Research on the Recognition of ECG Signals based on Integrated LSTM-GRU Network

The use of artificial intelligence technology to achieve automatic diagnosis of ECG signals is an inevitable trend in future technological development. This design targets the nonstationarity and individual variability of ECG signals, proposing an integrated model of Long Short-Term Memory networks (LSTM) and Gated Recurrent Units (GRU) for heart rate signal recognition. Wavelet scattering transform is employed to extract the time-frequency features of ECG signals, thereby improving the model's convergence speed and recognition accuracy. The experiment uses the MIT-BIH Arrhythmia Database, which includes five categories of ECG signal data: N, A, V, L and R. Test results show that the integrated LSTM-GRU network outperforms traditional LSTM networks in both training speed and recognition accuracy, with an accuracy rate as high as 98.7%, which is of great significance for the future development of automatic ECG signal diagnosis technology.