search

jcwang123 / Separate_CL

[AAAI 2022 Oral] Separate Contrastive Learning for Organs-at-Risk and Gross-Tumor-Volume Segmentation with Limited Annotation
15 stars 3 forks source link
aaai2022 contrastive-learning segmentation

readme

Separate Contrastive Learning for Organs-at-Risk and Gross-Tumor-Volume Segmentation with Limited Annotation

Introduction

This is an official release of the paper Separate Contrastive Learning for Organs-at-Risk and Gross-Tumor-Volume Segmentation with Limited Annotation.

It is accepted by AAAI-2022 Oral and has been awarded an AAAI student scholarship.

Separate Contrastive Learning for Organs-at-Risk and Gross-Tumor-Volume Segmentation with Limited Annotation,
Jiacheng Wang, Xiaomeng Li, Yiming Han, Jing Qin, Liansheng Wang, Zhou Qichao
In: Association for the Advancement of Artificial Intelligence (AAAI), 2022
[arXiv][Bibetex]

<div align="center" border=>

TODO List

  1. Complete the resources ...

  2. Evaluate the effectiveness on more vision tasks ...

Code List

Usage

  1. First, you can download the dataset at PDDCA. To preprocess the dataset and save as ".png", run:

    $ python utils/prepare_data.py

    Note that some cases lack the complete annotation, so that we can obtain 32 cases with full annotation in the end.

  2. To create the region set, alternatively run:

    $ python utils/prepare_segs.py --dataset pddca --filter_method all --seg_method fb --min_size 400
$ python utils/prepare_segs.py --dataset pddca --filter_method all --seg_method slic --n_segments 32
$ python utils/prepare_segs.py --dataset pddca --filter_method all --seg_method slice --n_segments 32

Citation

If you find SepaReg useful in your research, please consider citing:

@inproceedings{wang2022separated,
  title={Separated Contrastive Learning for Organ-at-Risk and Gross-Tumor-Volume Segmentation with Limited Annotation},
  author={Wang, Jiacheng and Li, Xiaomeng and Han, Yiming and Qin, Jing and Wang, Liansheng and Qichao, Zhou},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={36},
  number={3},
  pages={2459--2467},
  year={2022}
}