volgenmodel-nipype

volgenmodel-nipype is the port of volgenmodel to Nipype. It creates nonlinear models from a series of input MINC files.

Install for Windows Subsystem for Linux or Ubuntu 16.04

Install minc: https://bic-mni.github.io

    wget http://packages.bic.mni.mcgill.ca/minc-toolkit/Debian/minc-toolkit-1.9.16-20180117-Ubuntu_16.04-x86_64.deb
    sudo apt-get install libc6 libstdc++6 imagemagick perl octave
    sudo dpkg -i minc-toolkit-1.9.16-20180117-Ubuntu_16.04-x86_64.deb
    sudo apt-get install libgl1-mesa-glx libglu1-mesa
    rm minc-toolkit-1.9.16-20180117-Ubuntu_16.04-x86_64.deb

    vi .bashrc
    source /opt/minc/1.9.16/minc-toolkit-config.sh
    export PERL5LIB=/opt/minc/1.9.16/perl

Clone code and test-data:

git clone https://github.com/CAIsr/volgenmodel-nipype.git
git clone https://github.com/CAIsr/volgenmodel-fast-example.git

install miniconda:

wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

start new terminal and install packages required:

conda install --channel conda-forge nipype
pip install pydot

add additonal scripts to PATH

cd volgenmodel-nipype/extra-scripts
echo "export PATH="`pwd -P`":\$PATH" >> ~/.bashrc

you need a working octave installed. Somehow the minc libs break octave. Sometimes this fixes it: select /usr/lib/lapack/liblapack.so.3

sudo update-alternatives --config liblapack.so.3

or load octave using a module:

module load octave/4.2.1

start new temrinal and run volgenmodel with the test data:

cd volgenmodel-nipype/
python3 volgenmodel.py --input_dir ../volgenmodel-fast-example

The final model should look like this:

It's possible to apply the estimated flow fields to other images that where aligned with the original scans. Here is one example how this can be done (atlas was estimated on Magnitude data and applied to QSM data):

./new_data_to_atlas_space.py \
    --name "QSM template" \
    --xfm_dir "data/magnitude_template/workflow/xfmconcat_9" \
    --xfm_pattern "*/*.xfm" \
    --source_dir "data/qsm/qsm_final/mnc" \
    --source_pattern "*.mnc" \
    --atlas_dir "data/magnitude_template/workflow/voliso_9" \
    --atlas_pattern "*.mnc" \
    --work_dir "data/qsm_template/work" \
    --out_dir "data/qsm_template/out" 

Citation

This method is an implementation of the technique described in this paper:

http://www.ncbi.nlm.nih.gov/pubmed/25620005

If you use it in a publication please cite:

Janke AL, Ullmann JF, Robust methods to create ex vivo minimum deformation atlases for brain mapping. Methods. 2015 Feb;73:18-26. doi: 10.1016/j.ymeth.2015.01.005