Scholarlink

<p>Clinical evidence synthesis largely relies on systematic reviews (SR) of clinical studies from medical literature. Here, we propose a generative artificial intelligence (AI) pipeline named <Emphasis FontCategory="NonProportional">TrialMind</Emphasis> to streamline study search, study screening, and data extraction tasks in SR. We chose published SRs to build <Emphasis FontCategory="NonProportional">TrialReviewBench</Emphasis>, which contains 100 SRs and 2,220 clinical studies. For study search, it achieves high recall rates (Ours 0.711–0.834 v.s. Human baseline 0.138–0.232). For study screening, <Emphasis FontCategory="NonProportional">TrialMind</Emphasis> beats previous document ranking methods in a 1.5–2.6 fold change. For data extraction, it outperforms a GPT-4’s accuracy by 16–32%. In a pilot study, human-AI collaboration with <Emphasis FontCategory="NonProportional">TrialMind</Emphasis> improved recall by 71.4% and reduced screening time by 44.2%, while in data extraction, accuracy increased by 23.5% with a 63.4% time reduction. Medical experts preferred <Emphasis FontCategory="NonProportional">TrialMind</Emphasis>’s synthesized evidence over GPT-4’s in 62.5%-100% of cases. These findings show the promise of accelerating clinical evidence synthesis driven by human-AI collaboration.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Accelerating clinical evidence synthesis with large language models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月8日

查看全文
  • Zifeng Wang,
  • Lang Cao,
  • Benjamin Danek,
  • Qiao Jin,
  • Zhiyong Lu,
  • Jimeng Sun

摘要

Clinical evidence synthesis largely relies on systematic reviews (SR) of clinical studies from medical literature. Here, we propose a generative artificial intelligence (AI) pipeline named TrialMind to streamline study search, study screening, and data extraction tasks in SR. We chose published SRs to build TrialReviewBench, which contains 100 SRs and 2,220 clinical studies. For study search, it achieves high recall rates (Ours 0.711–0.834 v.s. Human baseline 0.138–0.232). For study screening, TrialMind beats previous document ranking methods in a 1.5–2.6 fold change. For data extraction, it outperforms a GPT-4’s accuracy by 16–32%. In a pilot study, human-AI collaboration with TrialMind improved recall by 71.4% and reduced screening time by 44.2%, while in data extraction, accuracy increased by 23.5% with a 63.4% time reduction. Medical experts preferred TrialMind’s synthesized evidence over GPT-4’s in 62.5%-100% of cases. These findings show the promise of accelerating clinical evidence synthesis driven by human-AI collaboration.