ChatRadio-Valuer: A Chat Large Language Model for Generalizable Radiology Impression Generation on Multi-Institution and Multi-System Data

Achieving clinical level performance and widespread deployment for generating radiology impressions encounters a giant challenge for conventional artificial intelligence models tailored to specific diseases and organs. Concurrent with the increasing accessibility of radiology reports and advancements in modern general AI techniques, the emergence and potential of deployable radiology AI exploration have been bolstered. Here, we present ChatRadio-Valuer, the first general radiology diagnosis large language model for localized deployment within hospitals and being close to clinical use for multi-institution and multi-system diseases. ChatRadio-Valuer achieved 15 state-of-the-art results across five human systems and six institutions in clinical-level events (n = 332,673) through rigorous and full-spectrum assessment, including engineering metrics, clinical validation, and efficiency evaluation. Notably, it exceeded OpenAI's GPT-3.5 and GPT-4 models, achieving superior performance in comprehensive disease diagnosis compared to the average level of radiology experts. Besides, ChatRadio-Valuer supports zero-shot transfer learning, greatly boosting its effectiveness as a radiology assistant, while ensuring adherence to privacy standards and being readily utilized for large-scale patient populations. Our expeditions suggest the development of localized LLMs would become an imperative avenue in hospital applications.

Authors

Institutions

Publication Details

Journal
IEEE Transactions on Biomedical Engineering
Published
2025-08-11
DOI
https://doi.org/10.1109/tbme.2025.3597325
Citations
1
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
article
Field-Weighted Citation Impact
0.65
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

ChatRadio-Valuer: A Chat Large Language Model for Generalizable Radiology Impression Generation on Multi-Institution and Multi-System Data

Peixin Dong, Zuowei Jiang, Youlan Shang, Jiaqi Yao et al.
1 citations
IEEE Transactions on Biomedical Engineering
Artificial Intelligence in Healthcare and Education
0.65
article

ChatRadio-Valuer: A Chat Large Language Model for Generalizable Radiology Impression Generation on Multi-Institution and Multi-System Data

Peixin Dong, Zuowei Jiang, Youlan Shang, Jiaqi Yao, Yaonai Wei, Haixing Dai, Zheng‐Xiang Li, Ying Zeng, Zhixue Zhang, Lei Guo, Zihao Wu, Xiaoyan Kui, Tianyang Zhong, Lei Guo, Ming Li, Jun Liu, Jun Liu, Jiaqi Wang, Shu Zhang, Lei He, Yiwei Li, Tianming Liu, Tuo Zhang, Chao Zheng, Hao Chen, Li Yang, Jiadong Zhang, Zhengliang Liu, Ning Zhu, Yisong Wang, Yifan Zhou, Zhao Lin, Zhao Huan, Yi Pan, Hanqi Jiang, Longtao Yang, Xin Zhang, Wei Zhao, Yuxiao Liu, Yutong Zhang
article en
1 citations

Abstract

Achieving clinical level performance and widespread deployment for generating radiology impressions encounters a giant challenge for conventional artificial intelligence models tailored to specific diseases and organs. Concurrent with the increasing accessibility of radiology reports and advancements in modern general AI techniques, the emergence and potential of deployable radiology AI exploration have been bolstered. Here, we present ChatRadio-Valuer, the first general radiology diagnosis large language model for localized deployment within hospitals and being close to clinical use for multi-institution and multi-system diseases. ChatRadio-Valuer achieved 15 state-of-the-art results across five human systems and six institutions in clinical-level events (n = 332,673) through rigorous and full-spectrum assessment, including engineering metrics, clinical validation, and efficiency evaluation. Notably, it exceeded OpenAI's GPT-3.5 and GPT-4 models, achieving superior performance in comprehensive disease diagnosis compared to the average level of radiology experts. Besides, ChatRadio-Valuer supports zero-shot transfer learning, greatly boosting its effectiveness as a radiology assistant, while ensuring adherence to privacy standards and being readily utilized for large-scale patient populations. Our expeditions suggest the development of localized LLMs would become an imperative avenue in hospital applications.

IEEE Transactions on Biomedical EngineeringVol. 73(3)
Harvard University (US), Central South University (CN), University of Electronic Science and Technology of China (CN), Xinjiang Medical University (CN), University of Georgia (US), Northwestern Polytechnical University (CN), Fudan University (CN), WinnMed (US), ShanghaiTech University (CN), Massachusetts General Hospital (US), Shanghai University of Traditional Chinese Medicine (CN), Huadong Hospital (CN), First Affiliated Hospital of Hunan University of Traditional Chinese Medicine (CN), The First People's Hospital of Changde (CN), Yueyang Second People's Hospital (CN), Yueyang Hospital (CN), Mayo Clinic in Florida (US), Second Xiangya Hospital of Central South University (CN), Second Affiliated Hospital of Xinjiang Medical University (CN)
Openalex Percentile: Top 23%
Artificial Intelligence in Healthcare and Education
0.65
DOISourceOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.