3DeeCellTracker

3DeeCellTracker is a deep-learning based pipeline for tracking cells in 3D time lapse images of deforming/moving organs (eLife, 2021).

To do:

In the next version of our PyPI package 3DeeCellTracker v0.4.1, following issues will be fixed:

• Solve the bug "the directions of arrows showing the accurate correction are opposite"
  • The codes here have been fixed
• Add functions for extracting/drawing activities from cell images based on tracked labels
• Add the function for storing coordinates of tracked cells

Other critical issues to be solved:

• Currently, 3DeeCellTracker only works with TensorFlow-gpu 1.x which does not support some latest NVIDIA GPU models using CUDA 11 drivers. This issue will be sovled in future versions of 3DeeCellTracker.

Updates:

2021.06.17

We have updated our program to 3DeeCellTracker v0.4.0:

• We modified the notebooks, and the underlying package to simplify the use.
• The intermediate results of segmentation and tracking can be visualized easily to assist the parameter optimization.

2021.03.29

We have updated our program to 3DeeCellTracker v0.3:

• By using vectorization, we remarkably reduced the runtime for tracking cells.

Installation

• Create a conda environment for a PC with GPU including prerequisite packages using the 3DCT.yml file:

$ conda env create -f 3DCT.yml

• (NOT RECOMMEND) Users can create a conda environment for a PC with only CPU, but it will be slow and may fail.

$ conda env create -f 3DCT-CPU.yml

• Install the 3DeeCellTracker package solely by pip

$ pip install 3DeeCellTracker

For detailed instructions, see here.

Quick Start

To learn how to track cells use 3DeeCellTracker, see following notebooks for examples:

1. Track cells in deforming organs:

   • Single mode (clear notebook);
   • single mode (results)

2. Track cells in freely moving animals:

   • Ensemble mode (clear notebook)
   • Ensemble mode (results)

3. Train a new 3D U-Net for segmenting cells in new optical conditions:

   • Train 3D U-Net (clear notebook).
   • Train 3D U-Net (results).

The data and model files for demonstrating above notebooks can be downloaded here.

Note: Codes above were based on the latest version. For old programs used in eLife 2021, please check the "Deprecated_programs" folder.

Video Tutorials

We have made tutorials explaining how to use our software. See links below (videos in Youtube):

Tutorial 1: Install 3DeeCellTracker and train the 3D U-Net

Tutorial 2: Tracking cells by 3DeeCellTracker

Tutorial 3: Annotate cells for training 3D U-Net

Tutorial 4: Manually correct the cell segmentation

A Text Tutorial

LINK: https://bio-protocol.org/prep1350

How it works

We designed this pipeline for segmenting and tracking cells in 3D + T images in deforming organs. The methods have been explained in Wen et al. bioRxiv 2018 and in Wen et al. eLife, 2021.

Overall procedures of our method (Wen et al. eLife, 2021–Figure 1)

Examples of tracking results (Wen et al. eLife, 2021–Videos)

Neurons in a ‘straightened’ freely moving worm	Cardiac cells in a zebrafish larva	Cells in a 3D tumor spheriod

Citation

If you used this package in your research and is interested in citing it here's how you do it:

@article{
author = {Wen, Chentao and Miura, Takuya and Voleti, Venkatakaushik and Yamaguchi, Kazushi and Tsutsumi, Motosuke and Yamamoto, Kei and Otomo, Kohei and Fujie, Yukako and Teramoto, Takayuki and Ishihara, Takeshi and Aoki, Kazuhiro and Nemoto, Tomomi and Hillman, Elizabeth MC and Kimura, Koutarou D},
doi = {10.7554/eLife.59187},
journal = {eLife},
month = {mar},
title = {{3DeeCellTracker, a deep learning-based pipeline for segmenting and tracking cells in 3D time lapse images}},
volume = {10},
year = {2021}
}

Acknowledgements

We wish to thank JetBrains for supporting this project with free open source Pycharm license.