Using electroencephalogram classification in a convolutional neural network, infer privacy on healthcare internet of things 5.0

Abstract

When enabled by the internet of health things (IoHT), brain neuroscience may conduct online analysis of brain information through multi-variate electroencephalogram (EEG) classification, which would be a requirement for the recent surge in biofeedback technologies and medical supervision. With the ever-increasing privacy issues and vulnerabilities of conventional methods, a universal and reliable-based authentication framework for smart IoHT application with 5G technology (healthcare 5.0) is needed. Research teams have come to trust the EEG features because of their reliability, durability and universality. Fortunately, the testing paradigm's restricted functionality and poor classification accuracy have kept an EEG-based identity authentication scheme from being widely seen in IoHT scenarios. However, due to unsatisfactory categories and the failure of a reliable identity authentication scheme, it remains important in research challenges. This research presents the design of an EEG identity authentication system supported via convolutional neural network classification includes cloud support storage methodology in the healthcare 5.0 environment, resulting in extremely high reliability, consistency and protection for the next generation of smart systems. The experimental results indicate that the accuracy and efficacy of the user authentication expect a higher legal probability of success and a lower unauthorized likelihood of success from a safety perspective. As compared to other frameworks, traditional EEG-based authorization approaches test results to reveal that the proposed methodology yields the desired classification accuracy of 97.6%. The experiment performance on an authentication scenario is structured to prove that the proposed method is efficient, reliable and accurate.