Reversible dynamic secure steganography for medical image using graph coloring

• A reversible image steganography algorithm using graph coloring is proposed.
• The key is derived at the extraction part from the stego-image.
• The proposed algorithm is tested with robustness, invisibility and capacity parameters.

Securing data in telemedicine applications is extremely essential and therefore it is mandatory to develop algorithms which preserve the data transmitted. Steganography (information hiding technique) plays a crucial role in telemedicine applications by providing confidentiality, integrity, availability and authenticity. This paper proposes a novel steganography technique that conceals patient information inside a medical image using a dynamic key generated by graph 3 coloring problem. The proposed method ensures reversibility as the original medical image is restored after extracting the embedded data from the stego medical image. Despite the embedding of patient information in the medical image, the visual quality of the image is preserved. Experimental results show that the proposed method is resistant against uniform affine transformations such as cropping, rotation and scaling. The proposed method is designed by considering issues related to transmission errors which could contaminate the medical images transmitted. The performance of the proposed method is compared to other information hiding methods against various parameters such as robustness of stego-image against affine transformations, toughness of the dynamic key generated, detection of transmission error, embedding rate and reversibility.