Tracking of Leukocytes in Time-Lapse Microscopy
Recent advances in time-lapse imaging have given rise to Two Photon Intra-Vital Microscopy (2P-IVM), which allows the visualization of cell interactions in vivo and opens up new possibilities of investigating the behavior of the immune system through the analysis of cell movement in organs and tissues.
However, the tracking of the movements of leukocytes (the key effectors of the immune system) in imaging data is far from automatic. Firstly, the identification of their shape is difficult, since cells cluster and boundaries are fuzzy. Secondly, cells may have different motion patterns. Lastly, the sampling rate is usually kept to a minimum in order to avoid photo-toxicity and to reduce the size of acquisition files.For these reasons, the currently available software and methods require manual correction, which is time-consuming and introduces bias.
We propose to develop new software and methods for the automatic analysis of leukocyte motion, to improve the tracking accuracy while keeping the execution feasible on commercial workstations.The proposed software and methods exploit space-time connected features in an inverse problem framework, by detecting forward models of leukocyte motion in 2P-IVM data. This will result in intrinsic segmentation and tracking.
Moreover, we set up an online leukocyte-tracking database at www.ltdb.info to collect the extensive amount of imaging data that is necessary to validate the new algorithms, and to predict accurate leukocyte motion without a complete understanding of the properties of their movements.
Mot-clé: Cell tracking, movements, leukocytes, immunity, imaging, inverse problemsretourner