• Citer

Cell division characterization by automatic cell tracking and cell lineage reconstruction

par Debeir, Olivier ;Kiss, Robert ;Decaestecker, Christine
Editeur scientifique Van Roy, Frans;Geelen, Danny
Référence Belgium Society for Cell and Developmental Biology Autumn Meeting 2006(2006: Ghent), Proceedings of the Belgium Society for Cell and Developmental Biology Autumn Meeting, Cytoskeleton and Cellular Imaging, page (14)
Publication Publié, 2006-10-01

Abstract de conférence

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

In vitro cell observation is widely used by biologists since years. Nowadays the use of computer assisted-microscopy (or time-laps microscopy) allows the management of huge amounts of image data generated during experiments of various durations. Periods of time from several hours to 1 day are generally sufficient to analyze cell trajectories in cell migration and chemotaxis studies. If cell cultures are observed during longer periods of time (3-4 days), it is possible to detect less frequent cell events, such as cell division or cell death. A wide range of applications can be covered by theses time-laps microscopy approaches, including embryology studies and anti-cancer drug testing.We showed in previous works that unmarked cells observed in vitro under classical phase-contrast microscopy during several days can be efficiently and automatically tracked, and their individual trajectories reconstituted and characterized by various measurements. We are now extending this technique for cell division detection using backward cell tracking (from the last to the 1st frame of the image sequence) and cell trajectory recombinations. In the present study, we describe a semi-automatic extension which is able to reconstruct tree structures characterizing cell lineages. The use of backward cell tracking enables easy cell division detection (by identifying cell trajectory merging), except in some cases when the tracking procedure fails. We thus added a “light” interactive step in order to select the actual cell division events among all the automatically detected ones; the operator has only to accept or to reject each proposed event.In order to characterize the action of specific drugs on cell motility and also on cell behavior during and just after cell division, specific features characterizing cell division events are extracted (in addition to those concerning cell movements). We illustrated our methodology on different examples including drug actions on cytokinesis, i.e. the cell separation event occurring at the end of cell division.