Cell tracking method based on correction by cellular partition combined with multi-frames and original image feedback

Abstract

The invention provides a cell tracking method based on correction by cellular partition combined with multi-frames and original image feedback. The method comprises the following steps of: segmenting cell sequence images, tracking inert cells by a regional overlap method, carrying out cellular partition for each frame of image, and building a cellular information storage center; applying a topological constraint algorithm introducing a Delaunay triangulation network combined with a cellular region to match the cells, storing a cell matching result into a temporary-storage track, then identifying the types of the cells in the temporary-storage track, and correcting the classified cells by a multi-frames and original images feedback method; and finally, updating the cellular information storage center and the storage information of the inner parts of the cells, and outputting a result. The method uses the multi-frames feedback and original images feedback method to correct the matching result and segmentation faults, and therefore the accuracy rate of a cell tracking system is further increased and meanwhile the higher efficiency can be realized.

Description

technical field [0001] The invention relates to a cell movement analysis method, in particular to a cell tracking method which uses topology constraints combined with cellular area division, and uses multi-frame and original image feedback to correct the results. Background technique [0002] Cell movement analysis is usually to track target cells to obtain information such as cell movement speed, displacement, trajectory, etc., which is crucial in the research of cell behavior, drugs and diseases. When we track a target cell in detail and accurately, especially when tracking a large number of cells, we must try to avoid the interference of other cells. This is the difficulty of cell tracking, and it is also a research hotspot in cell tracking in recent years. [0003] At present, more mature advanced cell tracking algorithms include level set algorithm (level set), active contour method (active contour), mean shift method (Mean Shift) and topological constraint method, etc....

AI Technical Summary

Problems solved by technology

However, this tracking method is not capable of processing image regions containing sparse cells and special cells, and is highly dependent on the segmentation effect of cell images.

Li (Kang Li, Mei Chen, Takeo Kanade, Cell population tracking and lineage construction with spatiotemporal context. in Proc. 10th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2007: 295302.) et al proposed multi-module Tracking system, this method integrates the cell information of multiple modules, and judges based on the advantages of each module, avoiding the tracking failure caused by a single method, but this method cannot correct segmentation errors, and only uses the information of each module between adjacent frames Fusion Corrects Tracking Errors

Images