Description
Date depot: 10 mai 2022
Titre: Leveraging Deep-Learning for the robust tracking of single particles in biological imaging. Application to the automatic, long-term tracking of the calcium activity of individual neurons in moving animals
Directeur de thèse:
Jean-Christophe OLIVO-MARIN (Analyse d'images biologiques)
Directrice de thèse:
Elsa ANGELINI (LTCI (EDMH))
Directeur de thèse:
Alasdair NEWSON (ISIR (EDITE))
Directeur de thèse:
Thibault LAGACHE (Analyse d'images biologiques)
Domaine scientifique: Sciences et technologies de l'information et de la communication
Thématique CNRS : Images et vision
Resumé: The objective of this PhD thesis is the development of innovative single-particletracking (SPT) methods in biological imaging. The developed methods will leverage deep-learning (DL) algorithms to increase the robustness of the tracking over long time-lapse sequences. Indeed, DL is already in use for the detection of biological particles, or the tracking itself but a DL-aided tracking framework that would concomitantly detect and track the particles using their appearance features and expected dynamics is missing. The development of such deeplearning-aided SPT algorithms is motivated by a collaboration established with a neuroscience lab at Columbia University that genetically engineers small animals (Hydra) and images the activity of single neurons during animal’s behaviour with fluorescent calcium probes. In that context, the automatic, long-term tracking of the calcium activity of individual neurons is a fundamental pre-requisite to the robust analysis of their functional connectivity and the emergent computational properties that drive the animal behaviour.
Doctorant.e: Reme Raphael