title: Tracking-by-Assignment as a Probabilistic Graphical Model with Applications in Developmental Biology creator: Kausler, Bernhard subject: ddc-000 subject: 000 Generalities, Science subject: ddc-004 subject: 004 Data processing Computer science subject: ddc-500 subject: 500 Natural sciences and mathematics subject: ddc-570 subject: 570 Life sciences description: This thesis presents a novel approach for tracking a varying number of divisible objects with similar appearance in the presence of a non-negligible number of false positive detections (more than 10%). It is applied to the reconstruction of cell lineages in developing zebrafish and fruit fly embryos from 3d time-lapse record- ings. The model takes the form of a chain graph—a mixed directed-undirected probabilistic graphical model—and a tracking is obtained simultaneously over all time slices from the maximum a-posteriori configuration. The tracking model is used as the second step in a two-step pipeline to produce digital embryos—maps of cell nuclei in an embryo and their ancestral fate; the first step being the segmentation of the fluorescently-stained cell nuclei in light sheet microscopy images. The pipeline is implemented as a software with an intuitive graphical user interface. It is the first freely available program of its kind and makes the presented methods accessible to a broad audience of users from the life sciences. date: 2013 type: Dissertation type: info:eu-repo/semantics/doctoralThesis type: NonPeerReviewed format: application/pdf identifier: https://archiv.ub.uni-heidelberg.de/volltextserverhttps://archiv.ub.uni-heidelberg.de/volltextserver/15296/1/bkausler_13_tracking.pdf identifier: DOI:10.11588/heidok.00015296 identifier: urn:nbn:de:bsz:16-heidok-152962 identifier: Kausler, Bernhard (2013) Tracking-by-Assignment as a Probabilistic Graphical Model with Applications in Developmental Biology. [Dissertation] relation: https://archiv.ub.uni-heidelberg.de/volltextserver/15296/ rights: info:eu-repo/semantics/openAccess rights: Please see front page of the work (Sorry, Dublin Core plugin does not recognise license id) language: eng