The platform functions by leveraging Atlas 2, a proprietary pathology foundation model developed alongside the Mayo Clinic. By generating embeddings from these images, the system combs through a database currently containing over 310,000 whole slide images. Each match provides a path back to a real, diagnosed case, allowing pathologists to inspect the underlying data to inform their own clinical decisions. This approach replaces traditional, time-consuming methods like textbooks or external consultations, which can often cost thousands of dollars per request.

Dr. Simon Schallenberg, an associate professor at Charité – Universitätsmedizin Berlin, noted that the tool provides an immediate reference point that drastically reduces the wait time for second opinions. The current database covers more than 300 diagnostic entities across 27 organs, aligning with the WHO Classification of Tumours. As the system moves toward general availability, the company expects to grow its library to over one million whole slide images. While PathoSearch is currently restricted to research use, it has already been integrated into the Techcyte Fusion AP image management system to facilitate early access testing.