Enhancing personal health record security using decentralized blockchain based techniques in big data

Abstract

newline In the context of Big Data within healthcare, the analysis of Medical Personal Healthcare Records (MPHR) presents significant challenges related to data privacy and security. The increasing volume of sensitive health data being stored and exchanged amplifies concerns about potential breaches and unauthorized access. Existing methodologies often fall short in addressing these challenges, focusing insufficiently on the protection of sensitive data, which can lead to security vulnerabilities and data breaches. newlineTo address these issues, this thesis proposes a comprehensive solution combining advanced data analysis techniques with robust security measures. The proposed strategy integrates two main approaches to enhance data protection and privacy within healthcare systems. newlineIn first approach Pragmatic Quasi Sensitive Feature Clustering (PQSFC) Algorithm is utilized to cluster sensitive features within the healthcare data. This approach categorizes sensitive and non-sensitive features, which are then processed for further analysis and protection. Following feature clustering, the DenseNet Convolutional Neural Network (DenseNet-CNN) is employed to distinguish between sensitive and non-sensitive data classes. This neural network model enhances the accuracy of sensitive data prediction by leveraging deep learning techniques.