Chapter 6 - An evolutionary predictive security control architecture for healthcare transactions in IoT-based cloud storage

Abstract

IoT has advanced significantly in the arena of processing very large volumes of data since the introduction of IoT cloud-based devices. However, data privacy and overhead processing are two significant issues in the IoT-based cloud-care system. Another difficult component of health systems is forecasting illness using patient data from IoT devices. In order to provide excellent patient data protection and disease prediction in the IoT-based cloud system, this chapter presents a revolutionary predictive security architecture for IoT-based health clouds. Two linked cryptic and predictive security and prediction systems are included in the proposed design. This paradigm was created with IoT integration in cloud-based health systems in mind. The patient information, as well as the keywords, is encrypted and stored in the cloud. The healthcare provider decrypts the cloud data that has been provided. Using the stated prediction model, the decrypted data are then used to predict the patient's sickness. The proposed design is tested using data from the UCI repository, specifically heart disease and diabetic type. According to the findings of the performance investigation, the proposed model outperformed the present models in terms of query processing time and key generation time, as well as lower computing costs. The proposed revolutionary architecture correctly predicted both diabetes and heart disease with an accuracy of 90.13% and 94.62%, respectively. The proposed architecture has an overall precision of 92.57% when it comes to disease prediction. All of the features must be addressed in order to incorporate a stable and realistic prediction model based on the neural network in future work.