Objective:

To emphasize the indispensable role of pathologists in shaping AI tools for pathology, ensuring they are clinically relevant and effective.

Key Findings:

• Pathologists provide essential clinical expertise that ensures AI tools are relevant and usable, as evidenced by successful projects at IMP Diagnostics.
• Early digital pathology systems lacking pathologist input failed to meet clinical needs, demonstrating the risks of exclusion.
• Multidisciplinary collaboration enhances the development and effectiveness of computational pathology tools, leading to better clinical outcomes.
• Barriers to pathologist involvement include slow adoption of digital pathology and inconsistent training opportunities, which hinder effective collaboration.

Interpretation:

Involving pathologists in AI development is crucial for creating tools that genuinely address clinical workflow challenges and significantly improve patient outcomes.

Limitations:

• Slow adoption of digital pathology in hospitals limits meaningful work in computational pathology, as seen in various case studies.
• Inconsistent implementation of digital pathology hinders training and engagement of pathology residents, affecting future workforce readiness.

Conclusion:

AI in pathology must be developed with pathologists' input to ensure clinical relevance and effectiveness, emphasizing the need for ongoing collaboration across disciplines.