Multicellular multi-parameter joint estimation and accurate tracking system based on ant colony system

Abstract

The invention provides a multicellular multi-parameter joint estimation and accurate tracking system based on an ant colony system, and belongs to the field of cell tracking, wherein parameters include positions, speed, outlines and the like. Initial position distribution of an ant colony is generated due to the fact that a corresponding local area gray level variance of each gray image is defined. Two information pixel fields which work in parallel and are mutually independent are built on the basis, wherein the pixel fields comprise a position field and an outline field. A bounded ant colony decision making system is respectively built on a cell position estimation module and a cell outline estimation module. Finally, important parameters such as positions, speed and outlines of cells are precisely estimated. Therefore, accurate tracking of multiple cells is achieved. By means of mutual cooperation of the bounded heuristic type ant colony systems, a cell detection module is needless, a large number of cell training samples are needless, and the joint estimation and tracking problems of multicellular kinetic parameters and multicellular outlines in a low-contrast cell image sequence are solved.

Description

technical field [0001] The invention provides a multi-cell multi-parameter (position, speed, contour, etc.) joint estimation and precise tracking system based on the ant colony system, which belongs to the field of cell tracking. Background technique [0002] In the past few decades, the rapid development of biological imaging technology has provided a solid technical guarantee for human health. For example, fluorescence microscopy imaging technology has made it possible to study the dynamic behavior of cells. The structure in the cell is marked and tracked, and information such as the speed, acceleration and strength changes of the cell is obtained in the formed cell "life" history record, which is helpful for the research of cell migration and other related cell biology. For example, the neural crest cells of vertebrates will continue to migrate from the dorsal side to the ventral side during the embryonic period. If there is a mutation, the patient's body color, blood cel...

AI Technical Summary

Problems solved by technology

[0005] 2) Due to the trembling caused by the breathing of the organism, the change of the contrast caused by the cells entering or leaving the confocal plane, the quality of the acquired image is reduced, and the difficulty of cell tracking is increased.

[0006] 3) The number of cells changes with time, and there will be phenomena such as approaching or crossing during movement;

[0009] In recent years, compared with the deterministic cell tracking algorithm, the probabilistic (random) multi-cell tracking technology has developed rapidly. There are two typical technologies. One is that Dr. SMAL proposed a particle filter cell based on the detection link. Tracking technology, although the tracking accuracy is high, but many parameters are used, the applicability is general, and the generalization is poor

The second is the multi-cell tracking technology based on random finite sets and prior to measurement proposed by Professor REZA. This technology often has missed detection and false alarms of cells, and the accuracy and stability of tracking are insufficient. The performance of cell tracking also depends on Due to a large number of cell training samples, the differences in the dynamic characteristics of multiple cells and their tracking performance have not been studied, etc.

Images