[RESOLVED] Dwi_legacy_gpu extra_eddy_args parameter

Issues
1

Description:

Dear QuNex team,

I was trying to run dwi_legacy_gpu command with the –extra_eddy_args. However, dwi_legacy_gpu will only pass the first parameter to eddy command.
Any help would be greatly appreciated.

Many thanks,

Selena

Call:

qunex dwi_legacy_gpu \
    --sessionsfolder="${StudyFolder}/sessions" \
    --sessions="${SessionBatchFile}" \
    --sessionids="${Session}" \
    --overwrite="${Overwrite}" \
    --pedir='2' \
    --unwarpdir='y' \
    --echospacing="$echoSpacingMsec" \
    --diffdatasuffix="DWI_dir75_AP" \
    --usefieldmap='yes' \
    --te='2.46' \
    --extra_eddy_args='--niter=8 --fwhm=10,8,6,4,2,0,0,0 --nvoxhp=2000 --mb=2 --ol_nstd=5 --mporder=16 --s2v_niter=5 --data_is_shelled --repol --ol_type=both --with_outliers --estimate_move_by_susceptibility'

Logs:

Running dwi_legacy_gpu with the following parameters:
--------------------------------------------------------------
   Study Folder: /scratch/shingshiun/QuNexMamah_NAPLS3_dMRI
   Sessions Folder: /scratch/shingshiun/QuNexMamah_NAPLS3_dMRI/sessions
   Sessions: 10340_20170828
   Study Log Folder: 
   Using FieldMap: yes
   Echo Spacing: .339997000
   Phase Encoding Direction: 2
   TE value for Fieldmap: 
   EPI Unwarp Direction: y
   Diffusion Data Suffix Name: DWI_dir75_AP
   Overwrite prior run: yes
   No GPU: no
   Extra Eddy Args: --niter=8


--- Full QuNex call for command: dwi_legacy_gpu

/opt/qunex/bash/qx_utilities/dwi_legacy_gpu.sh     --sessionsfolder=/scratch/shingshiun/QuNexMamah_NAPLS3_dMRI/sessions     --session=10340_20170828     --usefieldmap=yes     --pedir=2     --echospacing=.339997000     --te=2.46     --unwarpdir=y     --diffdatasuffix=DWI_dir75_AP     --overwrite=yes     --nogpu=no     --extra_eddy_args='--niter=8'

** For an example of how to report an issue, please refer to this post.

2

Hi,

Thanks for reporting this. Yes, this seems like a bug with the current version. The next version that fixes this will be released in July/August.

The workaround is to call the dwi_legacy_gpu script directly, and bypass the QuNex entry point, the results will be the same. The only thing missing will be the output log, output will be now printed only to the console, but you can redirect that to a log file as well. So:

/opt/qunex/bash/qx_utilities/dwi_legacy_gpu.sh \
  --sessionsfolder="/scratch/shingshiun/QuNexMamah_NAPLS3_dMRI/sessions" \
  --session="10340_20170828" \
  --usefieldmap="yes" \
  --pedir="2" \
  --echospacing="0.339997000" \
  --te="2.46" \
  --unwarpdir="y" \
  --diffdatasuffix="DWI_dir75_AP" \
  --overwrite="yes" \
  --nogpu="no" \
  --extra_eddy_args="--niter=8 --fwhm=10,8,6,4,2,0,0,0 --nvoxhp=2000 --mb=2 --ol_nstd=5 --mporder=16 --s2v_niter=5 --data_is_shelled --repol --ol_type=both --with_outliers --estimate_move_by_susceptibility"

Best, Jure

3

Hi Jure,

I still got the same error when calling the dwi_legacy_gpu script directly.

I would love to test the next QuNex version when it’s available.

Best,

Selena

4

Ah my bad, there was another change in this script as well to fix the bug. What you can do is save the script attached below on your disk, say as /home/selena/dwi_legacy_gpu.sh and then bind it over the script in the container, so with docker you would add:

-v /home/selena/dwi_legacy_gpu.sh:/opt/qunex/bash/qx_utilities/dwi_legacy_gpu.sh

If you are using singularity I think the parameter is --bind or -B. This will replace the script in the conatiner with the attached one, which should work. I can also share a dev container with you once we have it that fixes this and you can run with the proper QuNex command.

Best, Jure

dwi_legacy_gpu.sh (36.4 KB)

5

Hi,

was this resolved? 1.5.0 will be released in August, so you will no longer need the workaround.

Best, Jure