Bacterial growth can be studied at the single cell level through time-lapse microscopy imaging. Technical advances in microscopy lead to increasing image quality, which in turn allows to visualize larger areas of growth, containing more and more cells. In this context, the use of automated computational tools becomes essential. In this paper, we present STrack, a tool that allows to track cells in time-lapse images in a fast and efficient way. We compared it to 3 recently published tracking tools on images ranging over 6 different bacterial strains with various morphologies. STrack showed to be the most consistent tracking tool, returning more than 80% of correct cell lineages on average, in comparison to manually annotated ground-truth. The python implementation of STrack, a docker structure, and a tutorial on how to download and use the tool can be found on the following github page: https://github.com/Helena-todd/STrack. IMPORTANCE Automated image analysis of growing prokaryotic cell populations becomes indispensable with larger data sets, such as derived by time-lapse microscopy. The tracking of the same individual cells and their daughter lineages is cumbersome and prone to errors in image alignment or poor resolution. Here, we present a simplified but highly effective tool for non-specialists to engage in cell tracking. The tool can be downloaded and run as a contained script-structure requiring minimal user input. Run times are fast, in comparison to other equivalent tools, and outputs consist of cell tables that can be subsequently used for lineage analysis, for which we offer examples. By providing open code, training data sets, as well as simplified script execution, we aimed to facilitate wide usage and further tool development for image analysis.
Keywords: algorithm; bioinformatics; cell tracking; image analysis; soil microbiology.