Rabbit Brain Analysis

The pipeline to manipulate and analyse MRI data from the "Encephalopaty of prematurity" project, in collaboration with KU Leuven.

This is a data-specific code developed for the encephalophaty of prematurity project.
Input of the pipeline is the subject scanned in the native Bruker format.
Output of the pipeline is the subject oriented and segmented provided with subject-wise and comparative data analysis.
WARNING the code is developed to be research code (read: 80% is an uncommented work in progress, not packaged to be a product. See research code definition)
The code is developed in collaboration with University College London (UCL) and Katholieke Universiteit Leuven (KUL).
Pipeline schema can be found under notes folder.
Based on libraries in requirements.txt (there are non pip-installable dependencies: follow installation instructions directly from the README library repository) other than niftyReg niftySeg and FSL
Package main_pipeline has the structure of the pipeline as in the schematic of the documentation.
Each module of the pipeline can run independently (for debugging and single step analysis) or as part of the whole pipeline, if called by the main.
The code can be used for any dataset (with the required changes) and the authors intend to make the acquired dataset publicly available at the end of the study.

Connect to the dataset (NAS Emporioum if at KCL) and the rabbit dataset folder structure (can be used also on the cluster after updating dataset).
Run A0_main/subject_parameters_creator.py to have the latest parameters file (created parameters file connects the subject name with its chart, containing all the information related to the subject - e.g. study, category, orientation...)
Select the parts that you want to run and the subject you want to apply the pipeline in the A0_main/main_executer.py
Run main_executer.py.
Raise an issue if something goes wrong.

Update the log of the received files in the root of rabbit analysis in Emporium.
Follow existing folder structure (see under notes) and store the .zip in the appropriate place.
Rename the zip to only the filename (following the existing structure).

add the chart of the new subject in the file A0_main/subject_parameters_creator.py with default values.
start the A0 phase of the pipeline.
see the converted subject and update the parameters under A0_main/subject_parameters_creator.py according to visual intuition.
re run A0_main/subject_parameters_creator.py to update parameters.
run the whole pipeline main_executer.py for the selected subject.

Main work in progress is under danny_approval. Danny (see crash testing) is a dummy dataset created with the github repository DummyForMRI. Under danny approval we intend to develop some automatic testing to automatically perform the pipeline on the Danny, and check possible issues.
Improve parameters files subject_parameters_creator.py. Currently there is a single file instantiating an element of the class SubjectParameters. Re-design with yaml parameter file yaml_parameters_checker method and a yaml_parameters_creator.

To push latest commits on the GIFTSurg repo use

#git remote add origin_gs https://github.com/gift-surg/RabbitBrainAnalysis.git
git push origin_gs master

To push latest commits on the CMIClab platform (until 15-sept 2018) use

git push origin_gs master

To check the remote origin in your local repository type:

git remote -v

Acknowledgements

This repository is developed within the GIFT-surg research project.
This work was supported by Wellcome / Engineering and Physical Sciences Research Council (EPSRC) [WT101957; NS/A000027/1; 203145Z/16/Z].