Deep Learning Based Wearable Sensor System for Realtime Cardia Monitoring and Anomaly Detection During Sleep

Cardiovascular health is critical for overall well-being, and continuous monitoring of heart activity, especially during sleep, can provide valuable insights into potential heart conditions. Traditional methods for cardiac monitoring, such as electrocardiograms (ECGs) and Holter monitors, are often bulky, invasive, and limited to clinical settings. A novel deep learning-based wearable sensor system for real-time heart monitoring and anomaly detection while you sleep is presented in this research. The proposed system leverages the Inception V3 architecture to accurately classify ECG signals into three categories: normal, bradycardia, and abnormal. The system provides real-time feedback through visual and auditory alerts and enables remote monitoring via an IoT module, allowing healthcare professionals to track the user's heart health continuously. Experimental results demonstrate high classification accuracy (98% for normal signals, 95% for bradycardia, and 92% for abnormal signals) and reliable real-time performance, making the system a promising solution for proactive cardiac health management.