摘要
在当前的医疗事业中,患者的部分医疗图像蕴含敏感的个人信息,直接上传到医疗信息系统中可能会造成个人隐私泄露;针对该问题,该文设计了一种基于便携式虹膜仪的医疗图像数据安全共享系统;首先,对虹膜仪采集到的图像定位,利用离散傅里叶变换技术压缩图像,然后提取图像有效特征,采用拉普拉斯机制对所挑选的有效特征添加噪音,使整个处理过程满足ε-差分隐私;最后,将处理后的图像上传到医疗图像数据安全系统,该系统分为客户端Ⅰ与客户端Ⅱ,客户端Ⅰ能够实现传统功能,包括医生给出诊断意见和患者了解自身病情等,方便医生与患者的沟通;客户端Ⅱ能够实现安全分析功能,即科研人员对此类病情的分析研究,给出具体治疗方案,方便此类病情患者的后续治疗;结果显示,该系统能够有效保护患者个人隐私信息,且能够保持较好的数据可用性和鲁棒性。
In the current medical business,some medical images of patients contain sensitive personal information,and uploading them directly to the medical information system may cause personal privacy disclosure.To solve this problem,this paper designs a medical image data safety sharing system based on portable iris instrument.First,locate the image collected by the iris instrument,use the discrete Fourier transform technology to compress the image,and then extract the effective features of the image,use the Laplace mechanism to add noise to the selected effective features,so that the entire processing process meets-differential privacy.Finally,upload the processed image to the medical image data security system.The system is divided into client I and client II.Client I can achieve traditional functions,including doctors giving diagnosis opinions and patients understanding their own conditions,etc,which is convenient for doctors Communicate with patients;Client II can realize the safety analysis function,that is,scientific researchers analyze and study such diseases and give specific treatment plans to facilitate the subsequent treatment of patients with such diseases.The results show that the system can effectively protect patients'personal privacy information,and can maintain good data availability and robustness.
作者
李健
Li Jian(Information Division,Xuzhou Medical University,Xuzhou 221000,China)
出处
《计算机测量与控制》
2020年第10期195-199,共5页
Computer Measurement &Control
基金
国家自然基金项目(6522000057)
江苏省高校自然科学基金(18KJB520049)。
关键词
医疗图像
隐私泄露
傅里叶变换
差分隐私
medical images
privacy leakage
Fourier transform
differential privacy
