High capacity, robust lossless EPR data hiding using CDCS with ROI tamper detection

In telemedicine applications, patient medical records are linked to the medical secrecy. Hence confidentiality of patient´s data can be achieved by hiding Electronic Patient´s Report (EPR) data in corresponding medical images. However, medical images require extreme care when embedding additional data within them because the additional information must not affect the image quality and readability. This paper proposes high capacity; robust blind lossless reversible data hiding technique along with Region of Interest (ROI) tamper detection. A new coding technique called Class Dependent Coding Scheme (CDCS) is used to increase the embedding capacity. The original image is recovered without any distortion from the marked medical image following the extraction of the hidden data. As the statistical quantity related to the correlation among the pixel values is used for embedding the data, the system becomes robust to JPEG compression to certain extent. The embedding scheme also takes care of sensitive part of the medical image responsible for accurate diagnosis called ROL For detecting any tamper in ROI region, a secure hash function is calculated on ROI region and embedded in Region of Non-Interest (RONI) along with EPR data bits. Experimental results show that the proposed scheme exhibits high imperceptibility as well as low perceptual variations in stego images. Even a slight modification in ROI gets detected at receiver side. Imperceptibility and robustness have been tested against various image manipulation attacks.