Image Analysis-Based Approaches for Scoring Mouse Models of Colitis.

Area of Science:

• Veterinary pathology
• Digital pathology
• Immunohistochemistry

Background:

• Mouse models are crucial for studying inflammatory bowel disease (IBD).
• Histologic assessment is a common endpoint for evaluating IBD mouse models.
• Digital pathology and image analysis offer objective quantification of histologic features.

Purpose of the Study:

• To develop an automated image analysis workflow for scoring IBD mouse models.
• To assess the utility of digital pathology for quantifying immune cell infiltrates.
• To evaluate the correlation between automated scoring and pathologist assessment.

Main Methods:

• A commercially available image analysis platform was utilized.
• The workflow was applied to two mouse models of colitis.
• Automated quantification of CD3 immunolabeling was performed.
• Hematoxylin and eosin-stained sections were assessed for segmentation accuracy.

Main Results:

• Automated quantification of CD3 immunolabeling strongly correlated with pathologist scores.
• The image analysis workflow identified 8 out of 9 treatment-related differences found by pathologists.
• The software struggled with accurate segmentation of hematoxylin and eosin-stained slides.

Conclusions:

• Digital image analysis can be integrated into pathology workflows for IBD mouse models.
• Immunohistochemical markers are valuable surrogates for automated histologic assessment.
• Automated quantification of CD3+ cells provides a reliable, nonbiased endpoint.