Sensitive multimedia information of all forms is encrypted, with key, before storage and transmission, to protect from illegal use and data manipulation. Since digital images are larger, it’s crucial to encrypt the content, specifically in medical images effectively. In medical image diagnosis, even a small manipulation of data may lead to misinterpretation. This paper addresses this concern by devising an algorithm suitable for encrypting DICOM and other types of images. RC6 cipher is used for encrypting the approximation coefficients (LL), obtained by applying the Haar wavelet transform on the plain image and combined with the redistributed (confused) detailed coefficients (LH, HL, HH). Generation of keys through governing equations of Combined Logistical Tent map, adds to the robustness of the algorithm against attacks. This algorithm works well for all types of images, including DICOM. Among several image databases available, 30 different modalities of images have been taken for experimentation, and promising results have been achieved. Results show that on an average, for an image of bit-depth eight, the proposed encryption algorithm provides the PSNR of 9.0955 dB, the entropy of 7.9990 bits for an encrypted image and, with a UACI of 33.4549 and NPCR of 99.6129, the algorithm could effectively defy the statistical and differential attacks.

在存储和传输之前，所有形式的敏感多媒体信息都使用密钥进行了加密，以防止非法使用和数据操纵。由于数字图像较大，因此有效地加密内容至关重要，特别是在医学图像中。在医学图像诊断中，即使很小的数据处理也可能导致误解。本文通过设计一种适用于加密DICOM和其他类型图像的算法来解决这一问题。RC6密码用于加密近似系数（LL），该近似系数是通过将Haar小波变换应用于纯图像并与重新分配的（混淆的）详细系数（LH，HL，HH）组合而成的。通过组合后勤帐篷图的控制方程生成密钥，增强了算法抵御攻击的鲁棒性。该算法适用于所有类型的图像，包括DICOM。在几个可用的图像数据库中，已采用30种不同的图像形式进行实验，并取得了可喜的结果。结果表明，对于平均八位深度的图像，所提出的加密算法平均可提供9.0955 dB的PSNR，加密图像的7.9990位熵，以及UACI为33.4549和NPCR为99.6129的情况。可以有效地抵抗统计和差异攻击。