Mediseckey: A Strong Stream Cypher for Health Imaging driven through Machine Learning

MediSecKey is an encryption system built on deep learning intended to protect medical images through the creation of private keys from sophisticated neural networks. The system uses a transformation domain to improve key generation to provide robust encryption and decryption of confidential medical information. The system was assessed using the chest X-ray dataset from Montgomery County a publicly available collection of X-ray pictures of the chest, to assess its performance in tasks such as disease detection, segmentation, or classification, demonstrating its effectiveness in medical image analysis applications, BraTS18 dataset, and Brachial Plexus dataset, proving its ability to maintain data integrity and confidentiality. By integrating deep learning with secure encryption technologies, MediSecKey ensures safe storage and transmission of medical images, hence protecting patient confidentiality and preventing illegal access to confidential healthcare data.