Enhancing Medical Image Security through Steganography and Ensemble Deep Authentication

Medical images often contain personally identifiable information (PII) of patients. Safeguarding the privacy and preventing unauthorized access to PII of patients is paramount as it involves sensitive data that demands protection. Steganography can contribute to enhancing the confidentiality and privacy of PII by adding an extra layer of protection. Authentication mechanisms ensures that only authorized individuals have access to view or manipulate these images, authentication mechanisms serve as a safeguard against unauthorized disclosure of sensitive medical information. This proactive approach reduces the potential risks associated with identity theft and other privacy breaches, fostering a secure environment for handling patient data. The proposed model provides an Integrated approach of Steganography and Ensemble Deep Authentication. The cover image is pre-processed for creating a better storage for hiding the secret image. The Ensemble Deep Authentication consists of feature extraction and feature visualization using Deep Learning. According to Ensemble Deep Authentication, the retrieved image at the recipient end is authenticated if the histograms of the original and the retrieved image match. The suggested model exhibits significant achievement, achieving an outstanding maximum PSNR (Peak Signal-to-Noise Ratio) value of 73.84.