Section: New Software and Platforms

CellStar: Long-term tracking of single cells from brightfield microscopy images

Participants : Grégory Batt, Pascal Hersen, Artémis Llamosi, Szymon Stoma.

In close collaboration with Kirill Batmanov, Cédric Lhoussaine and Cristian Versari from the LIFL (CNRS/Lille Univ), we developed CellStar , a tool-chain for image processing and analysis dedicated to segmentation and tracking of yeast cells in brightfield time-lapse microscopy movies. To estimate algorithm quality we developed a benchmark made of manually-verified images illustrating various situations. On this benchmark, CellStar outperformed 5 other state-of-the-art methods. The tool-chain is implemented in MATLAB and is provided together with the Python Yeast Image Toolkit benchmark tool.