AnalysisHealthOctober 29, 2026

EVLF-FM clinical vision-language model hits 0.895 diagnostic accuracy

Read original source →nature.com

Trained on 1.3M+ samples from 18 datasets across 10 imaging modalities and six specialties, EVLF-FM beat generalist and specialist models on internal diagnostics. Human raters scored its reasoning 4.25/5 under the S.C.O.R.E. framework; authors say prospective clinical studies are still needed.

1 source

More stories today

Open the live feed