Mutil-medical image encryption by a new spatiotemporal chaos model and DNA new computing for information security

Medical images are closely related to the patient's condition. We put forward the model of Sin-Arcsin-Arnold Multi-Dynamic random nonadjacent Coupled Map Lattice (SAMCML) and found that it has good chaotic properties, so we used it to design an encryption scheme for protecting medical multi-images. First, we randomly change the three bits of binary information in medical images, then the binary information before and after blurring is spliced into new pixels. We let them combine with SHA-512 to design and generate random keys. Convert the three binary bits before the blurred pixel into 0–7, then encode the image with DNA technology (Deoxyribonucleic acid technology). We designed 3D-Fisher, which realized random cross-plane scrambling. Then, new DNA operations and asymmetric DNA coding and decoding rules are designed to achieve high-quality diffusion, thus obtaining multiple ciphertext images with different sizes. Finally, the bit-plane information before blurring is used to generate new pixels. They are filled into images with a smaller size, and multiple ciphertext images with the same size are obtained. The NPCR, UACI, and entropy of this algorithm are very close to the ideal values. Compared with other algorithms, it still has some advantages in data and can be applied to medical scenes.