研究：大模型从临床文本提取旅行史，时间变量仍难可靠提取

【原文标题】Can large language models extract travel history from clinical text? A multi-annotator benchmark study