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

Douglas Mesadri Gewehr1,2,3, Allan Fernando Giovanini1, Sofia Inez Munhoz1,2

  • 1Mackenzie Evangelical Faculty of Paraná, Curitiba, Paraná, Brazil.

Insights

A new semi-automatic protocol quantifies centrilobular fibrosis (CF) and sinusoidal lumen (SL) in liver Masson

Area of Science:

  • Digital pathology
  • Histopathology
  • Medical image analysis

Background:

  • Heart dysfunction and liver disease frequently coexist due to systemic conditions.
  • Right ventricular failure can lead to hepatic congestion and fibrosis.
  • Quantitative assessment of liver biopsy sections is crucial for diagnosing and grading chronic liver disease.

Purpose of the Study:

  • Develop a semi-automatic computerized protocol for quantifying centrilobular fibrosis (CF) and sinusoidal dilatation (SL).
  • Utilize Masson's Trichrome-stained liver specimens for analysis.

Main Methods:

  • Liver samples were processed, stained with Masson's Trichrome, and scanned into whole-slide images.
  • Regions of interest (ROIs) were selected for software-assisted image analysis using ImageJ®.
  • An in-house Macro was developed to automatically process measurements and calculate CF and SL ratios.

Main Results:

  • Empirical determination of optimal settings for identifying CF and SL across 250 ROIs.
  • Establishment of color threshold settings for accurate quantification.
  • Successful automatic batch processing for calculating fractional area and total area ratios.

Conclusions:

  • A detailed method was successfully created to identify and quantify fibrous tissue and sinusoidal lumen.
  • This protocol enhances the objective assessment of liver pathology in Masson's Trichrome-stained specimens.
Abstract