Awesome
SAML & A Multi-site Dataset for Prostate MRI Segmentation
by Quande Liu, Qi Dou, Pheng-Ann Heng.
Introduction
- The Tensorflow implementation for our MICCAI 2020 paper 'Shape-aware Meta-learning for Generalizing Prostate MRI Segmentation to Unseen Domains'.
- A well-organized multi-site dataset (from six data sources) for prostate MRI segmentation, that can support research in various problem settings with need of multi-site data, such as Domain Generalization, Multi-site Learning and Life-long Learning, etc. For more details and downloading link of the dataset, please Find Here.
Setup & Usage for the Code
-
Check dependencies:
python==2.7.17 numpy==1.16.6 scipy==1.2.1 tensorflow-gpu==1.12.0 tensorboard==1.12.2 SimpleITK==1.2.0
-
To train the model, you need to specify the training configurations (can simply use the default setting) in main.py, then run:
python main.py --phase=train
-
To evaluate the model, run:
python main.py --phase=test --restore_model='/path/to/test_model.cpkt'
You will see the output results in the folder
./output/
.
Citation
If this repository is useful for your research, please cite:
@article{liu2020shape,
title={Shape-aware Meta-learning for Generalizing Prostate MRI Segmentation to Unseen Domains},
author={Liu, Quande and Dou, Qi and Heng, Pheng-Ann},
journal={International Conference on Medical Image Computing and Computer Assisted Intervention},
year={2020}
}
Questions
For further question about the code or dataset, please contact 'qdliu@cse.cuhk.edu.hk'