Real-Time ECG Monitoring through IoMT with Attention-Enhanced Deep Learning: Enabling Smart Cardiac Healthcare

Accurate and timely arrhythmia detection remains a critical component of cardiovascular disease management. We propose a smart ECG monitoring system that leverages the Internet of Medical Things (IoMT) for real-time signal acquisition and secure cloud–edge data transmission. At its core, our system employs a hybrid deep learning architecture combining Convolutional Neural Networks (CNN), Long ShortTerm Memory (LSTM) networks, and an attention mechanism , facilitating both spatial feature extraction and temporal pattern focus. Evaluated on the MITBIH Arrhythmia Database, the model achieves 98.5% accuracy, 98.2% precision, 98.0% recall , and an F1score of 98.1% — outperforming recent related studies while maintaining only 2.2 million parameters , making it suitable for edge deployment , including on devices like Raspberry Pi 4. Inference latency on edge remains under 250 ms, supporting practical realtime monitoring. The attention-enhanced framework also offers improved interpretability by highlighting diagnostically significant ECG segments. Our system’s balance of high accuracy, low computational footprint, realtime responsiveness , and clinically interpretable outputs positions it as an effective solution for smart, scalable cardiac healthcare, suitable for ambulatory and homebased monitoring applications.