Abstract:

Difficulty in safe sharing, openness and utilization of medical and health data is one of the most important factors hindering the development of informatization in the current health and health industry. This paper researches and discusses the difficulties faced by the health industry in the actual application scenarios of data sharing and utilization, proposes the application schemes of privacy computing in various application scenarios, and gives suggestions on the security of privacy computing models in corresponding scenarios according to actual needs.

Key words: privacy computing, data security, federated learning, model security, trusted execution environment