Skip to content

Latest commit

 

History

History
54 lines (45 loc) · 3.03 KB

README.md

File metadata and controls

54 lines (45 loc) · 3.03 KB
Raw

diffusion-mri

This projects contains code to download and process diffusion data from Healthy Brain Network with QSIprep and other tools.

Project for Neurohackacademy 2024. Contributors:

Background information about Diffusion Imaging

A Short Introduction to Diffusion MRI

Running QSIprep preprocessing pipeline

  1. Get the data: /code/001-get-hbn-data_lkpo.ipynb
    • Make sure to have utilities.py under /code/
    • We worked with the unprocessed data under BIDS_curated folder. Each subject should have an anat, dwi and fmap folder.
    • Data should be downloaded to a data folder to comply with BIDS format
    • Make sure to have data_description.json under the BIDS dataset folder (You can find it under /code/dataset_description.json).
    • You will need a txt file with the FS_license
    • Make sure that fmaps belong to the dwi images. We removed the fMRI fmaps manually: rm -rf /tmp/cache/data/sub-*/*/*fmri*
  2. Run QSIprep preprocessing: /code/002_Run_QSI_Prep.sh
    • Create singularity image in diffusion_mri folder by typing on terminal:
      • singularity build ./my-qsi-prep.sif docker://pennbbl/qsiprep:0.22.1
    • To run the script do on Terminal: ./002_Run_QSI_Prep.sh <SUBID>
    • Modify all paths according to
      1. singularity image
      2. BIDS formatted data directory
      3. Output directory
      4. No need to modify subject id within the script
      5. Look at acquisition parameters to obtain the voxel resolution or modify according to desired voxel size
      6. Point to your freesurfer license
    • After it is done, manually inspect the HTML and figures files for each subject
  3. Run QSIprep reconstruction /code/003_Run_QSI_Recon.sh
    • To run the script do on Terminal: ./003_Run_QSI_Recon.sh <SUBID>
    • Modify all paths according to
      1. Singularity image
      2. BIDS formatted data directory
      3. Output directory with qsirecon
      4. Input directory with the qsiprep outputs from the previous step
      5. Specify the reconstruction model you want to use. For reference: https://qsiprep.readthedocs.io/en/latest/reconstruction.html
      6. Look at acquisition parameters to obtain the voxel resolution or modify according to desired voxel size
      7. Point to your freesurfer license

More info

For more information regarding the present project and instructions to use the jupyter notebooks, visit the documentation page in this repository.