Research And Application Of ECG Monitoring And Early Warning System Based On Neural Network

Cardiovascular disease is the first cause of death,so the detection of arrhythmia is very important for patients with heart disease.The development of artificial intelligence technology and deep learning technology are expected to help predict ECG waveform and abnormal ECG events,so as to improve the prediction accuracy.By studying the ECG detection method based on neural network,an ECG monitoring system is established.The accuracy of the proposed neural network model and the performance of the established ECG monitoring system are detected by MIT-BIH arrhythmia data set.Firstly,the collected ECG images are processed to eliminate noise and interference.Wavelet transform has outstanding characteristics in time-frequency analysis.The results show that the SNR of ECG signal after wavelet transform denoising is significantly reduced,so the practicability of wavelet transform in ECG waveform de-noising direction can be obtained.Then,a Res Net-18 residual model based on convolutional neural network is proposed to classify the heartbeat of ECG signals.Based on the traditional neural network,the residual structure is introduced.Based on the classic Res Net structure,the network uses the similar improved Res Net-18 architecture.Through the model training and parameter adjustment on MIT-BIH arrhythmia database,the results show that the algorithm has higher accuracy than other classification methods.Especially in the sensitivity and positive predictive rate of ventricular ectopic beat detection.Finally,an on-line ECG detection system is built.The ECG anomaly recognition system is nested into the Internet of things intelligent pension system.Through the analysis of the elderly ECG data,real-time analysis report is generated for the elderly,which is helpful for the elderly to understand their physical state more effectively and adjust the training and conditioning plan in time.It shows the stability of the proposed ECG monitoring method based on neural network,and verifies the economy,accuracy and stability of the ECG online detection system.Figure 30;Table 11;Reference 60...