Problems with brain extraction in fMRIprep ants

[irebollo]

Hi All

I’m running fmriprep on an old dataset and brainextraction is failing in 48/49 subjects, giving outputs like this:

Im running fmriprep through docker using the following command

fmriprep-docker /mnt/RAIDSTORAGE/MONTEVIDEOREST/Nifti /mnt/RAIDSTORAGE/MONTEVIDEOREST/derivatives --participant-label c01 --fs-license-file /home/irebollo/freesurfer/license.txt --fs-subjects-dir /mnt/RAIDSTORAGE/MONTEVIDEOREST/freesurferdata --omp-nthreads 16 --output-spaces fsaverage MNI152NLin2009cAsym:res-3 --anat-only

Im using a linux mint PC, 64gb of ram, 32 cpus
fmriprep version 20.0.4

The problem seems related to this issue

fMRIPrep: spatial normalization failed

Dear all, My output from the fmripreproc preprocessing looks very strange! The co-registration is way off and the striatum has been moved to the frontal cortex? Does anyone know why this might of happened? I did not use skull stripped images - would this help? I’m attaching example Asym_desc_prerpoc_T1 Faith [image]

, although I understand that there is no working directory issue since im running it through docker?

Please find here the report of the subject

And here the defaced anatomy

I am now running fmriprep with freesurfer (instead of providing the path to freesurfer files) to see if it solves the issue

Thank you very much for your help!

[irebollo]

Strangely, recomputing freesurfer segmentation with fmriprep has solved the issue.
For info, I had the images processed with freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c