RTP2: Reproducible Tract Profiles
Garikoitz Lerma-Usabiaga
Mengxing Liu
BrainHack Donostia
Diffusion
Combining quantitative MRI (qMRI) maps with white matter tracts.
https://github.com/garikoitz/RTP-pipeline
4
Tasks for all levels
Basic
We will select some already preprocessed data so that we can only test the qMRI part.
https://github.com/garikoitz/RTP-pipeline
I think that the lowest level of involvement will be to learn how to launch the pipeline and be responsible of testing and visualizing results. The intermediate level of involvement will correspond on coding the data organization and container launching scripts. The highest level of involvement will be coding the Python and Matlab codes within the container so that it can read qMRI data and provide the tract specific metrics. The three levels or involvement will run in parallel. First milestone will be to plan what each parallel line will need to accomplish and how (so training on the tool will be given). The second milestone will be to do the actual coding. The third milestone will be to combine the three branches for testing and improving it iteratively until it is working.
What is diffusion imaging and how it works. qMRI. Running a diffusion pipeline from dicom images to tract metrics. Containerization technology (Docker and Singularity). Python and Matlab. Git (working with a branch within a fork and creating a pull request)
3
They will be in a contributing page in the main repo.
Pipeline development
Matlab, Python, Containerization (Docker and Singularity)
Basic (commit & push)
DWI, MRI, qMRI
AFNI, ANTs, BIDS, fMRIPrep, Freesurfer, FSL, MRtrix, the pipeline uses many different softwares…