Evaluating AI in Mohs Surgery Post-Operative Questions Supplemental Table 1
收藏Mendeley Data2026-04-09 收录
下载链接：
https://data.mendeley.com/datasets/kxwzmgx62v/2
下载链接
链接失效反馈官方服务：
资源简介：
Supplementary Table 1. Summary of reviewer-flagged deficiencies by AI model. Eight Mohs surgeons reviewed 36 responses (12 questions × 3 models) and flagged deficiencies using pre-specified categories. Multiple deficiencies could be flagged per response. LLaMA 4 accounted for the highest number of omissions and ambiguous guidance. Total counts per model and category are shown.
补充表1. AI模型审阅标记缺陷汇总。8名莫氏外科医师（Mohs surgeon）审阅了36份回复（12个问题×3个模型），并按照预先设定的类别标记缺陷。每份回复可标记多项缺陷。LLaMA 4的遗漏与指引模糊问题数量最多。各模型及各类别的总计数已列明。
提供机构：
Northwestern University; Oregon Health & Science University; Washington University School of Medicine in St. Louis; University of Wisconsin System; Mayo Clinic Health System; Brigham and Women's Hospital; University of California San Francisco
