The COVID-19 pandemic highlights the need for effective and non-intrusive methods to monitor the well-being of elderly individuals in their homes, especially for early detection of potential viral infections. Conspicuously, the present paper develops a Multi-scaled Long Short Term Memory (Ms-LSTM) model for the routine health monitoring of elderly patients to detect COVID-19. The proposed method offers home-based health diagnostics through urine analysis by leveraging the IoT-Fog-Cloud paradigm. Mainly, the proposed model constitutes a four-layered architecture: data acquisition, fog layer, cloud layer, and interface layer. Each layer serves distinct functionalities and provides specific services, thereby collectively enhancing the overall effectiveness of the model. The statistical results of the study demonstrate the superior performance of the proposed Ms-LSTM model in comparison to state-of-the-art methods, including Artificial Neural Networks (ANN), K-Nearest Neighbors (K-NN), Support Vector Machine (SVM), Random Forest, and LSTM. Further, the proposed model attains a mean temporal efficiency of 39.23 seconds. It exhibits high reliability (92.97%), stability (70.06%), and predictive accuracy (93.25%).
Keywords: Artificial intelligence; CNN-LSTM; IoT; Machine learning algorithms; Smart toilet system.
Copyright © 2023 Elsevier B.V. All rights reserved.