基于国密加性同态的医学影像可逆信息隐藏方法

National cryptographic algorithm and additive homomorphism based scheme for reversible data hiding in medical images

医学影像共享是医疗信息云共享中最重要的部分,因为医疗信息80%以上是医学影像,但信息共享面临数据安全、隐私保护和信息检索等问题。虽然已有很多密文域可逆信息隐藏(RDH-EI,Reversible Data Hiding in Encrypted Image)方案,但一般不能直接应用于DICOM医学影像上。为了满足云服务中DICOM文件的隐私保护和信息检索需求,文章提出一种基于ZUC加性同态和多层差值直方图平移的DICOM图像RDH-EI方案。所提方案不改变DICOM文件格式,不增加文件大小,且图像解密和信息提取可分离。实验结果表明,所提出的方案具有良好的灵活性和计算效率,是一种适用云共享的RDH-EI方案。

Medical image sharing is the most important part of medical information cloud sharing because more than 80% of medical information is medical image,but information sharing faces the problems of data sccurity,privacy protection and data retrieval.Although there are many reversible data hiding in cncrypted image(RDH-EI)schemes,they cannot be directly applied to DICOM medical images.In order to meet the privacy protection and data retrieval requirements of DICOM files in cloud scervices,the paper proposes an RDH-EI scheme for DICOM images based on ZUC additive homomorphism and multi-layerdifference histogram shifting.The proposed scheme does not change the DICOM file format or increase the file size,and the image decryption and information extraction are separable.The experimental results show that the proposed scheme has good flexibility and computational efficiency,which is an RDH-EI scheme suitable for cloud sharing.