Fog-Computing Based Healthcare Framework for Predicting Encephalitis Outbreak

Abstract

Acute Encephalitis Syndrome (AES) is a serious, life-threatening disease, which is endemic in India and South-East Asia, where it adds to the strain on healthcare systems. It is a communicable disease that is caused by a virus attacking the brain tissues. The main objective of this study is to present a system that allows for monitoring, controlling, and preventing the spread of Encephalitis. The system is based on Hybrid-Fog-Computing (HFC) for real-time alert generation and notifies medical caregivers in case of abnormal geospatial distribution of infections. Deep learning, incorporating Multi-scaled Long Short Term Memory (MLSTM) and Convolutional Neural Network (CNN), is used to identify the health risk in terms of the Outbreak Severity Index (OSI). Our deep-learning model is integrated with Geographical Information System (GIS) tools to analyze the spatial distribution of disease-ridden areas and a Self-Organized Mapping (SOM) technique is used to visualize AES hotspots using spatial cluster analysis techniques including Getis-Ord Gi*. To determine the category of a patient's health state, a Bayesian classifier is used. A Spatio-temporal prediction model is used to coordinate medical resources toward successful health-oriented decision-making and effective knowledge delivery. The system is validated using real datasets, and the results are compared to various state-of-the-art prediction models. The proposed model outperforms other decision models for accuracy, precision, f-measure, and overall system stability.