临床科研中数据隐私安全平台的应用和研究

Application and research of data privacy security platform in clinical research

目的:在临床大数据深化应用的背景下,通过对临床数据进行隐私安全计算,提供更精准、安全的信息,形成一套临床+科研的良性数据闭环。方法:通过对结构化、非结构化数据进行智能分型,利用矢量化分析与联邦学习等技术手段,结合人工智能动态分析,消除信息孤岛,去除或加密有关患者隐私或国家法规要求保护的敏感信息,将结果组装,形成统一标准,提供科研应用。结果:通过在我院临床数据中心(CDR)的应用层搭建隐私安全计算平台,去除敏感信息,规范数据出口,无需人工干预,使我院临床大数据应用更安全可靠。结论:通过隐私安全计算的应用,对CDR数据进行统一的安全计算与隐私加密,能够提供临床更精准、有效的科研数据,并保证其安全可靠。

Objective To provide more accurate security information and form a set of positive data closed-loop of clinical plus scientific research through privacy and security calculations in the context of clinical big data's further application.Methods The intelligent classifications were conducted for the structured and unstructured data,and technical approaches such as vectorization analysis and federal learning were used to eliminate the information isolated islands and remove or encrypt sensitive information related to patient privacy or required by national regulations combined with artificial intelligence dynamic analysis.The results were assembled,and a unified standard was formed and applied in scientific research.Results A privacy security calculation platform has been built on the application layer of the clinical data center(CDR)of the hospital.It realizes the sensitive information removing and standardizing data exporting without manual interventions and makes the application of clinical big data in the hospital safer and more reliable.Conclusion The application of privacy security calculation attains the uniform security computing and privacy encryption of CDR data,so as to provide more accurate and effective clinical scientific research data,and ensure its security and reliability.