【Author】
Zhang, Huiru; Li, Guangshun; Zhang, Yue; Gai, Keke; Qiu, Meikang
【Source】KNOWLEDGE SCIENCE, ENGINEERING AND MANAGEMENT, PT III
【Abstract】With the booming development of big data technology and health care applications, data in the medical field is characterized by explosive growth, and medical data is valuable, which is the privacy data of patients. However, the characteristics and storage environment of medical big data have brought great challenges to the realization of privacy protection of medical data. In order to ensure the protection of data privacy when sharing medical data, we propose a medical data privacy protection framework based on blockchain (MPBC). In this framework, we protect privacy by adding differential privacy noise into federated learning. In addition, the growing volume of medical data could make blockchain storage problematic. Therefore, a storage mode is proposed to reduce the storage burden of blockchain. The raw data are stored locally and only the hash value calculated by IPFS are stored in blockchain. To enhance the performance, a mechanism is used to validate transactions and aggregate the model. Security analysis shows that our method is a safe and effective way to implement medical data.
【Keywords】Blockchain; Medical data; Privacy protection; Federated learning
【标题】使用联邦学习的基于区块链的隐私保护医疗数据共享方案
【摘要】随着大数据技术和医疗保健应用的蓬勃发展，医疗领域的数据呈现爆发式增长的特点，医疗数据具有价值，即患者的隐私数据。然而，医疗大数据的特点和存储环境给医疗数据隐私保护的实现带来了巨大挑战。为了保证共享医疗数据时的数据隐私保护，我们提出了一种基于区块链（MPBC）的医疗数据隐私保护框架。在这个框架中，我们通过将差分隐私噪声添加到联邦学习中来保护隐私。此外，不断增长的医疗数据量可能会给区块链存储带来问题。因此，提出了一种存储模式来减轻区块链的存储负担。原始数据存储在本地，只有 IPFS 计算的哈希值存储在区块链中。为了提高性能，使用一种机制来验证交易并聚合模型。安全性分析表明，我们的方法是实现医疗数据的一种安全有效的方法。
评论