Hybrid medical image encryption with compression framework for Internet of Medical Things

Abstract

Patient medical data’s security, storage, and transmission are critical challenges in healthcare systems, especially in the Internet of Medical Things (IoMT) environments. The vulnerability to attacks, higher computational costs, and loss of diagnostic quality are most often failures of the conventional encryption methods due to an imbalance between security and imperceptibility. This work focuses on developing a hybrid medical image encryption and compression (HMIEC) framework that uniquely integrates encryption, compression, and watermarking to address these issues. Initially, Improved Henon Chaotic Map Encryption (IHCME) was applied on the source image, which provides higher security. Then, the preprocessing operation is performed on the cover image, which converts the color space of the cover image. Further, the Discrete Karhunen–Loève Transform (DKLT) is applied to preprocessed and encrypted images. Moreover, a naturally inspired gray wolf optimization (GWO) algorithm selects the optimal embedding coefficients. Further, medical image embedding is performed using a GWO-based optimal embedding strength factor, where the preprocessed image hides the encrypted image and generates a watermarked image. Finally, a post-processing operation is performed on the watermarked image to generate a smoother watermarked image. The proposed HMIEC system resulted in improved peak signal-to-noise ratio (PSNR) by 77.04 dB, entropy of 41.923, mean square error (MSE) of 0.001283, structural similarity index metric (SSIM) of 0.991, normalizer correlation coefficient (NCC) of 0.992, compression ratio (CR) of 21.955%, the unified average change in intensity (UACI) of 99.60%, and the number of pixels change rate (NPCR) of 33.46% as compared to existing watermarking and security systems.