A Hybrid Reversible Digital Watermarking Algorithm Using Machine Learning for the Protection of Medical Images

There is a daily deluge of medical images created by medical institutes around the world using a wide variety of modalities. It is common practice to send these pictures to distant doctors online so that they can work together and get second opinions. We offer a medical image watermarking technology that incorporates Principal Component Analysis (PCA) to address concerns about the privacy and security of sensitive medical data during transmission. Firstly, we need to split the input medical image into two parts: the ROI and the RONI. Once these parts have been enhanced using PCA, they will be marked with two separate watermarks. In the PCA-enhanced ROI, a delicate binary design is incorporated, and in the PCA-transformed RONI, a robust composite watermark is applied. To achieve a perfect fit, the embedding procedure uses the LSB substitution technique. For experimental validation, a collection of CT scans and other medical pictures is used. While addressing issues with data authenticity and privacy in medical records, the simulation results show that the suggested PCA-enhanced method successfully increases the security of medical pictures.