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

Journal: IEEE transactions on bio-medical engineering
Published Date: Mar 1, 2026

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.

Authors

  • Tianyang Zhong
    School of Automation, Northwestern Polytechnic University, Xi'an, 710072, China.
  • Wei Zhao
    Key Laboratory of Synthetic and Biological Colloids, Ministry of Education, Jiangnan University, Wuxi 214122, Jiangsu Province, P. R. China. [email protected] [email protected].
  • Yutong Zhang
    Department of Biochemistry and Molecular Biology, Harbin Medical University, Harbin 150081, China.
  • Yi Pan
    Department of Neurosis and Psychosomatic Diseases, Huzhou Third Municipal Hospital, The Affiliated Hospital of Huzhou University, Huzhou, Zhejiang, China.
  • Peixin Dong
  • Zuowei Jiang
    School of Automation, Northwestern Polytechnical University, Xi'an 710072, China.
  • Hanqi Jiang
  • Yifan Zhou
    Department of Pharmacology, University of Oxford, Oxford, United Kingdom.
  • Xiaoyan Kui
    School of Computer Science and Engineering, Central South University, Changsha, Hunan, China.
  • Youlan Shang
  • Lin Zhao
    c Key Laboratory of Birth Defects and Related Diseases of Women and Children (Ministry of Education) , West China Second University Hospital Sichuan University , Chengdu , China.
  • Li Yang
    Department of Pharmacy, Beijing Tiantan Hospital, Capital Medical University, Beijing, China.
  • Yaonai Wei
  • Zhuoyi Li
  • Jiadong Zhang
  • Longtao Yang
  • Hao Chen
    The First School of Medicine, Wenzhou Medical University, Wenzhou, China.
  • Huan Zhao
    State Key Laboratory of Digital Manufacturing Equipment and Technology, Huazhong University of Science and Technology, Wuhan 430074, China.
  • Yuxiao Liu
    School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China.
  • Ning Zhu
    1 Google, Santa Clara County, CA, USA.
  • Yiwei Li
    New Cornerstone Science Laboratory, SEU-ALLEN Joint Center, Institute for Brain and Intelligence, Southeast University, Nanjing, Jiangsu 210096, China.
  • Yisong Wang
  • Jiaqi Yao
  • Jiaqi Wang
  • Ying Zeng
    Tongji University School of Medicine, Tongji University, Shanghai, China.
  • Lei He
    Guangxi Medical University, Nanning 530021; State Key Laboratory of Pathogen and Biosecurity, Beijing 100071, China.
  • Chao Zheng
    School of Biomedical Engineering, Southern Medical University, Guangzhou, Guangdong 510515 People's Republic of China.
  • Zhixue Zhang
  • Ming Li
    Radiology Department, Huadong Hospital, Affiliated with Fudan University, Shanghai, China.
  • Zhengliang Liu
    School of Computing, University of Georgia, Athens, GA, United States.
  • Haixing Dai
    School of Computing, University of Georgia, Athens, GA, United States.
  • Zihao Wu
    School of Computing, University of Georgia, Athens, GA, United States.
  • Lu Zhang
    Department of Computer Science and Engineering, The University of Texas at Arlington, Arlington, TX, United States.
  • Shu Zhang
    State University of New York, Department of Radiology, Stony Brook, New York, United States.
  • Xiaoyan Cai
    School of Automation, Northwestern Polytechnical University, Xi'an, 710 072, China.
  • Xintao Hu
  • Shijie Zhao
  • Xi Jiang
  • Xin Zhang
    First Department of Infectious Diseases, The First Affiliated Hospital of China Medical University, Shenyang, China.
  • Wei Liu
    Department of Radiation Oncology, Mayo Clinic, Scottsdale, AZ, United States.
  • Xiang Li
    Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States.
  • Dajiang Zhu
    Department of Computer Science and Engineering, The University of Texas at Arlington, Arlington, TX, United States.
  • Lei Guo
    Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China.
  • Dinggang Shen
    School of Biomedical Engineering, ShanghaiTech University, Shanghai, China.
  • Junwei Han
  • Tianming Liu
    School of Computing, University of Georgia, Athens, GA, United States.
  • Jun Liu
    Department of Radiology, Second Xiangya Hospital, Changsha, Hunan, China.
  • Tuo Zhang
    Weill Cornell Medical College, 1300 York Avenue, New York, New York, 10065.

Keywords

No keywords available for this article.

External Resources

Popular Topics
Recent Journals