基于差分隐私的医疗大数据隐私保护模型应用研究

Research on Privacy Protection Model and Application of Medical Big Data Based on Differential Privary

目的:随着医疗信息化应用的深入发展,以及医疗大数据挖掘、医疗大数据分析等深层次应用的普及,如何在利用医疗大数据的同时保护好患者的隐私数据,防止其敏感信息泄漏具有十分重要的意义。方法:差分隐私是一种严格且可被证明的隐私保护方法,近年来的研究使其在理论层面不断发展完善,并在数据挖掘、机器学习、推荐系统等领域得到了初步的应用。结果:在对医疗大数据领域的常用隐私保护技术进行综合叙述的基础上,对差分隐私保护技术的基本原理和研究方向进行了阐述。结论:针对不同类型医疗大数据的应用研究做了相应介绍,指出差分隐私技术存在的研究难点,最后展望了其在医学大数据隐私保护领域未来的发展方向。

Objective: With the further development of medical informatization applications, and the popularization of deep-level applications such as medical big data mining and medical big data analysis, how to protect the privacy data of patients while using medical big data to prevent their sensitive data from leaking has important practical significance. Methods: Differential privacy is a strict and proving method of privacy protection. In recent years, research has made it develop and improve at the theoretical level, and has been applied in statistics, machine learning, data mining and other fields. Results: Based on the comprehensive narrative of commonly used privacy preserving technologies in the field of medical big data, this paper describes the basic principles and research direction of differential privacy protection technologies, and presents a corresponding introduction to the application of different types of medical big data. Conclusion: This paper points out the difficulties in the research of differential privacy technology, and finally looks forward to its future development direction in the field of medical big data privacy preserving.