利用大型语言模型协助医生改善临床试验的患者预筛选（英文版）

英文标题：Improving Patient Pre-screening for Clinical Trials: Assisting Physicians with Large Language Models中文摘要：使用 InstructGPT 帮助医生根据患者的医疗概况确定临床试验的资格，通过组合一次性、选择推论和思考链技术，将 LLM 应用于 10 个合成患者概况，并有医生参与辅助判断，可实现 1.0 的召回率和 0.71 的精确率。英文摘要：Physicians considering clinical trials for their patients are met with thelaborious process of checking many text based eligibility criteria. LargeLanguage Models (LLMs) have shown to perform well for clinical informationextraction and clinical reasoning, including medic