首页> Clinical imaging> Performance of open-source and proprietary large language models in generating patient-friendly radiology chest CT reports.

Performance of open-source and proprietary large language models in generating patient-friendly radiology chest CT reports.

导出 LinkOut

翻译打印收藏纠错

第一作者： Philipp,Prucker

第一单位： Department of Diagnostic and Interventional Radiology, Technical University Munich, Ismaninger Str. 22, 81675 Munich, Germany.

作者： Philipp,Prucker ^[1] ; Felix,Busch ^[1] ; Felix,Dorfner ^[2] ; Christian J,Mertens ^[1] ; Nadine,Bayerl ^[3] ; Marcus R,Makowski ^[1] ; Keno K,Bressem ^[4] ; Lisa C,Adams ^[5]

作者单位： Department of Diagnostic and Interventional Radiology, Technical University Munich, Ismaninger Str. 22, 81675 Munich, Germany. ^[1] Charité - Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Department of Radiology, Charitéplatz 1, 10117 Berlin, Germany. ^[2] Institute of Radiology, Friedrich-Alexander-Universität Erlangen-Nürnberg and Uniklinikum Erlangen, Erlangen, Germany. ^[3] Department of Diagnostic and Interventional Radiology, Technical University Munich, Ismaninger Str. 22, 81675 Munich, Germany; Department of Radiology and Nuclear Medicine, German Heart Center Munich, Lazarettstraße 36, 80636 Munich, Germany. ^[4] Department of Diagnostic and Interventional Radiology, Technical University Munich, Ismaninger Str. 22, 81675 Munich, Germany. Electronic address: lisa.adams@tum.de. ^[5]

关键词 Artificial intelligence Chest CT GPT-4 Gemini-Ultra Large language models Llama-3 Patient communication Radiology reporting Readability metrics

医学主题词人类(Humans);体层摄影术, X线计算机(Tomography, X-Ray Computed);放射摄影术, 胸部(Radiography, Thoracic);结果可重复性(Reproducibility of Results);理解力(Comprehension)

DOI 10.1016/j.clinimag.2025.110557

PMID 40639135

发布时间 2025-08-25

翻译满意度评价：

准确一般字义偏差语意偏离译文作用小提交