Fibro-CoSANet: Pulmonary Fibrosis Prognosis Prediction using a Convolutional Self Attention Network
git clone https://github.com/zabir-nabil/osic-pulmonary-fibrosis-progression.git
cd osic-pulmonary-fibrosis-progression
- Install Anaconda Anaconda
conda create -n pulmo python==3.7.5
conda activate pulmo
conda install -c intel mkl_fft
(opt.)
conda install -c intel mkl_random
(opt.)
conda install -c anaconda mkl-service
(opt.)
pip install -r requirements.txt
- Download the kaggle.json from Kaggle account. Kaggle authentication
- Keep the kaggle.json file inside data_download folder.
sudo mkdir /root/.kaggle
sudo cp kaggle.json /root/.kaggle/
sudo apt install unzip
if not installed already
cd data_download; python dataset_download.py; mv osic-pulmonary-fibrosis-progression.zip ../../; unzip ../../osic-pulmonary-fibrosis-progression.zip -d ../../; cd ../; python train_slopes.py
- Set the training hyperparameters in
config.py
- Slope Prediction
- To train slopes model run
python train_slopes.py
- trained model weights and results will be saved inside
hyp.results_dir
- Quantile Regression
- To train qreg model run
python train_qreg.py
- trained model weights and results will be saved inside
hyp.results_dir
https://arxiv.org/abs/2104.05889