跳到正文
原文
Nature 健康护理主题·· 2026-10-08AI 评分44

深度学习模型 OCRiskNet 用于口腔潜在恶性疾患的时变恶性转化风险预测

Time-varying risk prediction for oral potentially malignant disorders using deep learning

AI 导读

研究团队开发并验证了深度学习模型 OCRiskNet,用于口腔潜在恶性疾患的个体化、时变恶性转化风险预测。该回顾性多中心研究基于2003—2024年香港队列开发与内部测试,口腔白斑和口腔苔藓样病变的 C 指数分别为0.9209和0.9207,并在英国和中国大陆独立队列中完成外部验证。模型还显示良好校准和临床效用

正文

Communications Medicine (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

Background

Patients with oral potentially malignant disorders require repeated risk assessment after diagnosis because malignant transformation risk varies across lesions and over time. Existing tools rarely provide individualized, interpretable, time-resolved estimates to guide surveillance.

Methods

A retrospective multicentre study was conducted with clinicopathologically confirmed oral leukoplakia or oral lichenoid conditions. Hong Kong cohorts from 2003 to 2024 were used for model development and internal testing, and independent United Kingdom and mainland China cohorts for external validation. OCRiskNet integrates demographic, exposure, clinical, treatment and histopathological data in a fixed deep survival pipeline combining probabilistic feature enrichment with an intrinsically interpretable free-knot Kolmogorov-Arnold network. Performance was evaluated using concordance index, area under the curve, Brier score, calibration and decision curve analysis. Interpretability was assessed using symbolic function extraction and feature attribution.

Results

Here we show that OCRiskNet provides strong internal discrimination for oral leukoplakia and oral lichenoid conditions, with concordance indices of 0.9209 and 0.9207. External validation supported model generalizability across geographically distinct cohorts. Interpretation analyses identified clinically meaningful contributors to risk. OCRiskNet also demonstrated favorable calibration and clinical utility, with clear stratification of risk groups over time. A web-based interface was further implemented to support practical use of individualized risk estimation in longitudinal management.

Conclusion

OCRiskNet provided individualized, time-resolved estimates of malignant transformation risk in OPMDs and demonstrated consistent performance across multicentre cohorts. This framework may support more transparent and risk-adapted surveillance planning in heterogeneous clinical settings, with potential public-health value by enabling risk-adapted surveillance of high-risk patients and more efficient allocation of clinical follow-up resources.

Plain Language Summary

Oral potentially malignant disorders (OPMDs) represent an important window for oral cancer prevention. These conditions may persist for years, with malignant transformation risk changing across the disease course, making long-term risk-adapted surveillance essential. Specialist oral medicine services remain limited in many regions, and general dental practitioners may face challenges in determining which patients require closer monitoring. We developed OCRiskNet using longitudinal real-world clinical and histopathological data from multicentre cohorts with extended follow-up. The model estimates how an individual patient’s future cancer risk changes over time using information routinely collected during care. This low-burden approach may help resource-limited settings prioritise surveillance for high-risk patients, use specialist services more efficiently, and strengthen oral cancer prevention.

Funding

D.Z. discloses support for the research of this work from the SZU-LU Joint Research Programme (2025014; SZU-LU009/2526), Lingnan Interdisciplinary & Strategic Research Grant (784009), Lingnan Faculty Research Grant (106125), and Lingnan Strategic Development Fund (785007). L.W.Z. discloses support from the Hong Kong Research Grants Council General Research Fund (17102526). O.S.L.discloses support from the National Natural Science Foundation of China (82470989) and Science and Technology Talent Support Project of the Hunan Association for Science and Technology (2023TJ-Z08).

Author information

Authors and Affiliations

  1. School of Data Science, Lingnan University, Hong Kong SAR, China

    Jing Wen Li & Dian Zhang

  2. Division of Oral & Maxillofacial Surgery, Faculty of Dentistry, University of Hong Kong, Hong Kong SAR, China

    Jing Wen Li, Jane Jing Ya Pu & Li Wu Zheng

  3. Department of Computer Science and Software Engineering, Shenzhen University, Shenzhen, China

    Zi Han Lin, Xiang Zhu, Liu Yang Zhao & Ya Fang Zhou

  4. Xiangya Stomatological Hospital and Xiangya School of Stomatology, Central South University, Changsha, Hunan, China

    Hui Feng & Ou Sheng Liu

  5. Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong SAR, China

    Hannah On Lap Yiu & Amy Bik Wan Chan

  6. Division of Applied Oral Sciences & Community Dental Care, Faculty of Dentistry, The University of Hong Kong, Hong Kong SAR, China

    John Adeoye & Colman McGrath

  7. College of Medicine and Dentistry, James Cook University, Townsville, Queensland, Australia

    Peter Thomson

Authors

  1. Jing Wen Li
  2. Zi Han Lin
  3. Xiang Zhu
  4. Liu Yang Zhao
  5. Ya Fang Zhou
  6. Hui Feng
  7. Hannah On Lap Yiu
  8. John Adeoye
  9. Jane Jing Ya Pu
  10. Amy Bik Wan Chan
  11. Ou Sheng Liu
  12. Peter Thomson
  13. Colman McGrath
  14. Dian Zhang
  15. Li Wu Zheng

Corresponding authors

Correspondence to Dian Zhang or Li Wu Zheng.

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, J.W., Lin, Z.H., Zhu, X. et al. Time-varying risk prediction for oral potentially malignant disorders using deep learning. Commun Med (2026). https://doi.org/10.1038/s43856-026-01951-6

Download citation

  • Received: 17 June 2026

  • Accepted: 22 September 2026

  • Published: 08 October 2026

  • DOI: https://doi.org/10.1038/s43856-026-01951-6

来源:Nature 健康护理主题 · nature.com