IoT-Enabled Smart Health Monitoring System with Deep Learning Models for Anomaly Detection and Predictive Health Risk Analytics Integrated with LoRa Technology
Antony Pradeesh D, N. P. Subiramaniyam · International Journal of Engineering Trends and Technology · 2025
The incorporation of Internet of Things (IoT) technology into advanced deep learning models has led to the development of complex health monitoring systems capable of determining anomalies and predicting health risk conditions in real-time. This research illustrates a system driven by Blood Pressure (BP), Heart Rate (HR), oxygen saturation, body temperature, Galvanic Skin Response (GSR), ECG, EMG, and particulate matter, among other sensors, in this study—all coordinated through a Raspberry Pi 5. This research considered three leading models of anomaly detection that exhibit high accuracy in handling diversified health data: Bidirectional Long Short-Term Memory (LSTM) using K-fold Cross Validation, eXtreme Gradient Boosting (XGBoost), and Random Forest. The system uses LSTM and Gated Recurrent Unit (GRU) to predict health risk conditions for health management, including hypertension, hypoxia, cardiac stress, fever, and stress. In the empirical approach, the system indicates impressive precision and accuracy in detecting anomalies and health risk prediction; thus, the system could be improved in remote health monitoring and patient care. Furthermore, the adaptive Long Range (LoRa) communication system ensures reliable data transmission without an internet connection, ensuring that data is transferred only when anomalies are detected. This paper represents the potential of integrating IoT and deep learning in bringing transformational changes in healthcare by providing scalable and efficient solutions for continuous health monitoring with early interventions.