3DeeCellTracker, a deep learning-based pipeline for segmenting and tracking cells in 3D time lapse images.

Journal: eLife

Published: 2021-03-30

DOI: 10.7554/elife.59187

Affiliations: 10

Authors: 14

Go to article

Research Highlight

Tracking cells with AI

© Anchalee Phanmaha/Moment/Getty Images

© Anchalee Phanmaha/Moment/Getty Images

Software that automatically tracks the movement of cells imaged by three-dimensional, time-lapse microscopy offers an unprecedented ability to monitor cellular activity in lab organisms and tissue models of disease.

Artificial intelligence techniques known as deep learning are opening up new possibilities in bioimage analysis.

Now, a team led by a researcher from Osaka University has designed one such deep-learning method that can efficiently track the position of individual cells imaged by a variety of optical systems.

They demonstrated the tool’s accuracy by following the fate of up to 150 neurons in the brains of worms. They also tracked around 100 cells in a beating zebrafish heart with high fidelity. Plus, they analysed close to 1,000 cancer cells in a cultured tumour model, a common platform for drug screening.

The cell-tracking software should empower scientists to study cellular dynamics with hitherto unachievable ease and precision.

Supported content

  1. eLife 10, e59187 (2021). doi: 10.7554/eLife.59187
Institutions Share
Osaka University, Japan 0.17
Columbia University in the City of New York (CU), United States of America (USA) 0.14
Kyushu University, Japan 0.14
National Institute for Physiological Sciences (NIPS), NINS, Japan 0.12
Exploratory Research Center on Life and Living Systems (ExCELLS), NINS, Japan 0.11
The Graduate University for Advanced Studies (Sokendai), Japan 0.11
Nagoya City University, Japan 0.10
National Institute for Basic Biology (NIBB), NINS, Japan 0.06
Hokkaido University, Japan 0.04
RIKEN Center for Advanced Intelligence Project (AIP), Japan 0.02