Fuzzy Sets Applied to Image Processing and Quantification of Interstitial Fibrosis and Glomerular Size in Computer Assisted Microscopy

摘要

The aim of the paper is to present the application of fuzzy sets to image processing and analysis of renal biopsies in light and confocal microscopy. Fuzzy segmentation was applied to processing of grey scale images. Colour images were processed by fuzzy-like approach. Segmented glomerular profiles and interstitial fibrosis were then quantitatively assessed. The mean glomerular volume and the degree of fibrosis were estimated in light microscopy. The volume of glomerular mesangium was assessed with a confocal microscope. Concluding, fuzzy processing makes morphometric studies of renal biopsies more practical for clinical diagnosis of diseases.