SwinCheX

This repo is the official implementation of "SwinCheX: Multi-label classification on chest X-ray images with transformer". (Link will be added soon)

Training on NIH (ChestX-ray14) dataset

• You can find initial setup and pretrain models in get_started.md. We used ImageNet-22K pre-trained model Swin-L with 224x224 resolution for our training.

• If you had problems installing Apex, you can install it using conda-forge:

  conda install -c conda-forge nvidia-apex
  

  Additionally, install numpy, pillow, pandas, scikit-learn and scipy packages with pip or conda:

  pip install [package-name]
  

• Download NIH (ChestX-ray14) dataset from Kaggle. Merge images from different folders into one folder. Optionally, you can have different folders for train, validation, and test data.

• For training the model on one gpu run:

  python -m torch.distributed.launch --nproc_per_node 1 --master_port 12345 main.py \
  --cfg configs/swin_large_patch4_window7_224.yaml --resume path/to/pretrain/swin_large_patch4_window7_224_22k.pth \
  --trainset path/to/train_data/ --validset path/to/validation_data/ --testset path/to/test_data/ \
  --train_csv_path configs/NIH/train.csv --valid_csv_path configs/NIH/validation.csv --test_csv_path configs/NIH/test.csv \
  --batch-size 32 [--output <output-directory> --tag <job-tag> --num_mlp_heads 3] > log.txt
  

  You can extract validation and test ROC_AUC scores from resulting log.txt file.

  Note: As mentioned above, if you have all data in one folder, then the trainset, validset and testset arguments would point to the same folder.

  Note: If you want to continue a half-trained model from a checkpoint, you should comment the line specified with "TODO" in utils.py.