文献速递腹部技术与安全CT

GPT-4 提升影像报告准确性:减少腹部盆腔 CT 报告错误

Improving radiology reporting accuracy: use of GPT-4 to reduce errors in reports.
PubMed 文献 · Abdom Radiol (NY) · 2025-06-27

本回顾性研究用 GPT-4 零样本评估 600 份已定稿腹部盆腔 CT 报告,91% 报告被提示问题但仅 23% 修订被放射科医师接受,多数修订无或低临床影响,低年资医师接受率更高。

要点

  • GPT-4 对 91% 定稿报告提出修订,但仅 23% 被放射科医师接受。
  • 多数修订针对语法(74%),临床影响多为无(44%)或低(46%)。
  • 低年资(≤7 年)医师对 GPT-4 修订的同意率和接受率均高于高年资医师。

核心结论

  • GPT-4 可识别定稿报告错误并提升清晰度,但多数错误为轻微、无或低临床影响,仅 23% 建议修订被接受。
  • GPT-4 作为报告辅助工具具潜力,但需进一步优化以用于临床实践。

研究设计与样本

  • 回顾性研究,评估三级医疗系统已定稿腹部盆腔 CT 报告,使用 GPT-4 零样本学习。
  • 6 名放射科医师各随机审阅 100 份自己定稿的报告,共 600 份,评估 GPT-4 建议修订的一致性、接受度和临床影响。
  • 比较不同执业年限和性别的放射科医师反应。

诊断标准与关键数值

  • GPT-4 对 91% 的 600 份报告识别问题并提出修订建议;74% 修订针对语法。
  • 放射科医师同意 27% 的修订,接受 23%;临床影响评为无 44%、低 46%、潜在伤害 8%(仅 2 例可能严重伤害)。
  • 执业≤7 年医师同意率 34% 高于 >7 年的 20%(P = .003),接受率 32% 高于 15%(P = .003)。

临床应用与阅片要点

  • GPT-4 可用于辅助发现报告语法错误和提升清晰度,但临床相关错误修订接受率低,需人工审核。
  • 低年资医师可能更受益于 GPT-4 建议,但需注意潜在伤害风险(8%),尤其严重伤害罕见但存在。

局限性

  • 回顾性研究,仅评估腹部盆腔 CT 报告,使用零样本学习,未与其他模型或人工校对比较。
  • 样本来自单一三级医疗系统,放射科医师仅审阅自己的报告,可能存在偏倚。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

PURPOSE: Radiology reports are essential for communicating imaging findings to guide diagnosis and treatment. Although most radiology reports are accurate, errors can occur in the final reports due to high workloads, use of dictation software, and human error. Advanced artificial intelligence models, such as GPT-4, show potential as tools to improve report accuracy. This retrospective study evaluated how GPT-4 performed in detecting and correcting errors in finalized radiology reports in real-world settings for abdominopelvic computed tomography (CT) reports.

METHODS: We evaluated finalized CT abdominopelvic reports from a tertiary health system by using GPT-4 with zero-shot learning techniques. Six radiologists each reviewed 100 of their finalized reports (randomly selected), evaluating GPT-4's suggested revisions for agreement, acceptance, and clinical impact. The radiologists' responses were compared by years in practice and sex.

RESULTS: GPT-4 identified issues and suggested revisions for 91% of the 600 reports; most revisions addressed grammar (74%). The radiologists agreed with 27% of the revisions and accepted 23%. Most revisions were rated as having no (44%) or low (46%) clinical impact. Potential harm was rare (8%), with only 2 cases of potentially severe harm. Radiologists with less experience (≤ 7 years of practice) were more likely to agree with the revisions suggested by GPT-4 than those with more experience (34% vs. 20%, P = .003) and accepted a greater percentage of the revisions (32% vs. 15%, P = .003).

CONCLUSIONS: Although GPT-4 showed promise in identifying errors and improving the clarity of finalized radiology reports, most errors were categorized as minor, with no or low clinical impact. Collectively, the radiologists accepted 23% of the suggested revisions in their finalized reports. This study highlights the potential of GPT-4 as a prospective tool for radiology reporting, with further refinement needed for consistent use in clinical practice.

来源:Abdom Radiol (NY),查看原文。版权归原作者所有。

引用

  1. Mayes CJ, Reyes C, Truman ME, et al. Improving radiology reporting accuracy: use of GPT-4 to reduce errors in reports. Abdom Radiol (NY). 2025. PMID: 40576664. doi:10.1007/s00261-025-05079-4

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。