An analysis on efficacy of applying β-elemene intervention on chemically -induced tongue lesions using SAM algorithm

Area of Science:

• Pathology
• Artificial Intelligence
• Computational Biology

Background:

• Chemical carcinogens induce structural changes in tongue lesions.
• Accurate quantification of these changes is crucial for diagnosis and treatment assessment.
• Digital pathology and AI offer novel approaches to analyze histopathological data.

Purpose of the Study:

• To develop and validate an AI model for quantifying structural changes in chemically induced tongue lesions.
• To assess the efficacy of β-elemene treatment by analyzing epithelial thickness and papilla-like protrusions.
• To leverage the Segment Anything Model (SAM) for enhanced image segmentation in histopathology.

Main Methods:

• Utilized a 4-nitroquinoline-N-oxide induced tongue cancer model in rodents.
• Processed 183 digital pathology slides using AI, including SAM for segmentation.
• Employed OpenCV for contour analysis and skeletonization to measure epithelial thickness and papillae.

Main Results:

• The AI model accurately measured tongue epithelial thickness and papilla count.
• Carcinogen-induced lesions showed significantly increased epithelial thickness and decreased papillae.
• β-elemene treatment partially reversed these changes, indicating therapeutic effects.

Conclusions:

• The SAM-based AI framework effectively quantifies key pathological features in tongue lesions.
• This AI tool assists pathologists in objective assessment of disease progression and treatment response.
• The findings highlight AI's potential in advancing cancer research and diagnostics.