I am running fmriprep --anat-only and in the mri folder it has appeared the following error.log that it says:
mghRead(/home/rosalia/Desktop/FREE_NEW/aseg.manedit.mgz, -1): could not open file
How are you running fmriprep? It’s possible that not all of those resources are being devoted to running it. For example, Docker has its own memory limitations which you have to change, separate from the overall machine. When I run fmriprep, the FreeSurfer part should not take anymore than 1 day, and I devote 20gb and 8 cpus. At what point does your original error occur, and does it cause the whole process to stop?
How could I change those memory limitations? Because the process is only using 1 cpu…and minimum percentage of the others. The process has not stopped but it is running from Tuesday, 8.30 am. The error occurred on Wednesday…
It would be nice if you tell me how to change the memory limitations…
You can try submitting it as an sbatch job where you can explicitly state the memory usage and have it run in the background. Do you know how to do that?
Then under that paste in your singularity command. Then in terminal, run sbatch fmriprep.sh and your job should begin running in the background. You will also notice a log file appear in the location you run the command which you can use to monitor your progress.
Thank you very much.
A subject finished without error, however, after I checked it I saw that some anatomical regions (mainly frontal and temporal regions) were not fully covered
Glad to hear it finished! Could you possibly post a screenshot of the HTML for that subject so I can see what you mean? And is the coverage there in the original T1 and fMRI file?
I did not use fMRI this time because the fMRI had problems. So, I thought just process the structural data for now…and once the fMRI problem is solved, include it.
If you have a brain viewing program, such as FSLeyes (included with FSL), Freeview (included with Freesurfer), or Mango (standalone application Research Imaging Institute — Mango), you can view the T1 file to see if this is a problem with fMRIPrep or a problem with the T1 file.