Answers are grounded in this paper and cite their sources.
Since the dynamics of immune cells change about phenomena in a living body, it is very important to observe and analyze cell dynamics in vivo in real-time. For this purpose, it is necessary to extract the information for the analysis by accurately tracking individual cells. As a method for this, general object tracking algorithms based on CNN (Convolutional Neural Networks) have been actively studied in the field of computer vision. However, in cell tracking, there are a large number of cells in fluorescent images that are similar in color and shape. It is not easy to recognize individual cells once they are lost due to overlap with other cells. Thus it is difficult to generate a large amount of training data with correct tracking trajectories. To cope with the problem of insufficient training data of cell images, our method extends the data by image processing and by assigning pseudo-labels. Furthermore, to obtain information more suitable for dynamic analysis, we propose to apply the re-identification function based on Euclidean distance. We demonstrate the effectiveness of our method with application to time-lapse images of immune cells against multiple inflammatory stimulations.
No takes yet. Share an insight, caveat, or question.
A 2023 study studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: