EUVIMEDScopri EUVIMED
Open access

Privacy-Preserving Framework for Multi-Institutional Medical Time-Series Analysis via Homomorphic Encryption: Design and Development Study

Yao Lu iD, Yu Tian iD, Tianshu Zhou iD, Jing Xia iD, Zhensong Zhao iD, Huiyao Sun iD, Kangli Liu iD, Jingsong Li iD

JMIR Formative Research · 2026-08-07 · 10 · e87306-e87306

DOI: 10.2196/87306 ↗

At a glance

Journal article in JMIR Formative Research, published online on 2026-08-07. Written by Yao Lu and 7 co-authors. Institution: Research Center for Scientific Data Hub, Zhejiang Lab, Hangzhou, Zhejiang, China. Major MEDLINE topics: Computer Security, Confidentiality, Privacy. Cited 1 times so far (Crossref); 71 references. The full text is freely available in PubMed Central.

AbstractSubjectsDati bibliograficiCiteRelated studies

Abstract (original language)

Background

The development of robust medical AI for knowledge discovery and decision support commonly necessitates large-scale datasets from multiple institutions. However, such data aggregation is severely constrained by privacy regulations and the inherent risk of sensitive information leakage, making it difficult to navigate the utility-privacy trade-off.

Objective

We aimed to design a secure multiparty deep learning system that enables privacy-preserving modeling from distributed medical time-series data without centralizing raw information or exposing model parameters. Our goal was to achieve predictive accuracy comparable to nonsecure models while providing strong security and efficiency.

Methods

We developed a framework using threshold homomorphic encryption to securely train recurrent neural networks on distributed longitudinal data. To improve the efficiency, we proposed an optimized encrypted matrix multiplication scheme, a secure ciphertext refresh protocol, and used lightweight encryption parameters and low-degree approximated activation polynomials. The system was evaluated on 4 real-world intensive care unit datasets for tasks like mortality and sepsis prediction.

Results

The system demonstrated practical efficiency, requiring approximately 1 minute per training iteration for processing 125 local batches over 39 variables and 48 time steps, and scaling well with data size and participant number. Securely trained models achieved predictive performance that was comparable to, and in some cases superior to, nonsecure centralized models, highlighting their ability to learn generalizable patterns in different unseen data distributions. For example, on the PhysioNet Challenge 2012 dataset, our secure model achieved an area under the curve (AUC) of 0.8480, outperforming the nonsecure baseline AUC of 0.8404.

Conclusions

This work provides a viable and efficient solution for cross-institutional, privacy-preserving analysis of longitudinal medical data. The framework successfully bridges the utility-privacy gap, facilitating safer collaborative research and enabling robust knowledge discovery and decision support while adhering to strict data protection standards.

Subjects

MeSH terms (MEDLINE)

Computer SecurityConfidentialityPrivacyHumansRecurrent Neural Networks

Author keywords

homomorphic encryptionmedical time-series analysispatient data privacyprivacy-preserving learningrecurrent neural networkssecure multiparty computation

Authors and institutions

  • Research Center for Scientific Data Hub, Zhejiang Lab, Hangzhou, Zhejiang, China.
  • Engineering Research Center of EMR and Intelligent Expert System, Key Laboratory for Biomedical Engineering of the Ministry of Education, College of Biomedical Engineering and Instrument Science, Zhejiang University, 38…
  • School of Intelligent Science and Technology, Hangzhou Institute for Advanced Study, University of Chinese Academy of Sciences, Hangzhou, Zhejiang, China.

ORCID profiles: Yao Lu, Yu Tian, Tianshu Zhou, Jing Xia, Zhensong Zhao, Huiyao Sun, Kangli Liu, Jingsong Li

Cite

APA

Lu, Y., Tian, Y., Zhou, T., Xia, J., Zhao, Z., Sun, H., Liu, K., & Li, J. (2026). Privacy-Preserving Framework for Multi-Institutional Medical Time-Series Analysis via Homomorphic Encryption: Design and Development Study. JMIR Formative Research, 10, e87306-e87306. https://doi.org/10.2196/87306

Vancouver

Lu Y, Tian Y, Zhou T, Xia J, Zhao Z, Sun H, et al. Privacy-Preserving Framework for Multi-Institutional Medical Time-Series Analysis via Homomorphic Encryption: Design and Development Study. JMIR Formative Research. 2026;10:e87306-e87306. doi:10.2196/87306

BibTeX
@article{Lu2026,
  author = {Lu, Yao and Tian, Yu and Zhou, Tianshu and Xia, Jing and Zhao, Zhensong and Sun, Huiyao and Liu, Kangli and Li, Jingsong},
  title = {Privacy-Preserving Framework for Multi-Institutional Medical Time-Series Analysis via Homomorphic Encryption: Design and Development Study},
  journal = {JMIR Formative Research},
  year = {2026},
  volume = {10},
  pages = {e87306--e87306},
  doi = {10.2196/87306}
}
Start for free →