This is the official implementation of CShaper published by Nature Communications. It is developed for segmenting densely distributed cells at single-cell elevel, supporting detailed morphological analysis on developmental cells with live-cell microscopy. CShaper was extensively tested on C. elegans, but it should work for other kinds of membrane images (possibly with re-retraining process).

Usage

Note: This repository targets for runing CShapper with command lines. If you want to process your data with user-friendly GUI, pls prefer to CShaperApp.

Step 1: clone this repository and install requirements with conda:

git clone --depth 1 https://github.com/cao13jf/CShaper.git
cd CShaper
conda env create -f environment.yml

Step 2 (optional): train CShaper. Parameters in configs/train.txt should be adjusted accordingly. You may refer to the example data to organize your own dataset.

python train.py --cf configs/train.txt

Step 3: Test CShaper. Parameters for testing CShaper are listed in configs/test.txt. Pretrained model can be downloaded at Google Drive.

python test.py --cf configs/test.txt

Note: Code for shape analysis is not included in this repository.

Correspondence

Please contact Jianfeng Cao at jianfeng13.cao(at)gmail.com if you have any question about the source code or dataset.

Citation

If this work is useful for you, pls consider the citation.

@article{cao2020establishment,
  title={Establishment of a morphological atlas of the Caenorhabditis elegans embryo using deep-learning-based 4D segmentation},
  author={Cao, Jianfeng and Guan, Guoye and Ho, Vincy Wing Sze and Wong, Ming-Kin and Chan, Lu-Yan and Tang, Chao and Zhao, Zhongying and Yan, Hong},
  journal={Nature communications},
  volume={11},
  number={1},
  pages={6254},
  year={2020},
  publisher={Nature Publishing Group UK London}
}