RadiFlow：AI Agents重构放射科工作流程

张濛濛; 丛福泽; 王静; 张慧; 李娟娟; 倪清桦; 丁炫婷; 田永林; 吕宜生; 薛华丹; 王飞跃

doi:10.11959/j.issn.2096-6652.202535

您当前的位置：

首页 >

文章列表页 >

RadiFlow：AI Agents重构放射科工作流程

学术论文 | 更新时间：2025-11-18

- RadiFlow：AI Agents重构放射科工作流程
- RadiFlow: AI agents reconstructing radiology workflow
- 智能科学与技术学报 2025年7卷第3期页码：396-407
- 作者机构：
  
  1.中国科学院自动化研究所多模态人工智能系统全国重点实验室，北京 100190
  2.中国科学院大学人工智能学院，北京 100049
  3.中国医学科学院北京协和医院放射科，北京 100730
  4.澳门科技大学创新工程学院，澳门 999078
  5.北京交通大学计算机科学与技术学院，北京 100044
  6.清华大学信息科学技术学院，北京 100084
  7.中国科学院自动化研究所复杂系统管理与控制国家重点实验室，北京 100190
- 作者简介：
  
  [ "张濛濛（2000- ），女，中国科学院自动化研究所多模态人工智能系统全国重点实验室博士生，主要研究方向为通用医学影像分割模型、合成数据辅助分割、基于医学影像进行诊断与推理的视觉语言模型等。" ]
  [ "丛福泽（1993- ），男，中国医学科学院北京协和医院放射科科研博士后，主要研究方向为智能医学影像处理、医学视觉问答。" ]
  [ "王静（1992- ），男，中国科学院自动化研究所多模态人工智能系统全国重点实验室工程师，澳门科技大学创新工程学院博士生，主要研究方向为平行医疗理论与方法、智慧医疗系统的构建与应用。" ]
  [ "张慧（1993- ），女，北京交通大学计算机科学与技术学院副教授，主要研究方向为多智能体协同、多模态感知、具身智能。" ]
  [ "李娟娟（1986- ），女，博士，中国科学院自动化研究所多模态人工智能系统全国重点实验室副研究员，主要研究方向为区块链治理、DAO与决策智能。" ]
  [ "倪清桦（1999- ），女，澳门科技大学创新工程学院工程科学系博士生，主要研究方向为平行戏剧、社会计算、智能决策。" ]
  [ "丁炫婷（2006- ），女，清华大学信息科学技术学院在读，主要研究方向为数字生物与智能健康。" ]
  [ "田永林（1994- ），男，博士，中国科学院自动化研究所多模态人工智能系统全国重点实验室助理研究员，主要研究方向为平行智能、自动驾驶、智能交通。" ]
  [ "吕宜生（1983- ），男，中国科学院自动化研究所多模态人工智能系统全国重点实验室研究员，主要研究方向为人工智能、机器学习、深度学习、智能驾驶、智能交通和交通大数据。" ]
  [ "薛华丹（1979- ），女，博士，中国医学科学院北京协和医院放射科副主任、主任医生、教授、博士生导师，主要研究方向为腹盆影像诊断。" ]
  [ "王飞跃（1961- ），男，博士，中国科学院自动化研究所复杂系统管理与控制国家重点实验室主任，澳门科技大学特聘教授，主要研究方向为平行系统的方法与应用、社会计算、平行智能、知识自动化。" ]
- 基金信息：
  
  国家自然科学基金项目(82372051);澳门特别行政区科学与技术发展基金(0093/2023/RIA2;0145/2023/RIA3;0157/2024/RIA2);北京市自然科学基金海淀联合基金(L222099);首都卫生发展科研专项(2024-2-4015);四川省科技厅重点研发计划项目(2024YFHZ0011)
- DOI：10.11959/j.issn.2096-6652.202535
  中图分类号： TP391
- 收稿：2025-08-28，
  
  修回：2025-09-09，
  
  录用：2025-09-12，
  
  纸质出版：2025-09-15
- 稿件说明：
移动端阅览
张濛濛,丛福泽,王静等.RadiFlow：AI Agents重构放射科工作流程[J].智能科学与技术学报,2025,07(03):396-407.

ZHANG Mengmeng,CONG Fuze,WANG Jing,et al.RadiFlow: AI agents reconstructing radiology workflow[J].Chinese Journal of Intelligent Science and Technology,2025,07(03):396-407.
张濛濛,丛福泽,王静等.RadiFlow：AI Agents重构放射科工作流程[J].智能科学与技术学报,2025,07(03):396-407. DOI： 10.11959/j.issn.2096-6652.202535.

ZHANG Mengmeng,CONG Fuze,WANG Jing,et al.RadiFlow: AI agents reconstructing radiology workflow[J].Chinese Journal of Intelligent Science and Technology,2025,07(03):396-407. DOI： 10.11959/j.issn.2096-6652.202535.

摘要

随着现代医学的飞速发展和诊疗需求的日益增长，放射科作为临床诊断的核心部门，工作流程面临着多重挑战。这些挑战突出体现在多设备区域的复杂患者调度、人工操作导致的流程效率瓶颈、诊断辅助缺乏多模态临床信息综合推理能力以及报告撰写耗时等方面。提出了基于代理智能（agentic AI）的放射科全流程智能系统RadiFlow，构建了预约与患者管理智能体、影像采集协同智能体、分析与诊断推理智能体以及报告生成与解读智能体。通过智能体的协同，系统能够高效处理复杂患者调度与跨区域设备分流、实现影像采集标准化与智能容错、进行精准辅助诊断、自动化生成规范化报告并进行解读。本文为初步总结报告，仅通过案例研究，验证RadiFlow在提升放射科效率、诊断准确性、患者满意度及减轻医护负担方面展现出显著潜力，后续将基于此框架构建放射科多智能体系统，在实际场景中应用验证，以期为构建更智能、高效的放射科提供创新思路。

Abstract

With the rapid advancement of modern medicine and the increasing demand for clinical diagnosis and treatment

radiology

as a core department of clinical diagnostics

faces multiple challenges. These are particularly reflected in the complexity of patient scheduling across multi-device environments

workflow inefficiencies caused by manual operations

the lack of multimodal clinical information integration for diagnostic reasoning

and the time-consuming nature of report writing. To address these issues

RadiFlow

an end-to-end intelligent radiology workflow system based on Agentic AI

was proposed. The system was composed of four cooperative agents: scheduling agents

radiological examination agents

diagnostic reasoning agents

and report generation and interpretation agents. Through agents collaboration

RadiFlow can efficiently handle complex patient scheduling and cross-regional equipment allocation

achieve standardized and fault-tolerant imaging acquisition

provide accurate diagnostic assistance

and automatically generate and interpret standardized reports. However

this work remains a preliminary research report

validated through case studies. RadiFlow demonstrates significant potential for improving radiology efficiency

diagnostic accuracy

patient satisfaction

and reducing the workload of healthcare professionals. Future work will extend this framework to build a multi-agent radiology system for real-world deployment and validation

aiming to provide innovative insights into developing a more intelligent and efficient radiology practice.

关键词

Keywords

references

RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer International Publishing, 2015: 234-241.

ISENSEE F, JÄGER P F, KOHL S A A, et al. Automated design of deep learning methods for biomedical image segmentation[J]. arXiv preprint, 2019, arXiv: 1904.08128.

MA J, YANG Z, KIM S, et al. MedSAM2: segment anything in 3D medical images and videos[J]. arXiv preprint, 2025, arXiv: 2504.03600.

TADAVARTHI Y, MAKEEVA V, WAGSTAFF W, et al. Overview of noninterpretive artificial intelligence models for safety, quality, workflow, and education applications in radiology practice[J]. Radiology Artificial Intelligence, 2022, 4(2): e210114.

JING A B, GARG N, ZHANG J J, et al. AI solutions to the radiology workforce shortage[J]. NPJ Health Systems, 2025, 2(20): 1-6.

SCHNEIDER J. Generative to agentic AI: survey, conceptualization, and challenges[J]. arXiv preprint, 2025, arXiv: 2504.18875.

GRIDACH M, NANAVATI J, ABIDINE K Z E, et al. Agentic AI for scientific discovery: a survey of progress, challenges, and future directions[J]. arXiv preprint, 2025, arXiv: 2503.08979.

ACHARYA D B, KUPPAN K, DIVYA B. Agentic AI: autonomous intelligence for complex goals: a comprehensive survey[J]. IEEE Access, 2025, 13: 18912-18936.

FERBER D, EL NAHHAS O S M, WÖLFLEIN G, et al. Autonomous artificial intelligence agents for clinical decision making in oncology[J]. arXiv preprint, 2024, arXiv: 2404.04667.

YILDIRIM N, RICHARDSON H, WETSCHEREK M T, et al. Multimodal healthcare AI: identifying and designing clinically relevant vision-language applications for radiology[C]//Proceedings of the CHI Conference on Human Factors in Computing Systems. New York: ACM, 2024: 1-22.

JIN D, PAN E, OUFATTOLE N, et al. What disease does this patient have? a large-scale open domain question answering dataset from medical exams[J]. Applied Sciences, 2021, 11(14): 6421.

FENG J, ZHENG Q, WU C, et al. M3Builder: a multi-agent system for automated machine learning in medical imaging[J]. arXiv preprint, 2025, arXiv: 2502.20301.

BIZZO B C, ALMEIDA R R, ALKASAB T K. Artificial intelligence enabling radiology reporting[J]. Radiologic Clinics of North America, 2021, 59(6): 1045-1052.

DO H M, SPEAR L G, NIKPANAH M, et al. Augmented radiologist workflow improves report value and saves time: a potential model for implementation of artificial intelligence[J]. Academic Radiology, 2020, 27(1): 96-105.

YI Z, XIAO T, ALBERT M V. A multimodal multi-agent framework for radiology report generation[J]. arXiv preprint, 2025, arXiv: 2505.09787.

KOCAK B, PONSIGLIONE A, ROMEO V, et al. Radiology AI and sustainability paradox: environmental, economic, and social dimensions[J]. Insights into Imaging, 2025, 16(1): 88.

PANDEY H, AMOD A. Advancing healthcare automation: multi-agent system for medical necessity justification[J]. arXiv preprint, 2024, arXiv: 2404.17977.

CHEN Y J, ALBARQAWI A, CHEN C S. Reinforcing clinical decision support through multi-agent systems and ethical AI governance[J]. arXiv preprint, 2025, arXiv: 2504.03699.

ZHAO H X, XU Z Y, CHEN L, et al. Large-scale pretrained frame generative model enables real-time low-dose DSA imaging: an AI system development and multi-center validation study[J]. Med, 2025, 6(1): 100497.

ZHANG J D, SUN K C, YANG J W, et al. A generalized dual-domain generative framework with hierarchical consistency for medical image reconstruction and synthesis[J]. Communications Engineering, 2023, 2: 72.

HOSSAIN M I, ZAMZMI G, MOUTON P R, et al. Explainable AI for medical data: current methods, limitations, and future directions[J]. ACM Computing Surveys, 2025, 57(6): 1-46.

RASHIDI H H, PANTANOWITZ J, CHAMANZAR A, et al. Generative artificial intelligence in pathology and medicine: a deeper dive[J]. Modern Pathology, 2025, 38(4): 100687.

ZHANG J D, CUI Z M, SHI Z W, et al. A robust and efficient AI assistant for breast tumor segmentation from DCE-MRI via a spatial-temporal framework[J]. Patterns, 2023, 4(9): 100826.

ALMANSOORI M, KUMAR K, CHOLAKKAL H. Self-evolving multi-agent simulations for realistic clinical interactions[J]. arXiv preprint, 2025, arXiv: 2503.22678.

MA J, HE Y T, LI F F, et al. Segment anything in medical images[J]. Nature Communications, 2024, 15: 654.

FALLAHPOUR A, MA J, MUNIM A, et al. MedRAX: medical reasoning agent for chest X-ray[J]. arXiv preprint, 2025, arXiv: 2502.02673.

TANNO R, BARRETT D G T, SELLERGREN A, et al. Collaboration between clinicians and vision-language models in radiology report generation[J]. Nature Medicine, 2025, 31(2): 599-608.

LAI Y, ZHONG J, LI M, et al. Med-r1: Reinforcement learning for generalizable medical reasoning in vision-language models[J]. arXiv preprint, 2025, arXiv: 2503.13939.

KAPADNIS M N, PATNAIK S, NANDY A, et al. SERPENT-VLM: self-refining radiology report generation using vision language models[J]. arXiv preprint, 2024, arXiv: 2404.17912.

DOSHI R, AMIN K S, KHOSLA P, et al. Quantitative evaluation of large language models to streamline radiology report impressions: a multimodal retrospective analysis[J]. Radiology, 2024, 310(3): e231593.

LIU H, DING N, LI X Y, et al. Artificial intelligence and radiologist burnout[J]. JAMA Network Open, 2024, 7(11): e2448714.

SIMON B D, OZYORUK K B, GELIKMAN D G, et al. The future of multimodal artificial intelligence models for integrating imaging and clinical metadata: a narrative review[J]. Diagnostic and Interventional Radiology, 2025, 31(4): 303-312.

张志成, 王静, 张阳, 等. OrthoGPT: 面向精准诊疗的多模态骨科大模型[J]. 智能科学与技术学报, 2024, 6(3): 338-346.

ZHANG Z C, WANG J, ZHANG Y, et al. OrthoGPT: multimodal generative pre-trained transformer models for precise diagnosis and treatment of orthopedics[J]. Chinese Journal of Intelligent Science and Technology, 2024, 6(3): 338-346.

林飞, 王飞跃, 田永林, 等. 平行药物系统: 基于大语言模型和三类人的框架与方法[J]. 智能科学与技术学报, 2024, 6(1): 88-99.

LIN F, WANG F Y, TIAN Y L, et al. Parallel drug systems: framework and methods based on large language models and three types of humans[J]. Chinese Journal of Intelligent Science and Technology, 2024, 6(1): 88-99.

王飞跃, 金征宇, 苟超, 等. 基于ACP方法的平行医学图像智能分析及其应用[J]. 中华放射学杂志, 2021, 55(3): 309-315.

WANG F Y, JIN Z Y, GOU C, et al. ACP-based parallel medical imaging for intelligent analytics and applications[J]. Chinese Journal of Radiology, 2021, 55(3): 309-315.

赵毅飞, 申乐, 叶佩军, 等. 平行麻醉: 从麻醉自动化走向智慧型全周期麻醉平台[J]. 智能科学与技术学报, 2023, 5(2): 234-246.

ZHAO Y F, SHEN L, YE P J, et al. Parallel anesthesia: from anesthesia automation to intelligent full-cycle anesthesia platform[J]. Chinese Journal of Intelligent Science and Technology, 2023, 5(2): 234-246.

WANG J, WANG X, GUO Y Y, et al. A parallel medical diagnostic and treatment system for chronic diseases[C]//Proceedings of the 2020 Chinese Automation Congress (CAC). Piscataway: IEEE Press, 2020: 7412-7416.

NI Q H, VIKTÓRIA B, LIN F, et al. Parallel theaters in CPSS: from shadows of ISDOS to intelligence of decision theaters[J]. IEEE/CAA Journal of Automatica Sinica, 2025, 12(6): 1059-1062.

李娟娟, 秦蕊, 丁文文, 等. 基于Web3的去中心化自治组织与运营新框架[J]. 自动化学报, 2023, 49(5): 985-998.

LI J J, QIN R, DING W W, et al. A new framework for Web3-powered decentralized autonomous organizations and operations[J]. Acta Automatica Sinica, 2023, 49(5): 985-998.

浏览量

256

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

智能的本质与意义

LinYun：林业领域智能问答与决策支持的大型语言模型研究

智能时代的“呼吸主权”：从“创世纪计划”观全球科技竞争逻辑

大模型驱动的社交智能体谣言易感性与干预策略研究

开源社区中项目热度阶段预测方法研究——以大语言模型项目为例