A visually meaningful medical image encryption scheme based on image steganography and memristive Hopfield neural networks

Abstract With the advancement of telemedicine technology, the security of digital medical images has become increasingly important. To address this issue, this paper proposes a visually meaningful color medical image encryption algorithm. First, a high-dimensional chaotic sequence is generated using a memristive Hopfield neural network. Subsequently, multi-channel pixel permutation is performed based on a chaos-driven pseudo-random strategy, followed by the implementation of a double-layer diffusion mechanism integrating cellular automata and dynamic deoxyribonucleic acid (DNA) coding. Finally, a chaos-driven cross-channel least significant bit (LSB) embedding approach is adopted. Simulation experiments and security analyses demonstrate that the proposed algorithm achieves excellent encryption performance, a large key space, and strong robustness against noise and data-loss attacks, thereby effectively ensuring the secure transmission of digital medical images.