文献速递腹部技术与安全CT

使用GPT-4将胰腺癌混合语言自由文本CT报告转换为美国国家综合癌症网络结构化报告模板

Conversion of Mixed-Language Free-Text CT Reports of Pancreatic Cancer to National Comprehensive Cancer Network Structured Reporting Templates by Using GPT-4.
PubMed 文献 · Korean J Radiol · 2025-04-17

本回顾性研究纳入两家机构2021年1月至2023年12月胰腺癌分期CT自由文本报告,评估GPT-4 Turbo与GPT-4o生成NCCN结构化报告及判断可切除性的准确性。结构化报告生成准确率两者相当(92.3% vs 92.2%,P=0.923),可切除性分类GPT-4 Turbo更高(81.7% vs 67.0%,P=0.002),但作者强调需放射科医师监督。

要点

  • GPT-4 Turbo与GPT-4o生成结构化报告准确率相当(92.3% vs 92.2%)
  • 可切除性分类GPT-4 Turbo准确率高于GPT-4o(81.7% vs 67.0%)
  • 作者认为基于CT判断可切除性仍需放射科医师监督

核心结论

  • GPT-4 Turbo和GPT-4o在从混合语言叙述性报告生成基于NCCN的PDAC结构化报告方面均表现出可接受的准确性,但基于CT判断可切除性仍需放射科医师监督。

研究设计与样本

  • 回顾性研究,纳入两家机构2021年1月至2023年12月胰腺癌分期胰腺方案CT的连续自由文本报告,使用115份来自机构A的报告评估GPT-4 Turbo和GPT-4o生成结构化报告及按NCCN指南2.2024版分类可切除性的性能,参考标准由腹部放射科医师手动生成。

诊断标准与关键数值

  • 结构化报告生成准确率GPT-4 Turbo 92.3%(1592/1725)vs GPT-4o 92.2%(1590/1725),P=0.923;可切除性分类准确率GPT-4 Turbo 81.7%(94/115)vs GPT-4o 67.0%(77/115),P=0.002。

临床应用与阅片要点

  • 作者认为基于CT判断可切除性仍需放射科医师监督;作者未据此给出其他临床推荐。

局限性

  • 原文未报告局限性。

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

OBJECTIVE: To evaluate the feasibility of generative pre-trained transformer-4 (GPT-4) in generating structured reports (SRs) from mixed-language (English and Korean) narrative-style CT reports for pancreatic ductal adenocarcinoma (PDAC) and to assess its accuracy in categorizing PDCA resectability.

MATERIALS AND METHODS: This retrospective study included consecutive free-text reports of pancreas-protocol CT for staging PDAC, from two institutions, written in English or Korean from January 2021 to December 2023. Both the GPT-4 Turbo and GPT-4o models were provided prompts along with the free-text reports via an application programming interface and tasked with generating SRs and categorizing tumor resectability according to the National Comprehensive Cancer Network guidelines version 2.2024. Prompts were optimized using the GPT-4 Turbo model and 50 reports from Institution B. The performances of the GPT-4 Turbo and GPT-4o models in the two tasks were evaluated using 115 reports from Institution A. Results were compared with a reference standard that was manually derived by an abdominal radiologist. Each report was consecutively processed three times, with the most frequent response selected as the final output. Error analysis was guided by the decision rationale provided by the models.

RESULTS: Of the 115 narrative reports tested, 96 (83.5%) contained both English and Korean. For SR generation, GPT-4 Turbo and GPT-4o demonstrated comparable accuracies (92.3% [1592/1725] and 92.2% [1590/1725], respectively; P = 0.923). In the resectability categorization, GPT-4 Turbo showed higher accuracy than GPT-4o (81.7% [94/115] vs. 67.0% [77/115], respectively; P = 0.002). In the error analysis of GPT-4 Turbo, the SR generation error rate was 7.7% (133/1725 items), which was primarily attributed to inaccurate data extraction (54.1% [72/133]). The resectability categorization error rate was 18.3% (21/115), with the main cause being violation of the resectability criteria (61.9% [13/21]).

CONCLUSION: Both GPT-4 Turbo and GPT-4o demonstrated acceptable accuracy in generating NCCN-based SRs on PDACs from mixed-language narrative reports. However, oversight by human radiologists is essential for determining resectability based on CT findings.

来源:Korean J Radiol,查看原文。版权归原作者所有。

引用

  1. Kim H, Kim B, Choi MH, et al. Conversion of Mixed-Language Free-Text CT Reports of Pancreatic Cancer to National Comprehensive Cancer Network Structured Reporting Templates by Using GPT-4. Korean J Radiol. 2025. PMID: 40288895. doi:10.3348/kjr.2024.1228

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。