Scholarlink

Objectives <p>To evaluate the performance of large language models (LLMs), including retrieval-augmented generation (RAG)-based approaches, in extracting components and management recommendations from structured coronary computed tomography angiography (CCTA) reports according to the Coronary Artery Disease Reporting and Data System (CAD-RADS&#xa0;2.0).</p> Materials and methods <p>A&#xa0;total of 320 fully structured CCTA reports were analyzed using LLM. Closed-source standard ChatGPT‑5, NotebookLM (RAG-based model), and a&#xa0;RAG-adapted ChatGPT‑5 model (ChatGPT-5-RAG) were used. Each model extracted the CAD-RADS category, plaque burden, presence of high-risk plaque (HRP), other modifiers, full score, and management recommendations in accordance with the CAD-RADS 2.0&#xa0;guidelines. We compared LLM outputs with reference standards determined by two expert cardiovascular radiologists.</p> Results <p>ChatGPT-5-RAG showed the highest accuracy for CAD-RADS classification (0.959, 95% CI: 0.932–0.976), plaque burden (0.912, 95% CI: 0.876–0.939), HRP detection (0.988, 95% CI: 0.968–0.995), other modifiers (0.950, 95% CI: 0.920–0.969), and full score (0.828, 95% CI: 0.783–0.866). Closed-source ChatGPT‑5 showed the weakest performance across all components. Significant statistical differences were found among the three models (<i>p</i> &lt; 0.001). Management recommendations were qualitatively rated on a&#xa0;three-point Likert scale; although agreement between models was low, ChatGPT-5-RAG and NotebookLM performed almost perfectly (median 3&#xa0;points).</p> Conclusion <p>This study demonstrates that RAG-enhanced LLMs significantly improve accuracy and reliability in extracting CAD-RADS&#xa0;2.0 components and generating clinical management recommendations. The findings highlight the potential of RAG-based LLMs as innovative, explainable tools for automated and standardized CCTA reporting in clinical radiology workflows.</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

Retrieval-augmented generation-enhanced large language models for comprehensive CAD-RADS 2.0 categorization from structured coronary CTA reports

  • 期刊论文

  • 发表日期：

    2026年2月20日

查看全文
  • Esat Kaba,
  • Yusuf Çubukçu,
  • Burak Uzunibrahimoğlu,
  • Yusuf Enes Yılmaz,
  • Mehmet Çınar,
  • Serdar Tabakoğlu,
  • Elif Merve Bal,
  • Yaprak Seren Beydüz,
  • Merve Solak,
  • Evin Oğuz,
  • Ayşenur Topçu Varlık,
  • Gökçen Malkoç,
  • Mehmet Beyazal,
  • Fatma Beyazal Celiker,
  • Selçuk Akkaya

摘要

Objectives

To evaluate the performance of large language models (LLMs), including retrieval-augmented generation (RAG)-based approaches, in extracting components and management recommendations from structured coronary computed tomography angiography (CCTA) reports according to the Coronary Artery Disease Reporting and Data System (CAD-RADS 2.0).

Materials and methods

A total of 320 fully structured CCTA reports were analyzed using LLM. Closed-source standard ChatGPT‑5, NotebookLM (RAG-based model), and a RAG-adapted ChatGPT‑5 model (ChatGPT-5-RAG) were used. Each model extracted the CAD-RADS category, plaque burden, presence of high-risk plaque (HRP), other modifiers, full score, and management recommendations in accordance with the CAD-RADS 2.0 guidelines. We compared LLM outputs with reference standards determined by two expert cardiovascular radiologists.

Results

ChatGPT-5-RAG showed the highest accuracy for CAD-RADS classification (0.959, 95% CI: 0.932–0.976), plaque burden (0.912, 95% CI: 0.876–0.939), HRP detection (0.988, 95% CI: 0.968–0.995), other modifiers (0.950, 95% CI: 0.920–0.969), and full score (0.828, 95% CI: 0.783–0.866). Closed-source ChatGPT‑5 showed the weakest performance across all components. Significant statistical differences were found among the three models (p < 0.001). Management recommendations were qualitatively rated on a three-point Likert scale; although agreement between models was low, ChatGPT-5-RAG and NotebookLM performed almost perfectly (median 3 points).

Conclusion

This study demonstrates that RAG-enhanced LLMs significantly improve accuracy and reliability in extracting CAD-RADS 2.0 components and generating clinical management recommendations. The findings highlight the potential of RAG-based LLMs as innovative, explainable tools for automated and standardized CCTA reporting in clinical radiology workflows.