一种基于区块链医疗数据共享的联邦学习方法
By introducing blockchain technology and differential privacy protection methods into federated learning, the problem of privacy leakage in medical data sharing is solved, the effectiveness of secure data sharing and model training is achieved, and the privacy and recognition accuracy of the data are ensured.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XI AN JIAOTONG UNIV
- Filing Date
- 2022-04-14
- Publication Date
- 2026-07-17
AI Technical Summary
Existing federated learning methods lack privacy protection in medical data sharing, which can easily lead to data leaks and fail to balance data privacy and validity.
The network is established using blockchain technology, the federated learning process is planned through smart contracts, training events are recorded using differential privacy and an immutable distributed ledger, and encrypted gradient aggregation is performed. Combined with noise addition and random selection of a secure aggregator, encrypted data transmission and privacy protection are achieved.
It effectively protects the privacy of medical data, prevents reverse engineering attacks, maintains the recognition accuracy and training effect of machine learning models, and ensures the security and privacy of data sharing.
Smart Images
Figure CN114912631B_ABST