Neuro-VL vision-language model trained on one million neuroimaging studies
November 13, 2026
Neuro-VL is a vision-language model trained on nearly one million clinical neuroimaging studies. It outperforms general-purpose and specialized medical models in diagnosing and describing abnormalities through natural language conversation.
HOW THIS AFFECTS YOU
●
builderYou can look at how large-scale clinical datasets are used to build specialized vision-language models.
●
healthThis could help clinicians interpret complex imaging examinations using natural language.