Feature based Encryption for Data Privacy and Access Control for Medical application

The healthcare organization collects large complex healthcare data that are not mined to uncover the hidden information for valid decision making. Discovery of hidden relationships and patterns often goes unexploited. Advanced data mining techniques can be a remedy this situation. Personal Health records have raised as a patient centric healthcare services where the patients health information is accumulated on the cloud. It gives an opportunity for the individuals to create, manage and distribute the data on the cloud. But this involves considering the issues of security and privacy. With technological development cloud services are being used widely and comprises of multiple users and owners of this cloud service, thus results in data management complexity. The objective of the paper is to predict Chronic Kidney Disease(CKD) using classification techniques and emphasize on the consequences of security and data access control of patient health information in the real time application. To manage huge amount of data, a flexible and fine-grained access control solution should be used. A favorable approach has emerged based on Feature Based Encryption, which provides flexible and fine-grained access control. Complexity of key management can be reduced by distributing users into multiple domains. But keys and data on clouds can be intercepted by semi trusted servers. Therefore Advanced Encryption Standard is used for providing high security of data.