Multimodal GPT model for assisting thyroid nodule diagnosis and management.

Journal: NPJ digital medicine
Published Date: May 3, 2025

Abstract

Although using artificial intelligence (AI) to analyze ultrasound images is a promising approach to assessing thyroid nodule risks, traditional AI models lack transparency and interpretability. We developed a multimodal generative pre-trained transformer for thyroid nodules (ThyGPT), aiming to provide a transparent and interpretable AI copilot model for thyroid nodule risk assessment and management. Ultrasound data from 59,406 patients across nine hospitals were retrospectively collected to train and test the model. After training, ThyGPT was found to assist in reducing biopsy rates by more than 40% without increasing missed diagnoses. In addition, it detects errors in ultrasound reports 1,610 times faster than humans. With the assistance of ThyGPT, the area under the curve for radiologists in assessing thyroid nodule risks improved from 0.805 to 0.908 (p < 0.001). As an AI-generated content-enhanced computer-aided diagnosis (AIGC-CAD) model, ThyGPT has the potential to revolutionize how radiologists use such tools.

Authors

  • Jincao Yao
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Yunpeng Wang
    College of Optical Science and Engineering, Zhejiang University, Hangzhou, China.
  • Zhikai Lei
    Department of Ultrasound, Zhejiang Provincial Hospital of Chinese Medicine, Hangzhou, China.
  • Kai Wang
    Department of Rheumatology, The Affiliated Huai'an No. 1 People's Hospital of Nanjing Medical University, Huai'an, Jiangsu, China.
  • Na Feng
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Fajin Dong
    Department of Ultrasound, The Second Clinical Medical College, Jinan University, Shenzhen, China.
  • Jianhua Zhou
    Department of Ultrasound, Sun Yat-sen University Cancer centre, State Key Laboratory of Oncology in South China, Collaborative Innovation centre for Cancer Medicine, Guangzhou, China.
  • Xiaoxian Li
    Department of Ultrasound, Sun Yat-sen University Cancer centre, State Key Laboratory of Oncology in South China, Collaborative Innovation centre for Cancer Medicine, Guangzhou, China.
  • Xiang Hao
    College of Optical Science and Engineering, Zhejiang University, Hangzhou, China.
  • Jiafei Shen
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Shanshan Zhao
    Department of Ultrasound, Shaoxing People's Hospital (Zhejiang University Shaoxing Hospital), Shaoxing, China.
  • Yuan Gao
    Engineering Research Center of EMR and Intelligent Expert System, Ministry of Education, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou Zhejiang Province, China.
  • Vicky Wang
    Wenling Medical Big Data and Artificial Intelligence Research Institute, Taizhou, China.
  • Di Ou
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Wei Li
    Department of Nephrology, The Second Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi, China.
  • Yidan Lu
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Liyu Chen
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Chen Yang
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Liping Wang
    School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai 200011, China.
  • Bojian Feng
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Yahan Zhou
    Wenling Medical Big Data and Artificial Intelligence Research Institute, Taizhou, China.
  • Chen Chen
    The George Institute for Global Health, Faculty of Medicine, University of New South Wales, Sydney, NSW, Australia.
  • Yuqi Yan
    Department of Diagnostic Ultrasound Imaging & Interventional Therapy, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, China.
  • Zhengping Wang
    Department of Ultrasound, The Affiliated Dongyang Hospital of Wenzhou Medical University, Dongyang, China.
  • Rongrong Ru
    Affiliated Xiaoshan Hospital, Hangzhou Normal University, Hangzhou, China.
  • Yaqing Chen
    Affiliated Xiaoshan Hospital, Hangzhou Normal University, Hangzhou, China.
  • Yanming Zhang
    Faculty of Applied Sciences, Macao Polytechnic University, Macao, China. zhangyanming_3@sina.com.
  • Ping Liang
    Department of Pharmacy, The Fourth Hospital of Hebei Medical University Shijiazhuang 050000, Hebei, China.
  • Dong Xu
    Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO, USA.

Keywords

No keywords available for this article.

External Resources

Popular Topics
Recent Journals