Predictive Healthcare Modeling using HESN with GPR for Scalable Cloud-Based Systems

Abstract

Cloud computing has completely revolutionized the healthcare industry by enabling scalable, secure, and efficient management of medical imaging, electronic health records (EHR), and real-time monitoring data from patients. However, current traditional healthcare cloud solutions suffer from limitations such as high delays, ineffective data retrieval, and low prediction accuracy with centralized storage and existing AI models. The current procedures SVM and Random Forest in addition to predicting uncertainty have been associated with the recognition of sequential patterns and create a wrong interpretation of how a condition would evolve. Moreover, static caching methods are inefficient with the highly used medical data, thus resulting in delays and high costs of computation. The present study proposes an AI-based framework for dynamic edge cache, Gaussian Process Regression (GPR), and Hierarchical Event-Driven Stochastic Networks (HESN) in order to mitigate these challenges. The HESN scheme aids in sequential pattern identification and supports the GPR with probabilistic forecast and uncertainty prediction. Considering the experimental results, it can said that cloud-based e-health is certainly optimally improved and consumes 40 percent less energy with a cache update delay of 15 ms and 95 percent accuracy in forecasting. The contribution made by the proposed system is toward scalable and intelligent healthcare clouding facilities by enhancing the prediction toward disease progression, optimizing data retrieval, and ensuring real-time decision-making.