A Differential Privacy-Preserving Framework for Secure Cloud-Based Medical Information Systems

Abstract

Ensuring patient privacy within healthcare systems is paramount, necessitating stringent measures to safeguard personal health information (PHI) from unauthorized access. This study is targeted at presents a robust model designed to facilitate secured access to patients' health data stored in a database, while preserving anonymity. Access control techniques are employed to tailor access privileges based on the role of users, including doctors, nurses, and administrative staff. This study employs the Laplace Mechanism as a strategic approach for the implementation of Differential Privacy when applied to a specific function denoted as f, which is intended to be executed on a designated database. The effective execution of the Laplace Mechanism is achieved by systematically introducing a calculated amount of noise into the output generated by the function f, wherein the magnitude of this noise is determined by a specific parameter \(\varepsilon\), as will be delineated in the subsequent sections of this discussion.The research entails the development and deployment of a web-based application utilizing contemporary web technologies, with data storage implemented on a cloud-based database hosting service. The front-end of the application is hosted on Netlify, while the backend is deployed on Heroku. To bolster security, patient and staff data are encrypted using the "simple-encryptor" JavaScript library at the front-end, mitigating the risk of unauthorized data access. This underscores its reliability in ensuring both data security and privacy within healthcare settings.