Cell Detection And Tracking Research For Immune Activation

Cellular immunity is an important part of the human defense system.T cell is a key part of cellular immunity.Specific binding of T cell to APC cell activates T cell,which produces cellular immunity against foreign antigens.The current way to study T cell activation is to manually label the synaptic positions and synaptic points of T cell in a fluorescent image,establish a model,and analyze the activation process of T cell.The method of manually labeling T cell synaptic positions and synaptic points is cumbersome and complicated,wastes financial resources,and the error rate is high,so a study of auto-labeling T cell synaptic points and synaptic positions by computer is necessary.This article is to complete the automatic labeling of T cell synaptic positions and synaptic points by means of computer vision.The main research contents include:(1)In order to determine the cells of interest,a method of first positioning and then segmentation was used in the T cell image,a binary map of T cell was obtained,and then positions of APC cell were obtained in the fluorescence image by Hough circle detection.Finally,through the test of the cell image,the accuracy and stability of the segmentation algorithm and detection algorithm for T cell segmentation and APC cell detection.(2)In order to obtain the T cell immune synaptic points of interest at a continuous time,First,fit the segmented T cells to an ellipse,and then the APC circle is used to obtain the initial synaptic points,Secondly,according to the initial synaptic points,the T cell and the APC cell of interest are respectively tracked to obtain the continuous time immune synaptic points,and the synaptic points is adjusted by an optimization algorithm.Finally,the accuracy of the algorithm is verified by comparing with the real(artificial marking)synaptic points.(3)In order to obtain the T cell immune synaptic position in continuous time,the deep learning algorithm Faster R-CNN was use d to detect the T cell immune synaptic position,and the effectiveness of the algorithm was verified by experiments.At the same time,T cell immune synaptic position detection software was fabricated by QT and algorithm fusion.