An Adaptive Neuro-Fuzzy Based Region Selection and Authenticating Medical Image Through Watermarking for Secure Communication

Healthcare has experienced significant growth in recent years, bringing both advantages and downsides. It is more difficult to provide a precise diagnosis using digitally transmitted medical images in today's digital age. Using an adaptive neuro-fuzzy model and an integrated watermark transformation mechanism, this study proposes a new approach for watermarking medical images based on Region of interest (ROI). Isolating the ROI from the original medical image will be accomplished first using adaptive neuro-fuzzy training techniques. Second, as a consequence of wavelet decomposition, where subbands are exchanged using a logistic mapping based on the magnitude value of each subband, all pixels are exchanged, resulting in medical data that is encrypted and unbreakable. To make the watermark image more stable, singular values are computed, and major components are calculated to prevent false-positive mistakes. When the source and watermark images are changed to separate values, this key variable is used. Finally, through an adaptive authentication code, the scheme suggested tests the strength of protection. Moreover, the authentication code is integrated into the medical image with a new image scale-up process results in better security.