Large Language Models for Diagnosing Focal Liver Lesions From <scp>CT</scp>/<scp>MRI</scp> Reports: A Comparative Study With Radiologists
Two-step ChatGPT-4o showed matching accuracies to real-world radiology reports and junior radiologists for diagnosing FLLs but was less accurate than middle-level radiologists and demonstrated little incremental diagnostic value.
Authors
- Liuji Sheng
- Yanshu Wang (ORCID: https://orcid.org/0000-0001-8473-8165)
- Yingyi Wu (ORCID: https://orcid.org/0000-0002-6499-3211)
- Hanyu Jiang (ORCID: https://orcid.org/0000-0002-7726-1618)
- Feng Che
- Mustafa Rifaat Bashir (ORCID: https://orcid.org/0000-0001-8800-5057)
- Bin Song (ORCID: https://orcid.org/0000-0002-7269-2101)
- Chongtu Yang (ORCID: https://orcid.org/0000-0001-5384-9778)
- Maxime Ronot (ORCID: https://orcid.org/0000-0001-7464-3939)
- Hong Wei (ORCID: https://orcid.org/0000-0002-1561-1512)
- Yidi Chen (ORCID: https://orcid.org/0000-0001-6293-4624)
- Qin Qin (ORCID: https://orcid.org/0000-0002-8118-3253)
- Jingwen Peng (ORCID: https://orcid.org/0000-0002-4283-1721)
Institutions
- North Sichuan Medical University (CN)
- Guangxi Medical University (CN)
- Université Paris Cité (FR)
- Sichuan University (CN)
- Hôpital Beaujon (FR)
- West China Hospital of Sichuan University (CN)
- Assistance Publique – Hôpitaux de Paris (FR)
- First Affiliated Hospital of GuangXi Medical University (CN)
- Duke Medical Center (US)
- Sanya Central Hospital (CN)
Publication Details
- Journal
- Liver International
- Published
- 2025-05-10
- DOI
- https://doi.org/10.1111/liv.70115
- Citations
- 11
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 14.96
Funders
- National Natural Science Foundation of China
- Natural Science Foundation of Sichuan Province