Watermarking Techniques for the Security of Medical Images and Image Sequences

Digital medical images and image sequences, such as magnetic resonance imaging, computed tomography, and ultrasound, are widely used to diagnose diseases and abnormalities. The secure storage and exchange of such images and image sequences present challenges for hospital management. The objective of this study is to provide a framework for the integrity verification and copyright protection of medical images and image sequences. We design a systematic dual-medium watermarking model that focuses on the security of these images and sequences. Watermark bits (Wbs) are embedded in the transformed and temporal domains of images and image sequences, respectively, instead of in the spatial domain. The images are transformed using integer wavelet transforms, and motion vectors (MVs) are estimated from the image sequences to transform them into the temporal domain using seven block-matching algorithms. The best-performing MVs during frame reconstruction are reconstructed using a genetic algorithm to improve imperceptibility in the watermark-embedded image sequences. To increase the embedding capacity, the Wbs are embedded in regions with minimum or no movement, contrary to the conventional method. Conventionally, Wbs are embedded in regions of image sequences with significant movements. We evaluate the proposed model on typically used MATLAB test images and video datasets available online in Luminance (Y), blue–luminance (U), red–luminance (V) format. A comparison with five state-of-the-art methods shows that the proposed model outperforms the other methods in terms of imperceptibility and embedding capacity.