Study finds top AI models still struggle with clinical reasoning

Researchers tested 21 frontier large language models on 29 stepwise MSD Manual clinical vignettes and found that, although many models performed well on final diagnosis, they remained much weaker at differential diagnosis and diagnostic testing. They also introduced the PrIME-LLM score, a multidimensional benchmark designed to better capture balanced clinical reasoning across the full workflow rather than raw accuracy alone.

from News Medical Medical Research News Feed https://ift.tt/hOaCUf7