|Microscopy Home | YGMBG Graz Home | IMB Graz Home|
Automated registration of yeast cells in dense cell populations
A new tool (CellSerpent) for registration of yeast cells in conventional transmission images was designed.
The algorithm was developed by Kristian Bredies, Institute of Mathematics and Scientific Computing, University of Graz.
Kristan Bredies, Heimo Wolinski. 2012. An active-contour based algorithm for the automated segmentation of dense yeast populations on transmission microscopy images. Computing and Visualization in Science. In press.
The method overcomes known limitations of detection of yeast cells in transmission images such as:
- very dense cell populations
- variations in image quality
- optical dense subcellular structures (e.g. large vacuoles, lipid bodies)
- variations of the cell morphology
The software tool runs on Matlab platforms and allows batch processing of image data acquired in high-content screens.
CellSerpent can be downloaded here:
Matlab + Image Processing toolbox required; run the 'main.m' file in Matlab.
Please note: the tool was designed to process images of eliptical S. cerevisiae cells as acquired with our high-content screening method (large number of densily associated cells; specific zoom factor); the software is not useful for detection of elongated cells (e.g. for Schizosaccharomyces pombe cells). Furthermore, the parameters can be/have to be modified in order to process yeast images which deviate from those generated in our screening approach.
For details about the algorithm please contact Kristian Bredies, Institute of Mathematics and Scientific Computing, University of Graz.
|UNI-NEWS SITEMAP SEARCH UNI GRAZ Webmaster | 07.09.2017|