asl-preprocessing

Workflow to obtain ASLPrep output

1. Run fMRIPrep

   • ASLprep reuses the anatomical derivatives from fMRIPrep output

2. Create a new dataset of fMRIPrep output using Datalad

   • Run datalad install to make a copy of the fMRIPrep output: code/datalad_install_dataset.sh
   • Rename files and modify folder hierarchy: code/datalad_rename_dataset.sh
     • datalad unlock the dataset
     • Run rename.ul
     • datalad save the changes

3. Run ASLprep

   • Create a json file describing custom BIDS input filters: code/aslprep_filter.json
     • To only include baseline T1w and ASL scans
   • Run the bash script: code/run_tay_aslprep.sh
     • Submit SLURM jobs
     • Use output spaces MNI152NLin2009cAsym and asl
     • Set --smooth_kernel to 0
     • Set --m0_scale to 100
     • Include --bids-filter-file, --anat-derivatives, [--skip_bids_validation

4. QC ASLprep output

   • QC file: sub-CMH*/ses-01/perf/*desc-quality_control_cbf.csv

5. Run statsitical analysis

   • Havard Oxford atlas: sub-CMH*/ses-01/perf/*desc-HavardOxford_mean_cbf.csv

Contact

Ghazaleh Ahmadzadeh