VU-Cog-Sci / mp2rage_preprocessing

mp2rage_preprocessing

This is the preprocessing workflow of the Knapen lab at the Spinoza Centre for Neuroimaging in Amsterdam, to get Freesurfer and CBS-tools segmentations of a combined MP2RAGE and MEMP2RAGE protocol.

Installation

1. Install Docker
2. Install docker-compose
3. Clone this repository

Running

Set environment variables

• $SOURCEDATA to your folder with data
• $DERIVATIVES to the folder where you want the output of the pipeline to go
• $FREESURFER_HOME to the Freesurfer folder with a license.txt-file

Then go to the folder where you cloned the repo and just run ./go

You will now be in an virtual environment with ZSH, where you can run all the scripts.

Step 1: combining the two scans

python /src/combine_scans.py <SUBJECT> <SESSION>

• Calculates UNI, T1map, T2*-map, S0-map.
• Registers the T1w-images of the mp2rage to the me-mp2rage space
• Makes average images in this common space
  • /derivatives/average_space

Step 2: masking the image

python /src/mask_averages <SUBJECT> <SESSION>

• Makes a brain mask using
  • Inhomogeniety-corrected average INV2 and BET
  • Dura filters of CBS-tools
  • Manual mask, of non-brain matter (sagital sinus)
    • Should be stored in /derivatives/manual_nonbrainmask/sub-<SUBJECT>/ses-<SESSION>/sub-<SUBJECT>_ses-<SESSION>_manual_nonbrainmask.nii.gz
• Outputs to /derivatives/masked_averages and /derivatives/sourcedata_fmriprep

Step 3: fmriprep

To be implemented (inside this docker or outside this docker?)

3 experiments to run:

• Only average T1w
• Average T1w and average INV2 as FLAIR
• Average T1w and average T1map as T2w