Mysterious recon-all failure

nipreps / fmriprep

fMRIPrep is a robust and easy-to-use pipeline for preprocessing of diverse fMRI data. The transparent workflow dispenses of manual intervention, thereby ensuring the reproducibility of the results.

https://fmriprep.org

Apache License 2.0

637 stars 295 forks source link

Mysterious recon-all failure #511

Closed effigies closed 7 years ago

effigies commented 7 years ago

Moving from #425, since this seems to be a legitimately different issue.

cc @mike21499

mike21499 commented 7 years ago

Thanks for your comments: Given that the entire pipeline ran (AFAICT), there are two possibilities I can see:

1 ) Our use of an external skull-stripping pipeline produced a mask that FreeSurfer could not use. 2 )There's some issue with your data that FreeSurfer is unable to automatically correct for.

To check for (1), if it isn't evident in the report, you can overlay the brainmask (/fmriprep/anat/sub-01_T1w_brainmask.nii.gz) on the preprocessed T1 (.../sub-01_T1w_preproc.nii.gz) in mricron or Freeview, and verify that gray matter hasn't been removed. recon_all_failure_overlay

I've tried your approach 1). Attached is the result. It seems that grey matter has been removed.....

thanks for your efforts, michael

effigies commented 7 years ago

Hi @mike21499. It's difficult to be sure from two slices and an opaque overlay, but that looks like a pretty reasonable mask. If anything, it seems to have included a bit too much near the occipital lobe.

I think at this point I'd recommend going ahead and trying recon-all in a fresh directory without going through fmriprep.

If you're able to provide a defaced copy of your T1w image, I can try running the anatomical pipeline and see if I can reproduce your issue.

mike21499 commented 7 years ago

@effigies . That would be great. Attached is a copy of the anat-folder (no worries, it's my brain :) anat.zip

effigies commented 7 years ago

Edit: Nevermind. It just took a very long time to download.

effigies commented 7 years ago

I cannot reproduce this bug. fmriprep ran without error, including recon-all. Have you tried running on fmriprep 0.4.3?

effigies commented 7 years ago

@mike21499 Just checking in.

A few questions:

How are you running this? Our recommended approach is the fmriprep-docker wrapper script, or Singularity, if you're running on a cluster environment.
- Do you have enough memory and disk space?
- Can you try re-running with the most recent release?

mike21499 commented 7 years ago

Hi effigies

Problem solved. It was a memory issue. We assigned 16GB of RAM to one subject and it ran smootly!

thanks for your efforts, michael

effigies commented 7 years ago

Great! Thanks for checking back in.