M-ary modulation-based medical image watermarking for accessing quality of service of communication channel
Ritu Agrawal, Маниша Шарма · International Journal of Biometrics · 2023
In an e-health environment, medical image protection is of prime importance as patient data is interleaved with medical images and broadcasted over the public internet. This paper presents a new electronic patient record (EPR) watermarking scheme for medical images in the mid-band discrete cosine transform using M-ary modulation. Robustness of EPR is enhanced by using a convolution encoder before embedding into the medical cover image. The imperceptibility and robust performance of the proposed watermarking scheme is tested on standard brain imaging modality using measures such as peak signal to noise ratio (PSNR), structural similarity index measure (SSIM), normalised cross-correlation (NCC) and bit error rate (BER) parameters. The proposed method achieves high PSNR and SSIM values, respectively at different watermark embedding factors for different modalities, which indicates better visual quality. The NCC and BER values of the scheme are high, indicating that the watermarking technique is appropriate for the protection of patient information and secure medical record dispersal over the public network. Furthermore, quality of service (QoS) of the communication channel is evaluated in terms of the embedded EPR, thus any separate mechanism is not required to evaluate the QoS of the communication channel.