We keep hearing that one of the promises of digital pathology is the ability to develop AI based algorithms and tools that will allow us to do things that we never have before. This gets many people excited in the industry - particularly computer scientists, programmers and machine learning specialists - but what about the role of the pathologist? How can we the keep pathologist engaged in the development of AI algorithms and digital pathology workflows?

One of the things that has hindered surgical pathology (maybe just in perception) is lack of standardization and reproducibility both in terms of molecular and IHC testing as well as interpretation of H&E images. Part of the promise of digital pathology is that we could eliminate this variability or subjectivity. But wouldn't pathologist developing their own algorithms recreate similar problems?

Rish Pai, MD, PhD is Professor of Laboratory Medicine & Pathology and serves as a Consultant Pathologist at the Mayo Clinic, Arizona and Associate Chair of Research. Dr. Pai has lectured nationally and internationally on many aspects of gastrointestinal and hepatic pathology and has directed multiple GI and Liver CME courses. He is Past-President of the Rodger C. Haggitt Gastrointestinal Pathology Society. Dr. Pai is a regular reviewer for numerous pathology and gastroenterology journals and serves on the Editorial Board of the American Journal of Surgical Pathology. He has published 130 articles and written multiple book chapters in the fields of gastrointestinal and hepatic pathology.

Pathology is a unique specialty. Much of what we do becomes a “permanent record,” and yet there remains an inherent subjectivity and imperfection. Do we have a special obligation to get it right?

Ver todos los episodios