Artificial intelligence in surgical pathology - Where do we stand, where do we go?

Area of Science:

• Computational pathology
• Digital pathology
• Artificial intelligence in medicine

Background:

• Pathologists seek novel, disease-specific features for tumor prognosis and classification.
• Artificial intelligence (AI) and machine learning (ML) offer potential solutions for tumor outcome prediction and biomarker discovery.
• AI integration in pathology workflows enhances accuracy, reproducibility, and productivity.

Purpose of the Study:

• To review the infrastructure required for digital and computational pathology.
• To address barriers to AI deployment in clinical pathology settings.
• To summarize AI applications in intraoperative and postoperative pathology, including quality assessment and emerging technologies.

Main Methods:

• Review of existing literature and technologies in digital and computational pathology.
• Discussion of AI/ML applications in analyzing histological slides.
• Exploration of infrastructure needs, clinical deployment challenges, and quality assessment.

Main Results:

• AI/ML systems can improve tumor grading, staging, and the identification of prognostic biomarkers.
• Digital pathology infrastructure is crucial for AI integration.
• Barriers to clinical AI deployment include data standardization and validation.

Conclusions:

• AI and ML are transforming pathology by enhancing diagnostic capabilities and biomarker discovery.
• Digital and computational pathology require robust infrastructure and standardized workflows for full clinical integration.
• Future innovations like large language models and optical biopsies promise further advancements in pathology.