Computerized quantitative color analysis for histological study of pulmonary fibrosis

Color segmentation is a process through which particular structures can be identified and isolated by the same image according to their color. This may prove particularly useful for the quantification of human pulmonary fibrosis through computerized color segmentation methods applied to trichromic stainings. The aim of this study was to develop and verify a reliable method in order to obtain reproducible parameters of fibrosis using trichromic stainings.

The system used is composed of a PC, hardware and software (Image-Pro Plus 4.1-Media Cybernetics, U.S.A.) for image acquisition and analysis, TV camera and optical microscope. The method requires preliminary software procedures of spatial calibration (micron scale) and setting of color segmentation for quantitative color analysis. Massonstained histological specimens of pulmonary parenchyma, with various degree of fibrosis, were analyzed from 10 cases of different lung diseases.

Statistical data of the morphological components identified (space, tissue and fibrosis) support the hypothesis that fibrosis distribution depends on relative amount in the tissue and at the same time attest to the precision of the method.

This procedure could represent a standard method for histopathological evaluation of tissular damage, with fibrosis outcome, especially in connection with the diagnosis and the seriousness of the damage. This would be extremely useful in both the anatomo-pathological and pathological-forensic fields, to make more objective and reproducible quantitative evaluations of pulmonary damage on autopsy or of bioptic material.