Kaya Y., Gürkaş Aydın G. Z.
Computers and Electronics in Medicine, cilt.2, sa.2, ss.53-59, 2025 (Hakemli Dergi)
Özet
Data privacy in healthcare system is getting more importance day by day. In this study we introduce a novel chaos-based encryption scheme for medical X-ray images. We used a double Lorenz map as chaotic system to produce random key streams in confusion phase and in diffusion phase. We also used hash value of the original file to determine initial parameters of second chaotic map. So, our scheme is highly resistant to differential attacks. In confusion phase, we employed row-based and column-based shift operation to confuse original data. In diffusion phase we used XoR operation to encrypt data. Since Lorenz map has 3 dimensions, we used each one for different purposes like X dimension is for row-based shift operation, Y dimension is for column-based shift operation, Z dimension is for data distribution, XoR operation etc. On average, our scheme generates a NPCR value exceeding 99.66% and an UACI value of approximately 33.39% when there is a one-pixel alteration in the plaintext. Additionally, it yields an average information entropy value greater than 7.9976. Since our method utilizes a double Lorenz map, it is resilient against brute force attacks. The results from our tests and analyses indicate that our schema is pretty fast, dependable, resilient, practical, and effective. It serves as a solid encryption scheme option for medical images.