Role-based policy to maintain privacy of patient health records in cloud

Modern healthcare system collects health information from health assisted gadgets of different sources and stores them in the cloud storage servers as an electronic record called the patients health records (PHR) and ensures the availability whenever and wherever needed. An important issue in this centralized cloud storage is the loss of privacy and security of sensitive PHR. Existing and the most recent solutions on privacy and security provisioning are purely based on role-based access control (RBAC). However, these RBAC schemes suffer from role explosion due to the increasing number of different roles. Furthermore, managing all those roles in order to provide proper access permissions can become a complex problem. Dynamic segregation of duty relations reduces the number of potential permissions that can be made available to a user by placing constraints on the users by assigning a set of roles. In order to address the above stated problem, this paper proposes a hybrid framework called MediTrust. The proposed MediTrust combines two schemes namely RBAC and attribute-based encryption (ABE) and works on semantic database, ensuring the accessibility of patient data for different access controls. The patient data are encrypted at the provider side before outsourcing it to the cloud server and then it is decrypted again at the user end after being downloaded from the cloud server. The general information of the patient collected as PHR is stored in a separate cloud server, and the medical reports are stored separately in yet another cloud server. A second-step security control is provided using CAPTCHA which is mainly used as a security check to ensure that only human users can log in to the MediTrust. A third-step security control is also provided in which one key is shared to user’s registered mobile number and another key is shared to user’s e-mail id. In MediTrust, combination of these two keys is required to decrypt the PHR. Further, ABE polices and access control security mechanisms for privacy preservation have been validated on PHR using Amazon AWS EC2 CA. Performance evaluation results show that the proposed MediTrust is better than existing work in terms of time complexity and computational overhead.