Cell cycle phase detection with cell deformation analysis

Abstract

Medical experts and pathologists studies on cells to examine the effects on cell activities due to external factors. In order to observe the live cell activity over a long period of time, new imaging techniques such as differential interference contrast microscopy have been deployed to overcome the shortcomings of transparent cells under the conventional microscopy. A normal healthy cell regulates the mitosis process (cell splitting) to prevent over growing. Abnormal cell cycle pattern may provide an important clue to disease or mutation, for example, cancerous cells expand uncontrollably with irregular mitosis process. In order to analyze the cell patterns, it is important to identify different stages in a cell cycle, and such process often requires human interactions. In this paper, proposed an automatic cell cycle phase detection process by analyzing the variations in the shapes and sizes of cells in the life cycle. We first adopted the geodesic active contour segmentation algorithm to extract the cells and by adjusting the initial contour for the GAC process, we can speed up the segmentation process almost half of the original processing time. By analyzing the deformation features such as the cell area and strain of the cell contours deforming over long period of time, we established the behaviors and the states of cells. Our results have demonstrated that the strain analysis of the cell deformation can detect the interphase and mitosis phase in cell cycle, and by analyzing the changes in cell area during mitosis phase, cytokinesis can be identified.