Updated version of "https://github.com/poldracklab/ds003-post-fMRIPrep-analysis/tree/" package for FMRI brain imaging analysis

Clone or download the repository

git clone https://github.com/joshjzhou/post-fMRIPrep-analysis.git

Install dependencies

pip install -r requirements.txt

Update the contrast in workflows.py in line 56. An example contrast looks like: contrasts=[('intask', 'T', ['word', 'pseudoword'], [1, 1])].

The file structure should follow the BIDS standard format. An example file structure looks as follows:

• run.py
• interfaces.py
• workflows.py
• Preprocessed (folder)
  • fmriprep (folder) <--- this is the derivatives directory
    • sub-(number) (folder)
      • anat (folder)
      • figures (folder)
      • ses-post (folder)
      • ses-pre (folder)
• Nifti (folder) <--- this is the bids directory
  • sub-(number) (folder)
    • ses-pre
    • ses-post

Run the following command to start the analysis

python [relative filepath of run.py] [relative filepath of the derivatives directory] [relative filepath of the output folder] [analysis level (either participant or group)] --space [specific space to be processed] --bids-dir [relative filepath of bids directory] --participant-label [can be a single participant or multiple] --task [type of task] -w [relative filepath where intermediate documents should be stored]

Example command:

python run.py "preprocessed/fmriprep" "results" participant --space MNI152NLin2009cAsym --bids-dir "Nifti" --participant-label 104 119 --task nback -w "work"

Everything in quotes is a folder. More possible inputs can be found at the beginning of run.py. If space is not specified, it will default to fsaverage. Here is a list of the possible space inputs: 'MNI152Lin', 'MNI152NLin2009cAsym', 'MNI152NLin6Asym', 'MNI152NLin6Sym', 'MNIInfant', 'MNIPediatricAsym', 'NKI', 'OASIS30ANTs', 'PNC', 'fsLR', 'fsaverage'.

You'll likely need to install additional modules such as fsl. The updated run.py has not yet been tested with group level analysis, so if you do run into any errors, please feel free to contact me.