An Image Analysis Algorithm for Kinematics and Morphological Measurement of Cells for the Evaluation of Tissue Construction

Image of cells has been analyzed in the various fields of biomedical engineering, such as evaluation of new drug and research of wound healing process, where the static or transient parameters, such as cell number, position, shape, and area, are usually quantified for small amount of the cells during a limited period. Therefore, the current image analysis techniques are not directly applicable for the analysis of the property of a tissue, which consists of huge number of cells. In this study, aiming at evaluation of tissue construction or reconstruction process, we developed a new image processing algorithm which could analyze the properties of huge number of cells in a tissue. The cultured human peritoneal mesothelial cells were injured mechanically by a scratch and were used as a tissue reconstruction model. The wound healing process in a CO2 incubator was visualized with a phase contrast microscope. The sequential images of the cells were acquired by a CMOS camera at intervals of one minute, and were stored in a personal computer. The stored image was converted into grayscale image with 256 levels. Based on the cell position in the flame immediately before, the region of the interest (ROI) was set for cell tracing, and the target cell was detected by processing the ROI with the Watershed method. Various parameters of the detected cell, such as centroid, area, surroundings length, circularity, and luminance, were quantified. The cell tracing algorithm was coded by the macro language of the image processing software. Through the preliminary experiments, the algorithm developed in this study could trace 88.4 ± 7.9 % of the cells detected in the initial flame.