1 paper
Pengyang Yu, Yiou Wang, Zhongping Dong +3
Vision-language models return structured chest-radiograph findings through interfaces exposing no confidence score, so a receiving institution cannot read off how far to trust an i…