Abstract
Digital subtraction angiography (DSA) devices guide procedures across numerous diseases, performed on more than 100,000 patients daily worldwide. However, these procedures expose patients and healthcare providers to radiation, increasing the risk of health issues. Despite many low-dose DSA imaging methods proposed, none have been prospectively clinically validated. In this study, 46,829 patients (over 5 million DSA images) from 70 centers were used to iterate our previously developed generative artificial intelligence system (named GenDSA-V2). A total of 1,068 patients (533 in intervention arm and 535 in control arm), with suspected cerebral aneurysms (n = 435), lung cancer (n = 417) or advanced liver cancer (n = 216), meeting surgical criteria, were enrolled to validate the GenDSA-V2. The primary outcome was radiation dose, while secondary outcomes included efficiency, operation time and intraoperative complications. Group assignments were blinded to patients, surgeons and investigators, while technicians were aware but not involved in data collection or analysis. The GenDSA-V2 group showed substantially reduced radiation exposure, with an air kerma (AK) of 151.3 ± 125.1 mGy compared to 457.4 ± 407.4 mGy in the standard clinical protocols (SCP) group (mean difference = −306.1 mGy, 95% confidence interval (CI) = −342.3 to −269.9, P < 0.001 for superiority) and a dose-area product (DAP) of 4009.7 ± 2767.9 μGy m2 versus 12531.6 ± 9145.9 μGy m2 (mean difference = −8521.9 μGy m2, 95% CI = −9333.1 to −7710.7, P < 0.001 for superiority). Mean operation time was 33.1 ± 10.8 min in the SCP group and 34.8 ± 11.8 min in the GenDSA-V2 group (mean difference = 1.7 min, 95% CI = 0.3 to 3.1, P < 0.001 for noninferiority). Complication rates were similar (SCP = 8.1%, GenDSA-V2 = 7.5%, mean difference = −0.6%, 95% CI = −3.8% to 2.6%, P < 0.001 for noninferiority). The GenDSA system reduces radiation exposure to both physicians and patients by approximately two-thirds during DSA-guided procedures, demonstrating substantial clinical and translational value. Chinese Clinical Trial Registry: ChiCTR2400084789.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$32.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$259.00 per year
only $21.58 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to the full article PDF.
USD 39.95
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
Supplementary Data 1 and 2 are publicly available (https://github.com/ZrH42/GenDSA_Data), and the remaining data are available upon request to the corresponding author (Huangxuan Zhao) within 30 days from the request, as long as they are in line with the ethics approvals. Source data are provided with this paper.
Code availability
The codes used for model construction are available at https://github.com/ZrH42/GenDSA-V2.
Change history
29 January 2026
In the version of the article initially published, affiliation 1 and 2 were mistakenly swapped around. The present address was also incorrect and should have read “School of Computer Science, Wuhan University, Wuhan, 430022, China.” These changes have now been corrected in the HTML and PDF versions of the article.
References
Richardson, D. B. et al. Cancer mortality after low dose exposure to ionising radiation in workers in France, the United Kingdom, and the United States (INWORKS): cohort study. BMJ 382, e074520 (2023).
El Ghissassi, F. et al. A review of human carcinogens—part D: radiation. Lancet Oncol. 10, 751–752 (2009).
Bosch de Basea Gomez, M. et al. Risk of hematological malignancies from CT radiation exposure in children, adolescents and young adults. Nat. Med. 29, 3111–3119 (2023).
Little, M. P. et al. Ionising radiation and cardiovascular disease: systematic review and meta-analysis. BMJ 380, e072924 (2023).
Chandra, R. A., Keane, F. K., Voncken, F. E. M. & Thomas, C. R. Contemporary radiotherapy: present and future. Lancet 398, 171–184 (2021).
Adam, A. & Kenny, L. M. Interventional oncology in multidisciplinary cancer treatment in the 21(st) century. Nat. Rev. Clin. Oncol. 12, 105–113 (2015).
Giri, J. et al. Interventional therapies for acute pulmonary embolism: current status and principles for the development of novel evidence: a scientific statement from the American Heart Association. Circulation 140, e774–e801 (2019).
Einstein, A. J. Medical imaging: the radiation issue. Nat. Rev. Cardiol. 6, 436–438 (2009).
Cai, Y., Wang, J., Yuille, A., Zhou, Z. & Wang, A. Structure-aware sparse-view X-ray 3d reconstruction. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 11174–11183 (IEEE, 2024).
Siow, T. Y., Ma, C.-Y. & Toh, C. H. Angular super-resolution in X-ray projection radiography using deep neural network: implementation on rotational angiography. Biomed. J. 46, 154–162 (2023).
Zhao, H. et al. Self-supervised learning enables 3D digital subtraction angiography reconstruction from ultra-sparse 2D projection views: a multicenter study. Cell Rep. Med. 3, 100775 (2022).
Ueda, D. et al. Deep learning-based angiogram generation model for cerebral angiography without misregistration artifacts. Radiology 299, 675–681 (2021).
Wang, L., Dou, Q., Fletcher, P. T., Speidel, S. & Li, S. (eds). NAF: neural attenuation fields for sparse-view CBCT reconstruction. Proc. Medical Image Computing and Computer Assisted Intervention—MICCAI 2022 442–452 (Springer, 2022).
Liu, Z. et al. Geometry-aware attenuation learning for sparse-view CBCT reconstruction. IEEE Trans. Med. Imaging 44, 1083–1097 (2025).
Greenspan, H. et al. (eds). Learning deep intensity field for extremely sparse-view CBCT reconstruction. Proc. Medical Image Computing and Computer Assisted Intervention—MICCAI 2023 13–23 (Springer, 2023).
De Bruijne, M. et al. (eds). Estimation of high framerate digital subtraction angiography sequences at low radiation dose. Proc. Medical Image Computing and Computer Assisted Intervention—MICCAI 2021 171–180 (Springer, 2021).
Tang, Z. et al. Radiation reduction for interventional radiology imaging: a video frame interpolation solution. Insights Imaging 15, 42 (2024).
OpenAI et al. GPT-4 technical report. Preprint at https://arxiv.org/abs/2303.08774 (2023).
Touvron, H. et al. Llama 2: open foundation and fine-tuned chat models. Preprint at https://arxiv.org/abs/2307.09288 (2023).
Brooks, T. et al. Video generation models as world simulators. OpenAI openai.com/index/video-generation-models-as-world-simulators/ (2024).
Zhao, H. et al. Large-scale pretrained frame generative model enables real-time low-dose DSA imaging: an AI system development and multi-center validation study. Med 6, 100497 (2025).
Pan, Y., Ye, Y., Zhang, Y., Xia, Y. & Shen, D. Draw sketch, draw flesh: whole-body computed tomography from any X-ray views. Int. J. Comput. Vis. 133, 2505–2526 (2024).
Brendlin, A. S. et al. Novel deep learning denoising enhances image quality and lowers radiation exposure in interventional bronchial artery embolization cone beam CT. Acad. Radiol. 31, 2144–2155 (2024).
Zhang, M., Gu, S. & Shi, Y. The use of deep learning methods in low-dose computed tomography image reconstruction: a systematic review. Complex Intell. Syst. 8, 5545–5561 (2022).
Xu, Z., Zhao, H., Liu, W. & Wang, X. GaraMoSt: parallel multi-granularity motion and structural modeling for efficient multi-frame interpolation in DSA images. In Proc. AAAI Conference on Artificial Intelligence 28530–28538 (AAAI, 2025).
Xu, Z. et al. MoSt-DSA: modeling motion and structural interactions for direct multi-frame interpolation in DSA images. In European Conference on Artificial Intelligence (2024).
Zucker, E. J., Sandino, C. M., Kino, A., Lai, P. & Vasanawala, S. S. Free-breathing accelerated cardiac mri using deep learning: validation in children and young adults. Radiology 300, 539–548 (2021).
Rastogi, A. et al. Deep-learning-based reconstruction of undersampled MRI to reduce scan times: a multicentre, retrospective, cohort study. Lancet Oncol. 25, 400–410 (2024).
Koetzier, L. R. et al. Deep learning image reconstruction for CT: technical principles and clinical prospects. Radiology 306, e221257 (2023).
Zhou, W., Bovik, A. C., Sheikh, H. R. & Simoncelli, E. P. Image quality assessment: from error visibility to structural similarity. IEEE Trans. Image Process. 13, 600–612 (2004).
Piaggio, G., Elbourne, D. R., Pocock, S. J., Evans, S. J. W. & Altman, D. G. Reporting of noninferiority and equivalence randomized trials: extension of the CONSORT 2010 statement. JAMA 308, 2594–2604 (2012).
Liu, X., Cruz Rivera, S., Moher, D., Calvert, M. J. & Denniston, A. K. Reporting guidelines for clinical trial reports for interventions involving artificial intelligence: the CONSORT-AI extension. Nat. Med. 26, 1364–1374 (2020).
CONSORT-AI and SPIRIT-AI Steering Group. Reporting guidelines for clinical trials evaluating artificial intelligence interventions are needed. Nat. Med. 25, 1467–1468 (2019).
Heidari, S., Babor, T. F., de Castro, P., Tort, S. & Curno, M. Sex and gender equity in research: rationale for the SAGER guidelines and recommended use. Res. Integr. Peer Rev. 1, 2 (2016).
Freidlin, B., Korn, E. L., George, S. L. & Gray, R. Randomized clinical trial design for assessing noninferiority when superiority is expected. J. Clin. Oncol. 25, 5019–5023 (2007).
Tannock, I. F. et al. The tyranny of non-inferiority trials. Lancet Oncol. 25, e520–e525 (2024).
Hendee, W. R. & Edwards, F. M. ALARA and an integrated approach to radiation protection. Semin. Nucl. Med. 16, 142–150 (1986).
Acknowledgements
This work was supported by the National Key Research and Development Program of China (grants 2023YFC2413500 and 2023YFC2705700) and the National Natural Science Foundation of China (grants 62225113, 82472070 and 82402411). We extend our sincere gratitude to D. Ouyang (Department of Cardiology, Division of AI in Medicine, Cedars-Sinai Medical Center) for his invaluable feedback during the experimental design process. We are also profoundly grateful to Xiaomi Corporation and Kingsoft Corporation for their support.
Author information
Authors and Affiliations
Contributions
H. Zhao, Y.B., L.C., C. Zhu, X.W., B.D. and C. Zheng had major involvement in study conception or design. J.M., Y. Lei, T.S., L.W., C.H., Z.M., L.L., H. Zhu, J.Z., L.F., Z.F., P.D., B.L., X.H. and X.X. had substantial involvement in data acquisition. Y.B., L.C., R.Z., L.Z. and B.W. had major involvement in data analysis. Z.X., X.L., Y. Li, Y.H., Y.F., X.K. and X.W. had major involvement in data interpretation. All authors had access to the data and participated in writing, reviewing and revising the paper. All authors approved the final version of the paper for publication. All authors accept responsibility for the accuracy and integrity of all aspects of the research.
Corresponding authors
Ethics declarations
Competing interests
X.L. and Y. Li are employees of the Institute of Research and Clinical Innovations, Neusoft Medical Systems. Y.H. serves as Vice President of Neusoft Medical Systems. The other authors declare no competing interests.
Peer review
Peer review information
Nature Medicine thanks Paola Berchialla, Gabriel Broocks, Rose Du and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Lorenzo Righetto, in collaboration with the Nature Medicine team.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Supplementary information
Supplementary Information (download PDF )
Supplementary Appendices 1–12, Tables 1–18 and Figs. 1–15.
Supplementary Video 1 (download MP4 )
3D cerebrovascular reconstruction in a case demonstrating the advantages of GenDSA-V2 under specific conditions.
Supplementary Video 2 (download MP4 )
Example of reconstructed images of 3D thorax–abdomen–CBCT.
Supplementary Video 3 (download MP4 )
Example of 2D cerebral angiography videos.
Supplementary Video 4 (download MP4 )
Example of celiac trunk angiography.
Source data
Source Data Fig. 3 (download XLSX )
Statistical source data.
Source Data Extended Data Table 1 (download XLSX )
Statistical source data.
Source Data Extended Data Table 2 (download XLSX )
Statistical source data.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhao, H., Bai, Y., Chen, L. et al. Generative AI-based low-dose digital subtraction angiography for intra-operative radiation dose reduction: a randomized controlled trial. Nat Med 32, 288–296 (2026). https://doi.org/10.1038/s41591-025-04042-6
Received:
Accepted:
Published:
Version of record:
Issue date:
DOI: https://doi.org/10.1038/s41591-025-04042-6
