Nature biomedical engineering
Benchmarking foundation models as feature extractors for weakly supervised computational pathology.
Peter Neidlinger, Omar S M El Nahhas, Hannah Sophie Muti, Tim Lenz, Michael Hoffmeister, Hermann Brenner, Marko van Treeck, Rupert Langer, Bastian Dislich, Hans Michael Behrens, Christoph Röcken, Sebastian Foersch, Daniel Truhn, Antonio Marra, Oliver Lester Saldanha
Published: 202510.1038/s41551-025-01516-3
Abstract
Numerous pathology foundation models have been developed to extract clinically relevant information. There is currently limited literature independently evaluating these foundation models on external cohorts and clinically relevant tasks to uncover a…
Preview only. Read the full abstract at the source