LabelFusion

This repo contains implementation of various label fusion approaches that can be used to fuse multiple labels.

Installation

For Usage

conda create -n venv_labelFusion python=3.6.5 -y
conda activate venv_labelFusion
pip install LabelFusion

For Development

# fork to your own repo
git clone ${yourFork_labelFusion_repo_link}
cd LabelFusion
conda create -p ./venv python=3.6.5 -y
conda activate ./venv
pip install -e .
# develop, push
# initiate pull request

Available LabelFusion:

Voting (ITK): DOI:10.1016/j.patrec.2005.03.017
STAPLE (ITK): DOI:10.1109/TMI.2004.830803
Majority Voting: DOI:10.1007/978-3-319-20801-5_11
SIMPLE: DOI:10.1109/tmi.2010.2057442

Usage

Command-Line interface

# continue from previous shell
python fusion_run -h
  -h, --help        show this help message and exit
  -inputs INPUTS    The absolute, comma-separated paths of labels that need to be fused
  -classes CLASSES  The expected labels; for example, for BraTS, this should be '0,1,2,4' - not used for STAPLE or ITKVoting
  -method METHOD    The method to apply; currently available: STAPLE | ITKVoting | MajorityVoting | SIMPLE
  -output OUTPUT    The output file to write the results

Example:

# continue from previous shell
python fusion_run \
-inputs /path/to/seg_algo_1.nii.gz,/path/to/seg_algo_2.nii.gz,/path/to/seg_algo_3.nii.gz \
-classes 0,1,2,4 \
-method STAPLE \
-output /path/to/seg_fusion.nii.gz

Python interface

# assuming virtual environment containing LabelFusion is activated
import SimpleITK as sitk
from LabelFusion.wrapper import fuse_images

label_to_fuse_0 = '/path/to/image_0.nii.gz'
label_to_fuse_1 = '/path/to/image_1.nii.gz'

images_to_fuse = []
images_to_fuse.append(sitk.ReadImage(label_to_fuse_0, sitk.sitkUInt8))
images_to_fuse.append(sitk.ReadImage(label_to_fuse_1, sitk.sitkUInt8))
fused_staple = fuse_images(images_to_fuse, 'staple') # class_list is not needed for staple/itkvoting
sitk.WriteImage(fused_staple, '/path/to/output_staple.nii.gz')
fused_simple = fuse_images(images_to_fuse, 'simple', class_list=[0,1,2,4])
sitk.WriteImage(fused_simple, '/path/to/output_simple.nii.gz')

Testing

This repo has continuous integration enbabled via Azure DevOps for the following operating systems:

Windows
Ubuntu
macOS

And for the following python versions:

kchsunny/Fusion_model_LabelFusion