Digital Watermarking in Astrocytoma Brain Tumor MRI Images Using Discrete Stockwell Transform
Nimmy George, Manju Manuel, Michael George · 2024
An efficient digital watermarking system for embedding Electronic Patient’s Record in brain tumor MRI images using Discrete Orthonormal Stockwell Transform for telemedicine applications is presented in this article. The patient details such as name, sex and age are embedded as a watermark in astrocytoma brain tumor MRI images. Initially local binary pattern features of the MRI image are extracted. These LBP features are used to find out the scaling and embedding factors which are the keys used for embedding as well as extraction. System performance is analysed using different performance measures like Peak Signal to Noise Ratio (PSNR), Structural Similarity Index (SSIM) and Normalized Cross Correlation (NCC). A large value of PSNR value of 96.29 dB is obtained with an embedding coefficient of 38.5 which indicates that the watermarking has not affected the quality of MRI image. The NCC between the embedded and extracted watermark is 1 which indicates that the extraction process has perfectly recovered the watermark without any alteration or distortion and the system has the ability to withstand attacks.