2609005274
  • Open Access
  • Article

FHE-Assisted Privacy-Preserving Medical Image De-Identification

  • Jinxiang Dai * and 
  • Li S. Cang

Received: 17 July 2026 | Revised: 17 August 2026 | Accepted: 10 October 2026 | Published: 21 October 2026

Abstract

Medical image de-identification is essential for data sharing, but conventional OCR-based redaction can expose protected health information (PHI) to the service that performs redaction. We present a client-cloud framework that applies fully homomorphic encryption (FHE) to privacy-preserving medical image de-identification. The client localizes text, segments character patches, encrypts normalized inputs with CKKS, and performs final correction and masking, while the cloud evaluates an FHE-compatible VGG-style character recognizer and CharCNN PHI classifier over ciphertexts. The system reduces encrypted inference cost through polynomial activations, average pooling, diagonal encoding, Toeplitz-style convolution lowering, and BSGS rotation reuse. Prototype results show 100% plaintext-ciphertext prediction consistency, 83.15% corrected character accuracy, and 97.1% encrypted PHI-classification accuracy, with bootstrapped nonlinear layers remaining the main bottleneck.

References 

  • 1.

    Qin, L. Shaping the Future: Emerging Technologies and Their Role in Industry 4.0 and Beyond. J. Emerg. Technol. Ind. Appl. 2026, 1, 4. https://doi.org/10.53941/jetia.2026.100004.

  • 2.

    Li, J.; Fu, L.; Wang, Y. Generative AI and management analytics: Survey. J. Manag. Anal. 2026, 13, 217–263. https://doi.org/10.1080/23270012.2026.2652312.

  • 3.

    U.S. Department of Health and Human Services. Guidance Regarding Methods for De-identification of Protected Health Information in Accordance with the HIPAA Privacy Rule. Available online: https://www.hhs.gov/hipaa/for-professionals/privacy/special-topics/de-identification/ (accessed on 29 June 2026).

  • 4.

    NEMA. Digital Imaging and Communications in Medicine (DICOM), Part 15: Security and System Management Profiles. Available online: https://dicom.nema.org/medical/dicom/current/output/html/part15.html (accessed on 29 June 2026).

  • 5.

    European Parliament and Council of the European Union. Regulation (EU) 2016/679: General Data Protection Regulation. Available online: https://eur-lex.europa.eu/eli/reg/2016/679/oj (accessed on 29 June 2026).

  • 6.

    Standing Committee of the National People’s Congress. Personal Information Protection Law of the People’s Republic of China. Available online: https://en.npc.gov.cn.cdurl.cn/2021-12/29/c694559.htm (accessed on 29 June 2026).

  • 7.

    Rivest, R.L.; Adleman, L.; Dertouzos, M.L. On Data Banks and Privacy Homomorphisms. In Foundations of Secure Computation; Academic Press: New York, NY, USA, 1978; pp. 169–180.

  • 8.

    Gentry, C. Fully homomorphic encryption using ideal lattices. In Proceedings of the STOC ’09: Symposium on Theory of Computing, Bethesda, MD, USA, 31 May–2 June 2009; pp. 169–178. https://doi.org/10.1145/1536414.1536440.

  • 9.

    Cheon, J.H.; Kim, A.; Kim, M.; et al. Homomorphic Encryption for Arithmetic of Approximate Numbers. In Advances in Cryptology – ASIACRYPT 2017; Springer International Publishing: Cham, Switzerland, 2017; Volume 10624, pp. 409–437. https://doi.org/10.1007/978-3-319-70694-8 15.

  • 10.

    Clunie, D.A.; Flanders, A.; Taylor, A.; et al. Report of the Medical Image De-Identification (MIDI) Task Group—Best Practices and Recommendations. arXiv 2023, arXiv:2303.10473.

  • 11.

    Pei, L.; Sutton, G.; Rutherford, M.; et al. Medical Image De-Identification Benchmark Challenge. arXiv 2025, arXiv:2507.23608v1.

  • 12.

    Rempe, M.; Heine, L.; Seibold, C.; et al. De-Identification of Medical Imaging Data: A Comprehensive Tool for Ensuring Patient Privacy. arXiv 2024, arXiv:2410.12402.

  • 13.

    Naddeo, K.; Koutsoubis, N.; Krish, R.; et al. DICOM De-Identification via Hybrid AI and Rule-Based Framework for Scalable, Uncertainty-Aware Redaction. arXiv 2025, arXiv:2507.23736v1.

  • 14.

    Jiang, H.; Xie, S.; Wan, Z. A DICOM Image De-identification Algorithm in the MIDI-B Challenge. arXiv 2025, arXiv:2508.07538.

  • 15.

    Du, Y.; Li, C.; Guo, R.; et al. PP-OCR: A Practical Ultra Lightweight OCR System. arXiv 2020, arXiv:2009.09941.

  • 16.

    de Campos, T.E.; Babu, B.R.; Varma, M. Character Recognition in Natural Images. In Proceedings of the International Conference on Computer Vision Theory and Applications, Lisboa, Portugal, 5–8 February 2009; pp. 273–280. https://doi.org/10.5220/0001770102730280.

  • 17.

    Centre for Vision, Speech and Signal Processing, University of Surrey. The Chars74K Dataset. Available online: https://infoee.surrey.ac.uk/CVSSP/demos/chars74k/ (accessed on 29 June 2026).

  • 18.

    Gilad-Bachrach, R.; Dowlin, N.; Laine, K.; et al. Crypto-Nets: Neural Networks over Encrypted Data. arXiv 2014, arXiv:1412.6181v2.

  • 19.

    Halevi, S.; Shoup, V. Algorithms in HElib. In Advances in Cryptology—CRYPTO 2014; Springer: Berlin/Heidelberg, Germany, 2014; Volume 8616, pp. 554–571. https://doi.org/10.1007/978-3-662-44371-2 31.

  • 20.

    Lee, J.; Lee, E.; Lee, J.-W.; et al. Precise Approximation of Convolutional Neural Networks for Homomorphically Encrypted Data. arXiv 2021, arXiv:2105.10879v4.

  • 21.

    Simonyan, K.; Zisserman, A. Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv 2015, arXiv:1409.1556v6.

  • 22.

    Zhang, X.; Zhao, J.; LeCun, Y. Character-level Convolutional Networks for Text Classification. In Proceedings of the Advances in Neural Information Processing Systems, Montréal, QC, Canada, 7–12 December 2015; Volume 28, pp. 649–657.

  • 23.

    Data Privacy Stack. Presidio: Data Protection and De-Identification SDK. Available online: https://github.com/data-privacy-stack/presidio (accessed on 29 June 2026).

Share this article:
How to Cite
Dai, J.; Cang, L. S. FHE-Assisted Privacy-Preserving Medical Image De-Identification. Journal of Emerging Technologies with Industrial Applications 2026, 1 (2), 9. https://doi.org/10.53941/jetia.2026.100009.
RIS
BibTex
Copyright & License
article copyright Image
Copyright (c) 2026 by the authors.
Article Metrics
36
Article Views
0
Citations