deep_learning_lungs_segmentation

Pre-requisites and installations

• Make sure you have python3 installed

• Clone this repository on your machine and go in it:

  cd deep_learning_lungs_segmentation/

• Create a virtual environments

  python3 -m venv env_seg

• Activate the virtual environment

  source env_seg/bin/activate

• Update pip3 repository and install dependencies listed in the requirements.txt

  pip3 install --upgrade pip
python3 -m pip install -r requirements.txt

Use our trained model on your data

To predict lungs mask (on your image input_data.mhd) with the model (located at checkpoint_best.pt), in the working directory, run the following command:

python3 trained_model_on_your_data.py -i input_data.mhd -o output_data.mhd -w checkpoint_best.pt

All paths are realtive to the working directory. Predicted lungs mask as .mhd and .raw files will be located in : results_showcase/

Acknowledments

Thanks to the authors of this repository : https://github.com/milesial/Pytorch-UNet for providing an efficient implementation of U-net.

Thanks to Olivier Bernard for getting us started with the project by providing examples of his codes.