CATS:COMPLEMENTARY CNN AND TRANSFORMER ENCODERS FOR SEGMENTATION

@inproceedings{li2022cats,
  title={Cats: Complementary CNN and Transformer Encoders for Segmentation},
  author={Li, Hao and Hu, Dewei and Liu, Han and Wang, Jiacheng and Oguz, Ipek},
  booktitle={2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI)},
  pages={1--5},
  year={2022},
  organization={IEEE}
}

CATS v2: Hybrid encoders for robust medical segmentation

@article{li2023cats,
  title={CATS v2: Hybrid encoders for robust medical segmentation},
  author={Li, Hao and Liu, Han and Hu, Dewei and Yao, Xing and Wang, Jiacheng and Oguz, Ipek},
  journal={arXiv preprint arXiv:2308.06377},
  year={2023}
}

train.py ------> train BTCV dataset with .jason file. (based on https://github.com/Project-MONAI/tutorials/blob/main/3d_segmentation/unetr_btcv_segmentation_3d.ipynb)

train_with_data_dir.py -----> changed data I/O, you can use a path that contains nifti files, in the following format

dataset_folder
└─train_set_image_folder
 ├── 1.nii.gz
 └── 2.nii.gz
 ...
 ...
 ...
 └── 100.nii.gz
└─train_set_label_folder
 ├── 1_label.nii.gz
 └── 2_label.nii.gz
 ...
 ...
 ...
 └── 100_label.nii.gz