DuAT

Feilong Tang, Qiming Huang, Jinfeng Wang, Xianxu Hou, Jionglong Su, and Jingxin Liu

This repo is the official implementation of "DuAT: Dual-Aggregation Transformer Network for Medical Image Segmentation".

1. Introduction

DuAT is initially described in PRCV.

Transformer-based models have been widely demon- strated to be successful in computer vision tasks by mod- elling long-range dependencies and capturing global rep- resentations. However, they are often dominated by fea- tures of large patterns leading to the loss of local details (e.g., boundaries and small objects), which are critical in medical image segmentation. To alleviate this problem, we propose a Dual-Aggregation Transformer Network called DuAT, which is characterized by two innovative designs, namely, the Global-to-Local Spatial Aggregation (GLSA) and Selective Boundary Aggregation (SBA) modules. The GLSA has the ability to aggregate and represent both global and local spatial features, which are beneficial for locat- ing large and small objects, respectively. The SBA mod- ule is used to aggregate the boundary characteristic from low-level features and semantic information from high-level features for better preserving boundary details and locat- ing the re-calibration objects. Extensive experiments in six benchmark datasets demonstrate that our proposed model outperforms state-of-the-art methods in the segmentation of skin lesion images, and polyps in colonoscopy images. In addition, our approach is more robust than existing meth- ods in various challenging situations such as small object segmentation and ambiguous object boundaries.

2. Framework Overview

3. Results

3.1 Image-level Polyp Segmentation

The polyp Segmentation prediction results in here.

4. Usage:

4.1 Recommended environment:

Python 3.8
Pytorch 1.7.1
torchvision 0.8.2

4.2 Data preparation:

Downloading training and testing datasets and move them into ./dataset/, which can be found in this Google Drive/Baidu Drive [code:dr1h].

4.3 Pretrained model:

You should download the pretrained model from Google Drive/Baidu Drive [code:w4vk], and then put it in the './pretrained_pth' folder for initialization.

4.4 Training:

Clone the repository:

git clone https://github.com/Barrett-python/DuAT.git
cd DuAT
bash train.sh

4.5 Testing:

cd DuAT
bash test.sh

4.6 Evaluating your trained model:

Matlab: Please refer to the work of MICCAI2020 (link).

Python: Please refer to the work of ACMMM2021 (link).

Please note that we use the Matlab version to evaluate in our paper.

4.7 Well trained model:

You could download the trained model from Google Drive and put the model in directory './model_pth'.

Citation If you find this code or idea useful, please cite our work:

Citation:

@inproceedings{tang2023duat,
  title={DuAT: Dual-aggregation transformer network for medical image segmentation},
  author={Tang, Feilong and Xu, Zhongxing and Huang, Qiming and Wang, Jinfeng and Hou, Xianxu and Su, Jionglong and Liu, Jingxin},
  booktitle={Chinese Conference on Pattern Recognition and Computer Vision (PRCV)},
  pages={343--356},
  year={2023},
  organization={Springer}
}

6. Acknowledgement

We are very grateful for these excellent works PraNet, Polyp-PVT and SSformer, which have provided the basis for our framework.