Screening For Bone Marrow Cellularity Changes in Cynomolgus Macaques in Toxicology Safety Studies Using Artificial...

Area of Science:

• Veterinary Pathology
• Computational Pathology
• Toxicologic Pathology

Background:

• Hematolymphoid organ cellularity, including bone marrow, is crucial in toxicologic safety studies.
• Evaluating bone marrow cellularity traditionally requires expert pathologist assessment.
• Artificial intelligence (AI) offers potential for automated diagnostic support.

Purpose of the Study:

• To assess the capability of a deep-learning AI model in evaluating whole slide images of macaque sternebrae.
• To identify and enumerate bone marrow hematopoietic cells using AI.
• To validate AI performance against established severity scores.

Main Methods:

• A deep-learning AI model was trained to differentiate hematopoietic cells from other sternebrae tissues.
• The AI model analyzed whole slide images of macaque sternebrae.
• AI-generated cell counts (cells/mm²) were compared to existing severity scores in a study with altered hematopoietic cellularity.

Main Results:

• The AI model successfully differentiated hematopoietic cells from surrounding tissues.
• A correlation was observed between AI-determined cell counts and severity scores, with lower counts corresponding to increased severity.
• The AI model demonstrated potential for rapid and agile generation by a single pathologist.

Conclusions:

• AI shows significant promise for bone marrow screening in toxicologic pathology.
• AI tools can provide efficient and accurate diagnostic support for pathologists.
• This study validates the feasibility of AI-driven bone marrow cellularity assessment.