neuro-ml / bidstransformers

BIDS Transformers for neuroimaging data preprocessing based on BIDS and sklearn architecture

BIDSTransformers

This is a very basic proof of concept of BIDS Transformers. It implements two transformers: one based on FSL ROI extractor and second based on FSL Skull Stripping.

Example notebook is BIDS_Transformers.ipynb. To run this notebook you will need the following:

1. Clone this repository
2. Working Nipype installation. You can also use docker container (see instructions below).
3. PyBIDS package (pip install pybids)
4. Download data folder from one of the following links and put it into the folder /bidst/tests/data

• Link1
• Link2

Installing Nipype

You can follow the Nipype tutorial.

Using docker

Getting neuro_docker container:

git clone https://github.com/Neurita/neuro_docker.git
cd neuro_docker
docker build -t="dockerfile/neuro" .
cd ..

Downloading repository

git clone https://github.com/neuro-ml/BIDSTransformers.git
cd BIDSTransformers

Running docker:

sudo docker run -v $PWD/:/work/soft/BIDSTransformers -it -p 8809:8809 dockerfile/neuro

Installing needed packages in the docker:

pip install notebook pybids duecredit
apt-get update
apt-get install tree

export PATH=/work/soft/ants/build/bin:$PATH
export ANTSPATH=/work/soft/ants/build/bin/

Starting jupyter notebook in the docker:

jupyter notebook --no-browser --ip="*" --allow-root --port 8809

You will see message:

  Copy/paste this URL into your browser when you connect for the first time,
  to login with a token:
    http://localhost:8809/?token=some_token

git checkout crn-code-sprint2017

You will need to open this url in your browser and copy the token.

bidstransformers's Issues

Implement NUCorrectionTransformer

What does: applies N3-correction to image.
Possible backends: ANTS, Freesurfer
Input: T1w of subject.
Outputs: N3-corrected image
BIDS reference: can’t find one. Maybe we will need to produce draft in their google doc.
Nipype references: http://nipype.readthedocs.io/en/latest/interfaces/generated/interfaces.ants/segmentation.html#n4biasfieldcorrection

Write BIDS transformer for dcm2nii

Write transformer for conversion from dicom to nii if dicoms in data structure are organized by bids.

https://github.com/neuro-ml/fmri_preprocessing/blob/master/fMRI_preprocessing.ipynb