Unsupervised Domain Adaptation for Anatomical Structure Detection in Ultrasound Images.

🔨 PostScript

  😄 This project is the pytorch implemention of ToMo-UDA;

  😆 Our experimental platform is configured with One RTX3090 (cuda>=11.0);

  😊 Currently, this code is avaliable for proposed dataset $FUSH^2$, public dataset CardiacUDA and MMWHS;

💻 Installation

1. You need to build the relevant environment first, please refer to : environment.yml

2. Install Environment:

   conda env create -f environment.yml
   

• We recommend you to use Anaconda to establish an independent virtual environment, and python > = 3.8.3;

📘 Data Preparation

1. FUSH^2 dataset

• This project provides the use case of UDA Ultrasound Anatomical Structure Detection task;

• The hyper parameters setting of the dataset can be found in the utils/config.py, where you could do the parameters modification;

• For different tasks, the composition of data sets have significant different, so there is no repetition in this file;

  1. Download & Unzip the dataset.

     The FUSH^2 dataset is composed as: /Heart & /Head.

  2. The source code of loading the $FUSH^2$ dataset exist in path :

     ..\data\fetus_dataset_coco.py
     and modify the dataset path in
     ..\utils/config.py

  3. In utils/config.py, you can set the part to select the anatomical slice, and choose the source and target domains using selected_source_hospital and selected_target_hospital, respectively.

2. FUSH^2 dataset access

• Dataset access can be obtained by contacting hospital staff (doc.liangbc@gmail.com) and asking for a license.

🐾 Training

1. In this framework, after the parameters are configured in the file utils/config.py and train.py , you only need to use the command:

   python train.py

2. You are also able to start distributed training.

   • Note: Please set the number of graphics cards you need and their id in parameter "enable_GPUs_id".

   python -m torch.distributed.launch --nproc_per_node=4 train.py

🐾 Testing

1. Download the TEST_CHECKPOINT here.

2. you only need to use the command:

    python test.py --path TEST_CHECKPOINT

🐾 citation

@inproceedings{puunsupervised,
  title={Unsupervised Domain Adaptation for Anatomical Structure Detection in Ultrasound Images},
  author={Pu, Bin and Lv, Xingguo and Yang, Jiewen and Guannan, He and Dong, Xingbo and Lin, Yiqun and Shengli, Li and Ying, Tan and Fei, Liu and Chen, Ming and others},
  booktitle={Forty-first International Conference on Machine Learning}
}

🚀 Code Reference

• https://github.com/CityU-AIM-Group/SIGMA