search

pengsl-lab / DHUnet

15 stars 1 forks source link

readme

DHUnet

The codes for the paper "DHUnet: Dual-branch Hierarchical Global-local Fusion Network for Whole Slide Image Segmentation"

Abstract

DHUnet Hematoxylin and eosin stained whole slide images (WSIs) are the gold standard for pathologists and medical professionals for tumor diagnosis, surgery planning, and postoperative examinations. In recent years, due to the rapidly emerging field of deep learning, there have been many convolutional neural networks (CNNs) and Transformer based models applied to computational pathology for accurate segmentation. However, the generalization ability and robustness of models often affect the diagnosis and prognosis of cancer, and we attempt to effectively combine the advantage of CNN which excels in sparse WSI segmentation while Transformer excels in dense cases. In this paper, We propose a novel feature fusion strategy, DHUnet, which utilizes Swin Transformer and ConvNeXt modules with a dual-branch hierarchical U-shaped architecture to fuse global and local features for WSI segmentation. Firstly, a WSI is divided into small patches, which are put into the global and local encoders to generate hierarchical features in parallel. Then, with the help of global-local fusion modules and skip connections, the decoder can fully obtain the global coarse and local fine-grained information during the upsampling process. The proposed Cross-scale Expand Layer can make the patches with the same center but different scales recover the input resolution gradually at each stage. Finally, all the projected pixel-level patch masks are merged to restore the final WSI tumor segmentation. Extensive experiments demonstrate that DHUnet has excellent performance and generalization ability for WSI segmentation, achieving the best segmentation results on three datasets with different types of cancer, densities, and target sizes.

Requirements

pip install -r requirements.txt

Pre-trained Checkpoints

The official pre-trained checkpoints are available at Google Drive.

Dataset

The processed datasets and trained models are publicly available at Google Drive.

Liver

Dataset Organization

You should orginize your own dataset like below:

data
│───[Dataset Name 1]
│   └───WSI
│       │   wsi_1.svs / wsi_1.png
│       │   wsi_1.xml / gt_wsi_1.png
│       │   wsi_2.svs / wsi_2.png
│       │   wsi_2.xml / gt_wsi_2.png
│       │           ...
│       │   wsi_n.svs / wsi_n.png
│       │   wsi_n.xml / gt_wsi_n.png
│───[Dataset Name 2]
│   └───WSI
│       │   wsi_1.svs / wsi_1.png
│       │   wsi_1.xml / gt_wsi_1.png
│       │   wsi_2.svs / wsi_2.png
│       │   wsi_2.xml / gt_wsi_2.png
│       │           ...
│       │   wsi_n.svs / wsi_n.png
│       │   wsi_n.xml / gt_wsi_n.png
└───read_xml.py # convert *.xml file to *.png

Dataset Preprocessing and Spliting

python wsi2patch.py --dataset Liver --data_dir data/Liver/WSI --output data/Liver --lists_dir lists/lists_Liver/ --overlap 224 --patch_size 448

Train

python train.py --dataset Liver --network DHUnet --cfg configs/DHUnet_224.yaml --root_path data --max_epochs 50 --output_dir model-Liver/DHUnet --img_size 224 --base_lr 0.005 --batch_size 24

Test

python test.py --dataset Liver --network DHUnet --cfg configs/DHUnet_224.yaml --is_saveni --volume_path data --output_dir model-Liver/DHUnet --max_epochs 50 --base_lr 0.005 --img_size 224 --batch_size 24

Optional implementation networks

CNNs-based network Transformer-based network
UNet Medical Transformer
ResNet TransUnet
FCN SwinUnet
DeepLabV3 TransFuse
ConvNeXt -

References

If you have any questions or comments, please feel free to contact: lianwang@hnu.edu.cn