ONCO-RADS–guided Large Language Models for Extraction and Classification of Incidental Findings on Whole-Body Imaging Reports

Mickaël Tordjman; Murat Yuce; Zelong Liu; Anis Meribout; Ian Bolger; Amine Geahchan; Francisco Restrepo; Himanshu Joshi; H Lee; Timothy Deyer; Z Fayad; Sara Lewis; Giuseppe Petralia; Xueyan Mei; Bachir Taouli

doi:10.1148/rycan.250484

April 17, 2026Radiology Imaging CancerOpen Access

ONCO-RADS–guided Large Language Models for Extraction and Classification of Incidental Findings on Whole-Body Imaging Reports

View Full Paper

Ask AI

Bookmark

Share

Key Points

To evaluate the effectiveness of large language models in extracting and classifying incidental findings using ONCO-RADS prompts.
Utilized ChatGPT-o1 and Gemini-2.5-Pro language models for data extraction.
Guided the models with ONCO-RADS prompts for enhanced accuracy.
Compared results with models not utilizing ONCO-RADS prompts and a named entity recognition model.
Large language models showed improved accuracy in extracting findings with ONCO-RADS guidance.
Outperformed extraction strategies without prompts and the named entity recognition model.

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

Large language models accurately classify incidental findings in imaging reports, highlighting effective data extraction methods.

Abstract

The large language models ChatGPT-o1 and Gemini-2.5-Pro, guided with Oncologically Relevant Findings Reporting and Data System (ONCO-RADS) prompts, accurately extracted and classified incidental findings from whole-body imaging reports and outperformed these strategies without ONCO-RADS prompts and the medical named entity recognition model.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Tordjman et al. (2026) studied this question.

synapsesocial.com/papers/69e4739a010ef96374d8f5c0 https://doi.org/10.1148/rycan.250484

View Full Paper

Ask AI

Bookmark

Share

ONCO-RADS–guided Large Language Models for Extraction and Classification of Incidental Findings on Whole-Body Imaging Reports

Key Points

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

ONCO-RADS–guided Large Language Models for Extraction and Classification of Incidental Findings on Whole-Body Imaging Reports

Key Points

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading