SOFTWARE-ASSISTED IMAGE ANALYSIS FOR IDENTIFICATION AND QUANTIFICATION OF HEPATIC SINUSOIDAL DILATATION AND CENTRILOBULAR FIBROSIS

BACKGROUND: Heart dysfunction and liver disease often coexist because of systemic disorders. Any cause of right ventricular failure may precipitate hepatic congestion and fibrosis. Digital image technologies have been introduced to pathology diagnosis, allowing an objective quantitative assessment. The quantification of fibrous tissue in liver biopsy sections is extremely important in the classification, diagnosis and grading of chronic liver disease. AIM: To create a semi-automatic computerized protocol to quantify any amount of centrilobular fibrosis and sinusoidal dilatation in liver Masson’s Trichrome-stained specimen. METHOD: Once fibrosis had been established, liver samples were collected, histologically processed, stained with Masson’s trichrome, and whole-slide images were captured with an appropriated digital pathology slide scanner. After, a random selection of the regions of interest (ROI’s) was conducted. The data were subjected to software-assisted image analysis (ImageJ(®)). RESULTS: The analysis of 250 ROI’s allowed to empirically obtain the best application settings to identify the centrilobular fibrosis (CF) and sinusoidal lumen (SL). After the establishment of the colour threshold application settings, an in-house Macro was recorded to set the measurements (fraction area and total area) and calculate the CF and SL ratios by an automatic batch processing. CONCLUSION: Was possible to create a more detailed method that identifies and quantifies the area occupied by fibrous tissue and sinusoidal lumen in Masson’s trichrome-stained livers specimens.