跳到正文
原文
Nature 健康护理主题·· 16 天前AI 评分50

MedGenAgent:用于肿瘤CT基础模型预训练的闭合环路代理工具

MedGenAgent: a closed-loop agent for oncology CT foundation model pretraining

AI 导读

该研究提出了一种名为MedGenAgent的闭合环路代理工具,用于在数据受限环境中预训练肿瘤CT基础模型。通过合成数据集和自动化质量控制,该工具在三个下游任务中实现了与全数据基线相当的性能。

正文

npj Precision Oncology (2026) Cite this article

We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.

Abstract

Oncology computed tomography (CT) foundation models require large, high-quality training datasets. However, the collection and annotation of such data are costly and time-consuming because tumors occur across multiple anatomical regions and have complex three-dimensional (3D) structures. Strict privacy constraints and institutional variation further limit the availability of large-scale oncology CT datasets. Here, we propose MedGenAgent, a closed-loop agent that supports oncology CT foundation model pretraining in data-constrained settings by iteratively constructing a lesion-rich, anatomically balanced 3D pretraining dataset. MedGenAgent incorporates a Specification resolver tool for structured task parsing and a central Orchestrator tool to coordinate the entire closed-loop agent workflow by leveraging both an Execution layer (including the MAISI synthesis model node for 3D oncology CT sample synthesis, the MedGemma vision-language model node for automated quality control, and the foundation model training node) and a Memory store tool for iterative, feedback-driven updates. Through the synergy of these tools, MedGenAgent synthesized 11,159 CT samples and retained 11,000 quality-controlled samples. By first leveraging this synthetic dataset for pretraining and then aligning with real-world data using only 4400 DeepLesion samples, representing 38.37% of the data used for baseline pretraining, the MedGenAgent-pretrained foundation model achieved performance comparable to the full-data baseline across three retrospective downstream tasks: lesion anatomical site classification, nodule malignancy classification, and cross-institutional NSCLC overall survival classification. Furthermore, automated quality control, independent physician assessment, and workflow analysis demonstrated the construction of a quality-controlled synthetic CT dataset while revealing regional variation in sample quality. Together, these results show that MedGenAgent can autonomously construct quality-controlled synthetic CT data that support the competitive performance of oncology CT foundation models when annotated real-world data are limited.

Acknowledgements

We sincerely thank the clinicians at Sun Yat-sen Memorial Hospital, for their valuable time and clinical expertise in assessing the quality of the medical imaging data used in this study.

Funding

This work was supported in part by the National Natural Science Foundation of China under Grants U24A20241, 62271456, 62672381, 62671458, 62676375, U24A20241, and 62132013, the Fundamental Research Funds for the Central Universities under Grants QTZX26142 and QTZX25085, the Key Research and Development Programs of Shaanxi under Grant 2025SYS-SYSZD-080, the Guangdong Basic and Applied Basic Research Foundation under Grant 2025A1515012741, and the Industrial Internet Innovation and Development Project under Grant ZC25T320059.

Author information

Authors and Affiliations

  1. State Key Lab of Integrated Service Networks, Xidian University, Xi’an, China

    Chengxuan Li & Qingqi Pei

  2. Shaanxi Key Laboratory of Blockchain and Secure Computing, Xidian University, Xi’an, China

    Chengxuan Li & Qingqi Pei

  3. Department of Emergency Medicine, Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Guangzhou, China

    Yunfeng Xu & Tao Yu

  4. Department of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China

    Zhenwei Huang, Xiaodong Huang & Changdong Wang

  5. Department of Radiology, Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Guangzhou, China

    Zehong Yang

  6. The Hong Kong Polytechnic University, Hong Kong, China

    Jinzhao Chu

Authors

  1. Chengxuan Li
  2. Yunfeng Xu
  3. Zhenwei Huang
  4. Zehong Yang
  5. Xiaodong Huang
  6. Jinzhao Chu
  7. Qingqi Pei
  8. Changdong Wang
  9. Tao Yu

Corresponding authors

Correspondence to Qingqi Pei, Changdong Wang or Tao Yu.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, C., Xu, Y., Huang, Z. et al. MedGenAgent: a closed-loop agent for oncology CT foundation model pretraining. npj Precis. Onc. (2026). https://doi.org/10.1038/s41698-026-01743-0

Download citation

  • Received: 14 May 2026

  • Accepted: 28 September 2026

  • Published: 04 November 2026

  • DOI: https://doi.org/10.1038/s41698-026-01743-0

来源:Nature 健康护理主题 · nature.com