A Novel Hybrid Integration of Chaotic System and GAN for Robust Medical Image Security

Abstract

Healthcare needs specialized focus on protecting medical images because these medical data contain sensitive information coupled with possible unauthorized breaches. High-security encryption methods prove difficult to maintain both high-level protection and efficient processing, especially when dealing with large medical images. Research investigates a new combination technique to protect medical images through integration of chaotic systems and deep learning GANs. A GAN-based encryption method incorporates the chaotic logistic map to replace its generator section, thus adding randomness to secure data transmission. The encryption system implements key hashing with SHA-256, followed by chaotic sequence generation and confusion encryption through pixel-level permutation, until security enhancement through XOR-based diffusion completes the process. The decryption process returns the original image using an inverse permutation followed by reverse diffusion steps. The proposed encryption system achieves both robust cryptographic protection and superior picture quality with resistance against standard cryptographic threats according to experimental outcomes. The proposed model represents an efficient method for securing medical images effectively in digital healthcare scenarios.