Awesome

TransUNet

This repo holds code for TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation

📰 News

• [7/26/2024] TransUNet, which supports both 2D and 3D data and incorporates a Transformer encoder and decoder, has been featured in the journal Medical Image Analysis (link).

@article{chen2024transunet,
  title={TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers},
  author={Chen, Jieneng and Mei, Jieru and Li, Xianhang and Lu, Yongyi and Yu, Qihang and Wei, Qingyue and Luo, Xiangde and Xie, Yutong and Adeli, Ehsan and Wang, Yan and others},
  journal={Medical Image Analysis},
  pages={103280},
  year={2024},
  publisher={Elsevier}
}

• [10/15/2023] 🔥 3D version of TransUNet is out! Our 3D TransUNet surpasses nn-UNet with 88.11% Dice score on the BTCV dataset and outperforms the top-1 solution in the BraTs 2021 challenge and secure the second place in BraTs 2023 challenge. Please take a look at the code and paper.

Usage

1. Download Google pre-trained ViT models

• Get models in this link: R50-ViT-B_16, ViT-B_16, ViT-L_16...

wget https://storage.googleapis.com/vit_models/imagenet21k/{MODEL_NAME}.npz &&
mkdir ../model/vit_checkpoint/imagenet21k &&
mv {MODEL_NAME}.npz ../model/vit_checkpoint/imagenet21k/{MODEL_NAME}.npz

2. Prepare data (All data are available!)

All data are available so no need to send emails for data. Please use the BTCV preprocessed data and ACDC data.

3. Environment

Please prepare an environment with python=3.7, and then use the command "pip install -r requirements.txt" for the dependencies.

4. Train/Test

• Run the train script on synapse dataset. The batch size can be reduced to 12 or 6 to save memory (please also decrease the base_lr linearly), and both can reach similar performance.

CUDA_VISIBLE_DEVICES=0 python train.py --dataset Synapse --vit_name R50-ViT-B_16

• Run the test script on synapse dataset. It supports testing for both 2D images and 3D volumes.

python test.py --dataset Synapse --vit_name R50-ViT-B_16

Reference

• Google ViT
• ViT-pytorch
• segmentation_models.pytorch

Citations

@article{chen2021transunet,
  title={TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation},
  author={Chen, Jieneng and Lu, Yongyi and Yu, Qihang and Luo, Xiangde and Adeli, Ehsan and Wang, Yan and Lu, Le and Yuille, Alan L., and Zhou, Yuyin},
  journal={arXiv preprint arXiv:2102.04306},
  year={2021}
}