This is part seven of our report on the Digital Pathology & AI Congress: USA. For other parts, please see links toward the bottom of the article. The conference featured insightful discussions on the practical aspects and potential impact of machine learning applications in pathology, showcasing the potential of AI to enhance diagnostic accuracy and efficiency, improve patient care, and foster personalized medicine. Parts five through eight will cover these discussions.

Brittany Dugger of the University of California, Davis, and David A. Gutman of Emory University provided valuable perspectives from both pathology and machine learning disciplines, offering advice on enhancing cross-disciplinary dialogues and fostering collaboration between these two fields.[1] Their session aimed to facilitate the effective integration of machine learning approaches into digital pathology workflows.

Dr. Dugger, representing the pathology domain, shared insights into the unique challenges and requirements pathologists face when integrating machine learning solutions into their workflows. Pathologists know what is needed to deliver high-quality health care, whereas computer experts know what machine learning can and cannot do. Dr. Dugger emphasized that clear communication and understanding between the two disciplines are needed to ensure that the developed solutions address real-world needs and can be effectively integrated into routine practice. She also emphasized the importance of extending the dialogue to statisticians, clinical scientists, radiologists, students, and staff.

From the machine learning perspective, Dr. Gutman discussed the technical considerations and best practices for developing robust and generalizable models tailored to the complexities of digital pathology data. Using a brain bank as a case study, he highlighted the importance of domain expertise and close collaboration with pathologists to ensure that models capture clinically relevant features and provide interpretable and actionable outputs. Machine learning techniques can streamline these tasks; however, variations in sampling strategies and staining protocols create challenges. Multidisciplinary communication and collaboration are crucial for harmonizing neuropathology across brain banks.

Dr. Dugger and Dr. Gutman emphasized the need for continuous dialogue and knowledge exchange between pathologists, computer scientists, and machine learning experts throughout the development and deployment process. They noted that establishing a shared vision and goals,  facilitating communication through regular check-ins and brainstorming sessions, asking open-ended questions, defining roles and skillsets, breaking down jobs into smaller tasks, and providing parameters of needs and time commitments (e.g., SMART goals) are essential for building an effective multidisciplinary team. This collaborative approach is critical for building robust and clinically relevant AI solutions for digital pathology that can positively affect patient care.Dr. Gutman, Assistant Professor of Biomedical Informatics, Emory University.

Links To Other Parts Of The Series

Part 1: Highlights from the 10th Digital Pathology & AI Congress: USA

Part 2: Digital Pathology Implementation: Insights From Experts At DP&AI: USA

Part 3: Clinical Implementation Challenges And Potential Solutions

Part 4: Recent Advances In Digital Pathology

Part 5: Reshaping the Future of Medical Care, Education and Research: The Pivotal Roles of Synthetic Data, Generative AI, and Auto-MLs

Part 6: Revolutionizing Pathology Practices For Improved Precision Oncology: Leveraging AI to Enhance Efficacy of Immunotherapy

References

[1] Brittany Dugger and David A. Gutman. Cross-disciplinary dialogs for building out machine learning: Advice from the digital pathology road. Presented at the 10th Digital Pathology & AI Congress: USA; May 7-8, 2024; San Diego, CA.