Multiple sclerosis cortical and WM lesion segmentation at 3T MRI: a deep learning method based on FLAIR and MP2RAGE
This is the code repository of the Neuroimage: Clinical paper.
This software provide a multiple sclerosis cortical and white matter lesion segmentation. The input required are co-registered FLAIR and MP2RAGE volumes of the subject acquired at 3T. The code depends on the NiftyNet framework and it is based on Tensorflow. A GPU is required to train the network with new data, whereas a LINUX machine is sufficient for inference.
First clone the repository and pull the niftynet code in the submodule as follows
git clone https://github.com/FrancescoLR/MS-lesion-segmentation/
cd MS-lesion-segmentation
git submodule update --init
Optionally, install a Python virtual environment
# CPU compute
python3 -m venv niftynet-cpu --clear
# GPU compute
python3 -m venv niftynet-gpu --clear
Next, activate the environment and install the custom Niftynet and the required dependencies:
source $(pwd)/niftynet-cpu/bin/activate
# for the GPU version, run `source $(pwd)/niftynet-gpu/bin/activate`
python -m pip install -r NiftyNet/requirements-cpu.txt
# for the GPU version, run `python -m pip install -r NiftyNet/requirements-gpu.txt`
python -m pip install -e ./NiftyNet
deactivate
Before any use, the vitual environment needs to be activate
source $(pwd)/niftynet-cpu/bin/activate
# For the GPU version `source $(pwd)/niftynet-gpu/bin/activate`
A configuration file and a trained model are provided in separate folders in the root directory. For inference co-register FLAIR and MP2RAGE images and place them in newly created folders data/FLAIR and data/MP2RAGE, respectively. Then, run:
net_run inference -c path_to/MS-lesion-segmentatin/Configuration file/config_network.ini -a niftynet.application.segmentation_application.SegmentationApplication
The segmentations will be saved to data/inference/. Additional parameters can be set in the configuration file Configuration file/config_network.ini.
Further instructions and commands to train a network on new data are present in the NiftyNet website.
This software is released under the version 2.0 of the Apache License.
If you use this code please cite the following references:
- Gibson, Eli, et al. "NiftyNet: a deep-learning platform for medical imaging." Computer methods and programs in biomedicine 158 (2018): 113-122. doi
- La Rosa, Francesco, et al. "Multiple sclerosis cortical and WM lesion segmentation at 3T MRI: a deep learning method based on FLAIR and MP2RAGE." NeuroImage: Clinical (2020): 102335. doi
- La Rosa, Francesco, Abdulkadir. Ahmed, Thiran, Jean-Philippe, Granziera, Cristina, & Bach Cuadra, Merixtell. (2020, July 7). Software: Multiple sclerosis cortical and WM lesion segmentation at 3T MRI: a deep learning method based on FLAIR and MP2RAGE (Version v1.0). doi