Color-Based Segmentation vs. Stereology: A Simple Comparison Between Two Semi-Automated Methods of Image Analysis for the Quantification of Collagen
Paulo Salinas; Jorge Sanhueza & Carlos Sandoval
Image processing techniques are being widely developed for helping specialists in analysis of histological images and its application is especially useful in obtaining numerical data for the realization of the subsequent statistical analysis. The use of these methods makes the histological analysis of experts more objective and less time-consuming. In this paper we evaluate how well the quantitative methods - color-based image segmentation and stereology - agree on average, and how well they agree for the individuals when they are used to quantify type I and III collagen. Digital images of sections of Salmo salar jaws (5 μm, SiriusRed staining) were analyzed. Collagen quantification was performed by two methods in the same group of images: i) Color Based-Segmentation (K-means algorithm; pixel cluster; ImageJ32 v1.51p) and ii) Stereology (V ; M ; STEPanizer Stereological Tools). They were evaluated 200 V 36 images per group. The difference between groups and concordance was analyzed using t-Student (p<0.05) and Blant Altman Comparison Method, respectively. The data analysis of average and individual assessments showed that there is concordance between two methods. In conclusion, stereology and color-based image segmentation are powerful tools which quantify collagen in histological sections.
SALINAS, P.; SANHUEZA, J. & SANDOVAL, C. Color-based segmentation vs. stereology: a simple comparison between two semi- automated methods of image analysis for the quantification of collagen. Int. J. Morphol., 36(3):1118-1123, 2018.