Following the Changes: HeLa cells Lineage from Phase Contrast Microscopy Time-Lapse Data

Phase contrast time-lapse imaging is crucial in biology for long-term, living cells investigations. In particular, time-lapse data may be used to extract the cell lineage, i.e. the set of mother-daughters relationships among proliferating cells. Lineage information are essential when investigating on how phenotipic traits are passed to the progeny along different cell generations. However, the automatic processing of time-lapse images is a very challenging task. This is due both to the huge amount of data usually produced, and to the nature of the imaged targets: cells may move, entering and exiting the field-of-view; they live, proliferate and die, continously changing the density of the population.; they usally show very complex topologies, because they may overlap and modify their shape when in close contact with each other. In this work we present the results obtained with a computational workflow, specifically designed to extract cell lineages from phase contrast time-lapse movies of proliferating HeLa cells. We describe in detail the issues we faced and the solutions we found, starting from the raw movies and addressing the registration, segmentation, and tracking steps that are essential for a reliable lineage construction. To our best knowledge there are no publicly available annotated datasets containing segmentation, tracking and lineage data extracted from phase-contrast time-lapse microscopy movies of dividing HeLa cells. Hence, we built our own benchmark data in a semi-automatic way using the state-of-the-art ICY tool. The presented pipeline has been tested on these annotated datasets to assess its performance and reliability.