SFIWMark: Secure Fundus Image Watermarking Scheme for Smart Healthcare Applications

The increasing volume of medical data shows significant challenges for different healthcare information management (HIM), particularly in the form of image security, its accessibility, authentication, reliability and its protections against different security threats. In the proposed work, a Fundus medical image watermarking (FMIW) scheme based on Redundant Discrete Wavelet Transform (RDWT) and Multiple Matrix Decomposition (MMD) is suggested in which additionally chaotic encryption of watermark (Wmark) image is done to further enhance the security. RDWT enhances the robustness against various types of attacks, while multiple matrix decomposition (MMD) improves the invisibility of Wmark along with the robustness against affine attacks. In this approach, a FUNDUS image of patient used as cover image, and the patient ID i.e. 'Aadhar' image is used as Wmark image. The 'Aadhar' image is divided into two halves, each of which is separately encrypted using a Henon map chaotic encryption scheme and used as a Wmark, thereby enhancing the overall security of the system. The proposed method achieves a Peak Signal-to-Noise Ratio (PSNR) exceeding 40 dB, indicating high imperceptibility of the Wmark. The Wmark is later extracted and reconstructed by merging the two halves. Experimental results demonstrate that the proposed technique outperforms existing state-of-the-art methods, offering approximately 20% improvement in robustness against various attacks.