

This probably got lost in the refactor. What spaces were they resampled into? I can’t remember off the top of my head where they used to go. They are pretty easy to pull out of FreeSurfer, though. Here you can resample in MNI with a particular BOLD run to ensure that your voxels align with that run: mri_convert sourcedata/freesurfer/sub-XYZ/mri/aseg.{mgz,nii.gz} antsApplyTransforms \ -i sourcedata/freesurfer/sub-XYZ/mri/aseg.nii.gz \ -r sub-XYZ/func/sub-XYZ_task-TASK_space-MNI152NLin20…

[Shahar_Ben_Noun]

@effigies Thanks! fMRIPrep is still re-running and hopefully it’ll be solved now but if not I’ll be sure to check out your solution

[Shahar_Ben_Noun]

Continuing the discussion from Desc-aparcaseg_dseg.nii.gz and MNI files missing after running fMRIprep:

Summary of what happened:

In before, when I ran fMRIPrep it had automatically created files such

fmriprep_output\sub-01\anat\sub-01_desc-aparcaseg_dseg.nii.gz

For some reason it longer creates these files, at least not by default.
Why doesn’t it do it automatically as it used to? Does anyone know about an arguement I’m missing when running fMRIPrep?

Command used

 singularity run -B /{path}/BIDS_data/06_05_24_pilot/  -B /{path}/06_05_24_pilot/imaging/ 
-B /{path}/freesurfer/7.3.2/ --cleanenv /opt/ct/fmriprep_23.2.1.sif  /{path}/BIDS_data/06_05_24_pilot/ 
 /{path}/06_05_24_pilot/fmriprep_output  participant -w /{path}/06_05_24_pilot/imaging/fmriprep_work 
--fs-license-file /{path}/freesurfer/7.3.2/license.txt --output-space T1w MNI152NLin2009cAsym --nthreads 9 
--participant-label 01 02 03 --slice-time-ref 0

**Many thanks to @‘effigies’ who helped me find an older thread about converting the .mgz files to nifti in this post, allowing me to continue my work in the meantime

---

[Shahar_Ben_Noun]

Hi,

I actually came back to this matter. Thanks to the replies I got in this thread I’m no longet stuck at the matter, but I still don’t get -

Why doesn’t fMRIPrep create desc-aparcaseg_dseg files automatically as it used to? Am I missing an input when running fMRIPrep?

[bchaloup]

Hello,

I am having the same issue. I can also use the workaround and additionally preprocess the aparc+aseg.mgz file in the sourcedata/freesurfer/sub-XX/mri folder, but why doesn’t fmriprep create the desc-aparcaseg_dseg.nii.gz files?

[gusma]

I’m running into the same problem using fmriprep 24.1.1, would be nice to have the (apart)aseg outputs automatically, or an option to ask for them.

[bchaloup]

@gusma From what I’ve discovered, the last version of fMRIPrep that still creates the desc-aparcaseg_dseg.nii.gz files is version 23.1.4. However, running that version creates tons of files in the work folder (multiple TBs worth of space). I’m guessing that’s part of why they refactored the code base, but no version since then outputs these files automatically or even has a flag that you can request them with. Hopefully that will get fixed soon. For now, the options seem to be to run an older version and clean up the work folder as you go, or run a newer version and manually run the transforms on the aparc+aseg.mgz files from the freesurfer output.