Medical image zero-watermarking algorithm based on a reversible integer quaternionic Meixner transform and a hybrid chaotic system

The protection of medical images, often transmitted over unsecured networks in telemedicine contexts, requires techniques that ensure data confidentiality, integrity, and reversibility. To meet these requirements, we introduce a novel reversible and integer-based transform called the Quaternionic Integer Reversible Meixner Transform (QIRMT). This method enables a compact and accurate representation of multidimensional data 1D, 2D and 3D, while guaranteeing lossless reconstruction (MSE = 0, PSNR = ∞), thereby overcoming the numerical limitations of classical Meixner moments. Based on QIRMT, we have designed a robust zero-watermarking scheme for the authentication of medical images. This approach combines the extraction of features via QIRMT with a chaotic hybrid system by logistic-sine map and a generalized Arnold transform to ensure secure scrambling and spatial dispersion of the watermark-without altering the original image. Experimental results on benchmark medical images demonstrate high robustness against geometric and image processing attacks (BER < 0.03; NC > 0.97), even under combined distortions such as noise, JPEG compression, cropping, and rotation. Moreover, the proposed scheme achieves a significantly reduced execution time, outperforming existing comparative methods. These results confirm that the proposed method offers a reliable, efficient, and practical solution for protecting sensitive medical imaging data, in line with the stringent requirements of modern healthcare and telemedicine systems.