Medical image security using optimized chaotic affine assisted elliptic curve signcryption.

The increased need for digital images across many industries makes multimedia data security across unreliable connections a critical concern. Because of the critical nature of medical image security, numerous methods have been refined to offer effective image encryption. However, there are a number of drawbacks, such as issues with security during intricate attacks, a lengthy key generation procedure, a high error rate, and computational complexity. Hence, the proposed methodology is developed to provide an efficient medical image security system using an optimized chaotic affine-assisted Elliptic curve approach (OCAEC). Initially, the input image is signcrypted using the OCAEC method by generating the public and private keys. The Fruit Fly Optimization algorithm is utilized to select the optimal keys and ECC parameters for encryption. The E-Secrete Picture Sharing Scheme (ESPSS) is used to convert the input image into multiple shares with the RGB matrix of the image. This technique is performed to ensure the integrity and security of the proposed technique. Next, the image pixels are distorted using chaotic Arnold's cat map and the pixels are also scrambled using an affine transformation. Finally image is unsigncrypted using the private key. The proposed technique's performance was compared with the existing related strategies. The encryption performances are analysed in terms of NPCR (99.71%), UACI (33.45%), entropy (7.99), encryption time (0.1245 s), decryption time (0.0982 s), Correlation coefficient (0.00031), and BER (0.03546). The image reconstruction performance is analyzed based on Peak Signal to Noise Ratio (PSNR) (65.12 dB), SSIM (0.9645) and MSE of 0.02. Further, the performances are analysed based on statistical significance test, ablation study, computational complexity and statistical randomness test to show the effectiveness of the proposed method.