video1.0<iframe src="https://www.loom.com/embed/fb5f338d3276464c972632a8c40c2755" frameborder="0" width="1720" height="1290" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12901720Loomhttps://www.loom.com12901720https://cdn.loom.com/sessions/thumbnails/fb5f338d3276464c972632a8c40c2755-dfcc877944164deb.gif180.556This Loom presents a method for detecting cancer regions in extremely large whole slide images using a foundational model plus a trained classifier. It explains that whole slide images are very high resolution and that normal machine learning models do not work on these massive inputs. The author describes an application designed to augment pathologists, letting them agree or disagree with tumor labels and optionally change labels. They also discuss added features for annotations, including notes mode, and a narration and reporting workflow using 11labs and Gemini APIs.