Medical Image Encryption using Biometric Image Texture Fusion.

Area of Science:

• Medical Imaging
• Cryptography
• Biometrics

Background:

• Exponential growth in medical image data, especially during pandemics, necessitates enhanced security measures.
• Biometric data (fingerprints, iris, faces) collected by hospitals are vulnerable to hacking and privacy breaches.
• Existing encryption methods may not adequately protect sensitive medical images and patient privacy.

Purpose of the Study:

• To propose a new texture fusion medical image encryption (TFMIE) algorithm using biometric information.
• To prevent illegal tampering and privacy invasion of medical images during storage or transmission.
• To enhance the security and efficiency of medical big data encryption.

Main Methods:

• Medical images are decomposed into n-bit-planes.
• A fusion image is created using biometric data, circular local binary patterns, and pixel-weighted averaging.
• The fused image undergoes bit-plane decomposition and XOR operations with the original medical image.
• A one-dimensional fractional trigonometric function (1DFTF) chaotic map is used for scrambling and diffusion.

Main Results:

• The TFMIE algorithm achieved an average information entropy of 7.99.
• Average values for NPCR (Number of Pixels Change Rate) and UACI (Unified Average Changing Intensity) reached 0.9958 and 0.3346, respectively.
• Demonstrated strong key sensitivity, robustness, and anti-attack capabilities compared to existing methods.
• The encryption process is lossless with high transmission efficiency.

Conclusions:

• The proposed TFMIE algorithm effectively encrypts medical images using biometric information, ensuring data security and privacy.
• The method's high performance metrics (entropy, NPCR, UACI) indicate its suitability for securing medical big data.
• TFMIE offers a robust, efficient, and lossless solution for the growing demands of medical image encryption.