Color correction for automatic fibrosis quantification in liver biopsy specimens

Area of Science:

• Digital pathology
• Medical image analysis
• Liver fibrosis quantification

Background:

• Precise liver fibrosis quantification is crucial for patient management.
• Manual extraction of fiber areas in digital pathology images is labor-intensive due to color variations.
• Existing methods lack objectivity and automation in fibrosis assessment.

Purpose of the Study:

• To propose a color correction method for whole slide images (WSIs) of Elastica van Gieson (EVG) stained liver biopsy specimens.
• To enable automated image analysis for accurate liver fibrosis quantification.

Main Methods:

• Developed a color correction technique modifying color distribution in WSIs to match a reference standard.
• Utilized a dataset of 38 EVG-stained liver biopsy WSIs from chronic viral hepatitis patients.
• Performed fibrosis quantification via tissue classification analysis on color-corrected WSIs.

Main Results:

• Color correction significantly enhanced the correlation between fibrosis quantification and liver stiffness (r = 0.61-0.68).
• Without color correction, correlation coefficients were substantially lower (r = 0.40-0.50).
• The method demonstrated robust performance even with limited training data.

Conclusions:

• The proposed color correction method facilitates objective liver fibrosis evaluation.
• Automated fibrosis quantification complements existing semi-quantitative histological assessment methods.
• This approach offers a valuable tool for precise and reproducible liver fibrosis assessment.