Smart Health Monitoring & Anomaly Detection System using Machine Learning

Abstract

The constant measurement of physiological parameters helps in detecting early health issues that can lead to serious complications, especially for elderly and chronically ill patients. This report discusses an IoT-enabled smart health monitoring system that measures physiological parameters such as heart rate, $\mathbf{S p O}_{2}$, body temperature and physical activities with the help of wearable sensors. The information obtained through sensors is sent to the cloud computing platform where machine learning algorithms analyze physiological signs to classify different types of health issues in real-time. The IoT-enabled smart health monitoring system consists of a sensor anomaly detection system that helps in recognizing abnormal or invalid data sent by sensors due to noise or transmission issues, resulting in reduced false alarms. Furthermore, the system has a health trend detector that analyzes past data records of each user to identify any physiological trends. The system supports multi-user monitoring, allowing multiple patients to be tracked simultaneously through individual profiles while maintaining personalized health analysis and records. The system helps the user make timely decisions regarding health by recommending a visit to a physician based on health risks identified through data analysis.