This Arkangel AI whitepaper is a community evidence brief on artificial intelligence in digital pathology, built from 24 selected sources after an exploratory search and a focused July 2026 update in Europe PMC. It organizes the evidence by task (detection and prioritization in whole-slide images, grading and quantification, molecular and tissue-of-origin prediction, and workflow and telepathology) and by stage, from retrospective development to routine implementation.
The most mature uses are bounded detection or prioritization tasks and support for grading or quantification, such as lymph node metastasis detection and Gleason grading; predicting molecular alterations or tissue of origin from histology remains mostly retrospective and does not replace molecular testing. Only two direct Latin American sources were found, a Delphi panel spanning seven countries and an expert review, and both describe barriers of cost, connectivity, regulation, training and interoperability rather than measuring effectiveness. The authors recommend reviewable tasks, validation by laboratory, stain and scanner, and explicit professional accountability. It is a whitepaper, not a systematic review and not peer reviewed.
