QSM pipeline: version update error

admin · May 3, 2021, 2:49pm

Hi Alan,

For my project we’re collecting multi-echo GRE images at 9.4 T to generate QSM maps. Ali updated your QSM pipeline to v2.0.0 to account for some problems generating the phase images specific to my project. He also added qsm_sstv_v2.0.0 as a bidsBatch app.

However, I keep getting the same error message when trying to run the qsm pipeline and it is:

201119-12:59:51, 453 pipype.uits WARNING:
Could not check for version updates:
Could not find a suitable TLS CA certificate bundle, invalid path: /etc/pki/tls/certs/ca-bundle.crt

I receive this message no matter which way I try to run it (bidsBatch, batch script, singularity run) on graham. I also receive it when trying to run your original qsm pipeline, qsm_sstv_v0.1.1.

Ali was able to get the new pipeline to run on his end but he wasn’t able to get it to run when I was going over it with him on my account. He checked my singularity, python, etc. versions and they seemed okay.

Do you know what the environment issue may be and if there is something I can do on my end to get it running?

Thanks,
Sara

akuurstr · May 4, 2021, 3:50pm

Hi Sara,

Welcome to the beta version of the forum! Thanks for helping us out with the testing. If you are “Watching” this thread you will receive emails with my replies.

Nipype is a python package that is used as part of the pipeline. It seems that when Nipype is imported, it wants to check to see if it’s using the most recent version - however it seems to be having trouble finding the certificates it needs to do this. To be honest, this warning can be ignored for now because we are happy to use the version of nipype the pipeline was built on and we don’t care about newer versions.

It is actually a different reason that your data is not being processed.

The first thing I noticed in your bids data directory is that whereas there are magnitude images, there are no phase images (ie. sub-2087_ses-LH_part-phase_echo-1_run-02_GRE.nii.gz). We measure the susceptibility’s affect on the magnetic field using the phase images and these are necessary for the pipeline to run correctly.

Do you know why the phase images are missing? I would encourage you to look into the dicom to bids conversion.

akhanf · May 3, 2021, 7:17pm

Hi guys,

Sara, for this data (from the 9.4T) the images were complex valued and a script to generate the phase data was added on to tar2bids I believe – can you point me to the data too and what version of tar2bids you are using?

Ali

admin · May 3, 2021, 2:49pm

Hi,

Thanks for the help and for looking into this!

That makes a lot of sense, I didn’t realize the phase images were not in the bids folder. I’ll have to rename this topic I started!

I had a few different folders where I was testing this, all with variations of the tar2bids so probably was using the wrong one. I believe I was previously running it with khanlab_tar2bids_v0.05h.sif. There’s also the genPhaseFromComplexNii.m I found in another folder.

I’ve added these to your scratch folder Alan, and Ali you should have permission to the ‘test_akuurstra_qsm’ folder at /scratch/spac/.

I’ll try running it with these different tar2bids options to see which one is the correct one, maybe on a fresh dataset so I start from the beginning.

akhanf · May 3, 2021, 7:18pm

Btw this is where that script came from:

github.com

khanlab/tar2bids/blob/master/etc/genPhaseFromComplexNii.py

#!/usr/bin/env python

import sys
import numpy as np
import nibabel as nib
from glob import glob
import re
import os
import json

def usage():
  print('{}: <in_anat_dir> (containing part-imag and part-real images)'.format(sys.argv[0]))

if len(sys.argv)-1 < 1:
  usage()
  sys.exit(1)

in_anat_dir = sys.argv[1]

This file has been truncated. show original

Looks like a tar2bids release was never made since that addition (v0.05h was released days before that script was added!), so it would actually only be in docker://khanlab/tar2bids:latest