A large private multi-center benchmark of 19 pathology foundation models on 64 tasks finds Virchow2 and H-Optimus-1 best overall, with vision-language models lagging behind.
advances10, eadi0302 (2024)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
PathBench: A comprehensive comparison benchmark for pathology foundation models towards precision oncology
A large private multi-center benchmark of 19 pathology foundation models on 64 tasks finds Virchow2 and H-Optimus-1 best overall, with vision-language models lagging behind.