In medical and biological research, measurements of cellular dynamics have been paid much attention in recent years. Among such quantifications, particle tracking velocimetry (PTV) of cells is one of the major tools to elucidate the dynamics. For this purpose, it is critical to establish precise detection of individual cell positions at different timepoints. However, in live imaging of confluent cell system, cells are densely packed and touching images of nuclei prevent naive PTV analysis. In this work, we focus on precise detection of cell nucleus positions particularly in a very confluent situation and construct the detection algorithm with deep neural network (DNN). We have tested our system in the case of fly pupal wing epithelium and found the centers of the spots within nuclei with high probability. An implementation of this spot detection to PTV and an extension of DNN to recurrent neural network model (RNN) will also be discussed.
CITATION STYLE
Ishimoto, Y., & Watanabe, T. (2020). Towards Particle Tracking Velocimetry of Cell Flow in Developing Tissue Using Deep Neural Network. In Lecture Notes in Computational Vision and Biomechanics (Vol. 36, pp. 495–504). Springer. https://doi.org/10.1007/978-3-030-43195-2_40
Mendeley helps you to discover research relevant for your work.