Real-Time ECG Signal Classification Using Recurrent Neural Networks

This study investigates the use of Recurrent Neural Networks (RNNs) for real-time ECG signal classification. The CNN-RNN hybrid model achieved the highest accuracy of 97.5% and an F1-score of 96.5%, while the GRU model offered the best real-time performance with a latency of 40 ms and a throughput of 25 segments/s. Model optimization techniques reduced model sizes by 50%, facilitating deployment on edge devices. These results indicate that RNN-based models can enhance real-time cardiac monitoring, improving early detection and intervention for cardiovascular diseases.