MedStar rolls out PathAI platform

A hospital pathologist used to do the core of the job with glass slides, a microscope, and a bench full of folders. MedStar Health is now pushing that work onto a software platform from PathAI across its pathology services under a multi-year deal announced on April 7, 2026. (pathai.com) That shift sounds simple until you picture what a pathology case actually is. A tissue sample is cut into a thin slice, stained so cells stand out, mounted on glass, and then examined by a physician called a pathologist, who decides whether the tissue looks normal, inflamed, precancerous, or malignant. (cap.org) Digital pathology turns that glass slide into a very large image file. A scanner photographs the entire slide at high resolution so the pathologist can pan and zoom on a screen the way a person uses a map app to move from a city view down to a street corner. (cap.org) The file is usually called a whole slide image, which means one digital copy of the entire specimen slide rather than a few snapshots. That matters because diagnosis often depends on comparing tiny suspicious areas with the surrounding tissue architecture across the full specimen. (cap.org) Once slides become digital, the work around them changes too. Instead of moving trays of glass between rooms, a laboratory can route cases electronically, attach prior images and reports, and let specialists review the same case from different sites without waiting for a courier. (cap.org) Artificial intelligence enters after that basic digitization step. In pathology, these systems do not replace tissue preparation or final medical judgment; they are software tools that can highlight regions of interest, measure features, count cells, or organize work so the pathologist spends less time hunting and more time deciding. (pathai.com) That is the backdrop for the MedStar-PathAI announcement. PathAI said MedStar Health will deploy its AISight Dx digital pathology platform and various artificial intelligence algorithm products across MedStar’s multi-site network, moving the technology from a pilot into broader operational use. (pathai.com) PathAI’s platform is not just an image viewer. The company says AISight Dx is an image management system for primary diagnosis in clinical settings, and it describes the product as a cloud-native platform that centralizes case and image management. (pathai.com, pathai.com) Regulators have already looked at the core platform. PathAI announced on June 30, 2025 that the United States Food and Drug Administration granted 510(k) clearance for AISight Dx for use in primary diagnosis, and the Food and Drug Administration database lists decision date June 26, 2025 for clearance K243391. (pathai.com, fda.gov) The Food and Drug Administration review summary gives a more concrete picture of what that clearance covers. It describes AISight Dx version 2.9 as web-based software for viewing, interpreting, and managing digital whole slide images from surgical pathology slides scanned on specified Leica Aperio and Hamamatsu scanners. (fda.gov) That scanner detail is not trivia. Digital pathology systems depend on the chain from tissue preparation to scanner hardware to image display, so a health system cannot assume that any scanner, monitor, workflow, or algorithm will behave the same way in every lab. (fda.gov, cap.org) That is why large deployments come with validation work before they become ordinary daily practice. The College of American Pathologists says laboratories implementing whole slide imaging for clinical diagnostic purposes should perform their own validation studies to confirm diagnostic accuracy and equivalence with traditional light microscopy. (cap.org, cap.org) A system-wide rollout also changes training, not just technology. When sign-out moves from microscope-based review to digital review, pathologists, laboratory staff, and trainees need to learn new habits for scanning quality checks, case navigation, image sharing, reporting, and the limits of software-generated prompts. (cap.org, cap.org) For MedStar, the practical significance is that digital pathology is no longer being treated as a side experiment. The announcement describes a move from pilot work to deployment across a multi-site health system, which means image management, artificial intelligence-assisted review, and workflow routing are being built into routine pathology operations rather than tested at the edges. (pathai.com) For the industry, this is the part worth watching over the next year. The headline is a partnership, but the real test will be whether a large health system can make digital sign-out, algorithm oversight, scanner compatibility, local validation, and staff training work together every day without slowing diagnosis or adding new failure points. (pathai.com, cap.org)