高效隐私保护的跨模态医疗数据检索方法
网络出版日期: 2025-07-18
基金资助
国家重点研发计划(2024YFB3908500);国家自然科学基金青年科学基金(62402388);广东省基础与应用基础研究基金(2023A1515110941)
版权
Efficient and privacy-preserving cross-modal retrieval scheme over medical data
Online published: 2025-07-18
Copyright
面向多模态医疗数据的安全外包存储与跨模态检索需求,提出了一种高效隐私保护的跨模态医疗数据检索方案。该方案利用改进的跨模态哈希网络提取电子病历与医疗影像的归一化语义哈希编码,并采用基于容错学习问题的安全内积协议实现哈希编码的语义相似性度量,支持隐私保护的跨模态医疗数据检索。同时,基于凝聚型层次聚类和多重索引哈希表构造分层索引结构,将查询复杂度从线性降低到亚线性,支持大规模数据的高效检索。严格的安全性分析表明,所提方案可以抵抗选择明文攻击,保证外包数据及检索请求的安全性。基于真实和模拟数据集的实验评估表明,所提方案的搜索时间与明文方案接近,具有实用性。
蔡珊珊 , 朱丹 , 汪群泽 , 答乐梅 , 慕德俊 , 胡伟 . 高效隐私保护的跨模态医疗数据检索方法[J]. 网络空间安全科学学报, 2025 , 3(2) : 70 -83 . DOI: 10.20172/j.issn.2097-3136.250207
To achieve secure storage and cross-modal retrieval over outsourced multi-modal medical data, an efficient and privacy-preserving cross-modal retrieval scheme for smart healthcare was proposed. At first, the proposed scheme applied an improved cross-modal hashing network to extract normalized semantic hash codes from electronic medical records and medical images. Then, it employed a secure inner product protocol based on learning with errors problem to measure the semantic similarity between the encrypted hash codes. After that, a hierarchical index structure was constructed by combining agglomerative hierarchical clustering with multi-index hash tables. The constructed structure could reduce query complexity from linear to sub-linear, thereby achieving the efficient retrieval on large-scale datasets. Rigorous security analysis demonstrated that the proposed scheme could resist chosen plaintext attacks, and ensure the security of outsourced data and retrieval queries. Experimental evaluations on both real-world and synthetic datasets showed that the retrieval time of the proposed scheme was comparable to plaintext-based methods, showing its practicality.
表 2 符号定义Table 2 Symbol definitions |
| 符号 | 定义 |
| M、T | 医疗影像/电子病历数据集 |
| 哈希编码 | |
| p、q、n | 加密算法参数 |
| 查询哈希编码 | |
| k | 哈希编码维度 |
| 高斯分布标准差 | |
| 查询向量的簇中心 | |
| S | 子码数 |
| K | 簇中心数 |
| 多重索引哈希表 | |
| 实例的唯一标识符 |
表 3 簇集合存储实例Table 3 Examples of cluster sets |
| ID | M/T | ||
| 5769 | |||
| 5770 | |||
| 5771 | |||
表 4 簇中心集合存储实例Table 4 Examples of cluster center sets |
| Samples | |
| 201 | |
| 145 | |
| 162 | |
表 5 检索准确率对比Table 5 Comparison of retrieval accuracy |
| 16-bit | 32-bit | 64-bit | 128-bit | |
| 明文 | 0.532 6 | 0.555 7 | 0.535 1 | 0.542 3 |
| 密文 | 0.532 6 | 0.555 7 | 0.535 1 | 0.542 3 |
①
| 1 |
SHI Y, CHEN S, YOU X, et al. Deep supervised information bottleneck hashing for cross-modal retrieval based computer-aided diagnosis[J]. arXiv preprint, arXiv:, 2205, 08365, 2022.
|
| 2 |
ZHANG Y, OU W, SHI Y, et al. Deep medical cross-modal attention hashing[J]. World Wide Web, 2022, 25 (4): 1519- 1536.
|
| 3 |
张嘉诚, 欧卫华, 陈英杰, 等. 胸部X线影像和诊断报告的双塔跨模态检索[J]. 计算机应用研究, 2023, 40 (8): 2544- 2548.
ZHANG J C, OU W H, CHEN Y J, et al. Dual-tower cross-modal retrieval of chest X-ray images and diagnostic reports[J]. Application Research of Computers, 2023, 40 (8): 2544- 2548.
|
| 4 |
王祥宇, 马建峰, 苗银宾. 高效隐私保护的多用户图像外包检索方案[J]. 通信学报, 2019, 40 (2): 31- 39.
WANG X Y, MA J F, MIAO Y B. Efficient privacy-preserving outsourced image retrieval scheme for multi-users[J]. Journal on Communication, 2019, 40 (2): 31- 39.
|
| 5 |
谢晴晴, 宋亮晴, 冯霞. 面向医疗数据分享的轻量级且安全的搜索方案[J]. 通信学报, 2024, 45 (11): 206- 222.
XIE Q Q, SONG L Q, FENG X. A lightweight and secure search scheme for medical data sharing[J]. Journal on Communication, 2024, 45 (11): 206- 222.
|
| 6 |
LI Y, MA J, MIAO Y, et al. Secure and verifiable multikey image search in cloud-assisted edge computing[J]. IEEE Transactions on Industrial Informatics, 2020, 17 (8): 5348- 5359.
|
| 7 |
ZHU D, ZHU H, WANG X, et al. An accurate and privacy-preserving retrieval scheme over outsourced medical images[J]. IEEE Transactions on Services Computing, 2022, 16 (2): 913- 926.
|
| 8 |
LI M,ZHU Y,DU R,et al. LPCR-IoT:Lightweight and privacy-preserving cross-modal retrieval in IoT[J]. IEEE Internet of Things Journal,2025,DOI:10.1109/JIOT.2025.3526939.
|
| 9 |
CAO Y,ZHANG H,SHANG X. A privacy-preserving cross-modal retrieval scheme based on CLIP and deep hashing[C]//Proceedings of the 2025 IEEE International Conference on Acoustics,Speech and Signal Processing. IEEE,2025:1-5.
|
| 10 |
HOTELLING H. Relations between two sets of variates[M]. New York:Springer,1992:162-190.
|
| 11 |
LI D,DIMITROVA N,LI M,et al. Multimedia content processing through cross-modal association[C]//Proceedings of the 11th ACM International Conference on Multimedia. ACM Press,2003:604-611.
|
| 12 |
ANDREW G,ARORA R,BILMES J,et al. Deep canonical correlation analysis[C]//Proceedings of the 30th International Conference on Machine Learning. PMLR,2013:1247-1255.
|
| 13 |
BRONSTEIN M M,BRONSTEIN A M,MICHEL F,et al. Data fusion through cross-modality metric learning using similarity-sensitive hashing[C]//Proceedings of the 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE,2010:3594-3601.
|
| 14 |
KUMAR S,UDUPA R. Learning hash functions for cross-view similarity search[C]//Proceedings of the 22nd International Joint Conference on Artificial Intelligence. IJCAI Organization,2011:1360-1365.
|
| 15 |
WANG D,GAO X,WANG X,et al. Semantic topic multimodal hashing for cross-media retrieval[C]//Proceedings of the 24th International Joint Conference on Artificial Intelligence. IJCAI Organization,2015:3890-3896.
|
| 16 |
LIN Z,DING G,HU M,et al. Semantics-preserving hashing for cross-view retrieval[C]//Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition. IEEE,2015:3864-3872.
|
| 17 |
JIANG Q Y,LI W J. Deep cross-modal hashing[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. IEEE,2017:3232-3240.
|
| 18 |
LEE K H,CHEN X,HUA G,et al. Stacked cross attention for image-text matching[C]//Proceedings of the 15th European Conference on Computer Vision. Springer,2018:201-216.
|
| 19 |
KARUNARATHNE S M, SAXENA N, KHAN M K. Security and privacy in IoT smart healthcare[J]. IEEE Internet Computing, 2021, 25 (4): 37- 48.
|
| 20 |
KOCABAS O,SOYATA T. Towards privacy-preserving medical cloud computing using homomorphic encryption[M]. Hershey:IGI Global,2020:93-125.
|
| 21 |
LIU Q, ZHOU F, CHEN H. Secure medical data on cloud storage via DNA homomorphic encryption technique[J]. Physical Communication, 2024, 64, 102295.
|
| 22 |
PARTHASARATHY S,HARIKRISHNAN A,NARAYANAN G,et al. Secure distributed medical record storage using blockchain and emergency sharing using multi-party computation[C]//Proceedings of the 11th IFIP International Conference on New Technologies,Mobility and Security. IEEE,2021:1-5.
|
| 23 |
ZHANG C, LUO X, FAN Q, et al. Enabling privacy-preserving multi-server collaborative search in smart healthcare[J]. Future Generation Computer Systems, 2023, 143, 265- 276.
|
| 24 |
FU J, WANG N, CAI Y. Privacy-preserving in healthcare blockchain systems based on lightweight message sharing[J]. Sensors, 2020, 20 (7): 1898.
|
| 25 |
TANG Y,XIONG L,HE M,et al. Lightweight privacy-preserving medical diagnostic scheme for internet of things healthcare[C]//Proceedings of the Sixth International Conference on Frontiers in Cyber Security. Springer Nature Singapore,2023:613-625.
|
| 26 |
顾一凡,杨雪冰,朱承瑶,等. 面向患者跨模态检索的图增强哈希网络模型[J]. 计算机辅助设计与图形学学报,2024,DOI:10.3724/SP.J.1089.2024-00351.
GU Y F,YANG X B,ZHU C Y,et al. A graph-enhanced hashing network model for cross-modal retrieval of patient data[J]. Journal of Computer-Aided Design & Computer Graphics,2024,DOI:10.3724/SP.J.1089.2024-00351.
|
| 27 |
CHATFIELD K, SIMONYAN K, VEDALDI A, et al. Return of the devil in the details: Delving deep into convolutional nets[J]. arXiv preprint, arXiv:, 1405, 3531, 2014.
|
| 28 |
TAUD H,MAS J F. Multilayer perceptron (MLP)[M]. Cham:Springer International Publishing,2017:451-455.
|
| 29 |
KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[J]. arXiv preprint, arXiv:, 1609, 02907, 2016.
|
| 30 |
CHEON J H,KIM D,KIM D,et al. Lattice-based secure biometric authentication for hamming distance[C]//Proceedings of the 26th Australasian Conference on Information Security and Privacy. Springer International Publishing,2021:653-672.
|
| 31 |
MÜLLNER D. Modern hierarchical, agglomerative clustering algorithms[J]. arXiv preprint, arXiv:, 1109, 2378, 2011.
|
| 32 |
HAO R, QIN Y, QIANG Y. Fast retrieval of similar images of pulmonary nodules based on deep multi-index hashing[J]. International Journal of Wireless and Mobile Computing, 2023, 25 (4): 303- 308.
|
| 33 |
YANG L, ZHANG W, MIAO Y, et al. Secure and efficient cross-modal retrieval over encrypted multimodal data[J]. IEEE Transactions on Computers, 2025, 74 (4): 1405- 1417.
|
| 34 |
JOHNSON A E W, POLLARD T J, BERKOWITZ S J, et al. MIMIC-CXR: A de-identified publicly available database of chest radiographs with free-text reports[J]. Scientific Data, 2019, 6, 317.
|
/
| 〈 |
|
〉 |