Sensitivity-Aware Differential Privacy for Federated Medical Imaging
Sensors · 30.04.2025 · 25 (9) · 2847
DOI: 10.3390/s25092847 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 30.04.2025. Verfasst von Lele Zheng und 7 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science and Technology, Xidian University, Xi'an 710126, China. Hauptthemen laut MEDLINE: Computer Security, Privacy, Diagnostic Imaging. Die Arbeit wurde bisher 30-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the National Key R&D Program of China, the National Natural Science Foundation of China, the Technology Innovation Leading Program of Shaanxi.
Abstract (Originalsprache)
Federated learning (FL) enables collaborative model training across multiple institutions without the sharing of raw patient data, making it particularly suitable for smart healthcare applications. However, recent studies revealed that merely sharing gradients provides a false sense of security, as private information can still be inferred through gradient inversion attacks (GIAs). While differential privacy (DP) provides provable privacy guarantees, traditional DP methods apply uniform protection, leading to excessive protection for low-sensitivity data and insufficient protection for high-sensitivity data, which degrades model performance and increases privacy risks. This paper proposes a new privacy notion, sensitivity-aware differential privacy, to better balance model performance and privacy protection. Our idea is that the sensitivity of each data sample can be objectively measured using real-world attacks. To implement this new notion, we develop the corresponding defense mechanism that adjusts privacy protection levels based on the variation in the privacy leakage risks of gradient inversion attacks. Furthermore, the method extends naturally to multi-attack scenarios. Extensive experiments on real-world medical imaging datasets demonstrate that, under equivalent privacy risk, our method achieves an average performance improvement of 13.5% over state-of-the-art methods.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Computer Science and Technology, Xidian University, Xi'an 710126, China.
- Department of Computer Science, Institute of Science Tokyo, Tokyo 152-8550, Japan.
- Faculty of Data Science, Osaka Seikei University, Osaka 533-0007, Japan.
- Graduate School of Information Science, University of Hyogo, Hyogo 670-0092, Japan.
- Graduate School of Information Science and Technology, University of Tokyo, Tokyo 113-0033, Japan.
- Graduate School of Medicine, University of Tokyo, Tokyo 113-0033, Japan.
ORCID-Profile: Essam A Rashed, Tao Zhang
Förderung
- the National Key R&D Program of China
- the National Natural Science Foundation of China
- the Technology Innovation Leading Program of Shaanxi
- the Innovation Capability Support Program of Shaanxi
- the Fundamental Research Funds for the Central Universities
- National Key R & D Program of China
- National Natural Science Foundation of China
- Technology Innovation Leading Program of Shaanxi
Zitieren
APA
Zheng, L., Cao, Y., Yoshikawa, M., Shen, Y., Rashed, E. A., Taura, K., Hanaoka, S., & Zhang, T. (2025). Sensitivity-Aware Differential Privacy for Federated Medical Imaging. Sensors, 25(9), 2847. https://doi.org/10.3390/s25092847
Vancouver
Zheng L, Cao Y, Yoshikawa M, Shen Y, Rashed EA, Taura K, et al. Sensitivity-Aware Differential Privacy for Federated Medical Imaging. Sensors. 2025;25(9):2847. doi:10.3390/s25092847
BibTeX
@article{Zheng2025,
author = {Zheng, Lele and Cao, Yang and Yoshikawa, Masatoshi and Shen, Yulong and Rashed, Essam A and Taura, Kenjiro and Hanaoka, Shouhei and Zhang, Tao},
title = {Sensitivity-Aware Differential Privacy for Federated Medical Imaging},
journal = {Sensors},
year = {2025},
volume = {25},
number = {9},
pages = {2847},
doi = {10.3390/s25092847}
}
EUVIMED