Разработка программного продукта для подсчета клеточных структур гистологических снимков роговицы

Abstract

Manual cell counting is commonly used for the quantitative assessment of cellular structures; however, it is labor-intensive, time-consuming, and prone to fatigue. Most automated cell counting methods are expensive and require expert involvement. The use of image analysis software provides an affordable yet reliable automated cell counting solution, particularly for histological corneal image analysis. This study aims to develop a software product for counting cellular structures in histological corneal images. The program is implemented in Python within the Visual Studio Code environment and features a graphical user interface created with Tkinter. The software allows users to load images, mark different cell types by mouse clicks, and save annotations and statistics in Excel format. Testing was conducted on 50 histological images of Wistar rat corneas stained with hematoxylin and eosin. The software automatically identifies and counts fibroblasts, lymphocytes, macrophages, mast cells, basophils, eosinophils, and neutrophils. Initial testing revealed an issue with displaying resized images on the canvas, which made annotation difficult. Future modifications are planned to address this problem. The developed software successfully provides direct annotation of cellular structures in images and automated cell counting