Skip to content

Latest commit

 

History

History
64 lines (45 loc) · 3.27 KB

SettingUpDirectories.md

File metadata and controls

64 lines (45 loc) · 3.27 KB
Raw

Processing off-scanner data


Set up the directory structure


Set a subject ID: snum=$01; sbj=sub-${snum}

cd /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/Data/
mkdir ${sbj}
cd ${sbj}
mkdir DataOffScanner Unprocessed
cd DataOffScanner
mkdir biopac DICOM psychopy


Move data to correct directories post-scan


Document scanner notes:

  • Create a document named ${sbj}_ScanningNotes.txt in /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/${sbj}


Move PsychoPy logs & biopac .acq files to specified locations:

  • Copy PsychoPy logs for all tasks to /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/${sbj}/Data/DataOffScanner/psychopy
  • For the WordNonword task:

    • To extract timing files for AFNI GLMS, run a similar command:

      python [path to ComplexMultiEcho1 code repo]/PsychoPy/WordNonword/CreateEventTimesForGLM.py --sbjnum 3 --runnums 7 8 9 --rootdir /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/Data/

      Note: Look at the stdout of that script to confirm that there's little to no difference between expected and actual trial times, the TR is <1ms of 1.5sec, and the participant's behavioral response rates show some level of attention. sub-??_CreateEventTimesForGLM.log contains more details about responses during each run

  • Copy biopac data to /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/Data/${sbj}/DataOffScanner/biopac


Document DICOM scanner information (contains PII):

  • Put deidentified DICOM README-Study.txt in /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/Data/${sbj}/DataOffScanner/DICOM
  • README-Study.txt has PII in the following fields at the top: Subject Name, Subject ID, Study Date, Study ID, Accession Number and the scan date is in the repeated "Organized" field


On a local computer where the DICOMs with PII are downloaded run a similar command: python bespoke_cme.py --ignore 3 4 5 18 19 20 27 28 29 --start 2 ./DICOM ./tmp ${snum} > ${sbj}_dcm2nii_mapping.txt

  • Option parameters:
    • Ignore: ignored numbers are the mr_00?? directories that should not be reconstructed.
    • Start: DICOM volume at which to start?
  • Data Requirements & Assumptions:
    • This script assumes the data are collected in the order: MPRAGE, 3 runs of WNW, movie, breathing, movie, breathing (If that's not the order, corrections may be necessary.)
    • This script requires having an up-to-date version of dcm2niix in the path (dcm2niix_afni is shipped with AFNI)
    • Output is piped to dcm2nii_mapping.txt. (That file should be viewed to make sure the mapping of DICOM to meaningful names was done correctly)


Transfer BIDS-ified files to Biowulf & add header information:

  • Move BIDSified data to /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/${sbj}/Data/UnProcessed
  • Run AddingSliceTiming.sh ${sbj} to add slice timing information to the header. (The BIDSified data need the slice timing information in the nii file headers.)
  • Delete DICOM data with PII from computer


Set directory permissions for Data directory

chgrp -R SFIM /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/Data/${sbj}
chmod -R 2770 /data/NIMH_SFIM/handwerkerd/ComplexMultiEcho1/Data/${sbj}


To convert physiological BIOPAC Acqknowledge (.acq) files to BIDS format

See ../PhysioProcessing/Physiological_Proc.md