Application of Artificial Neural Network in HRV Analysis

HRV reflects the tension and balance of cardiac nerve activity, and it is a non-invasive index to detect autonomic nervous activity. In recent decades, the relationship has been fully affirmed between autonomic nervous activity and a variety of diseases through a lot of studies, especially mortality rate of some cardiovascular diseases, especially the rate of sudden death. Therefore, on the basis of application of the artificial neural Network to HRV analysis, the HRV signals can be preliminary identified through the application of error Back Propagation Network (Back Propagation Network, BP Network) and its improved algorithm. Through parameter setup and training test on different algorithms of neural network, the result shows that the additional momentum method is used when the number of hidden nodes is 10 and the learning rate is 0.5 (momentum learning rate is 0.3), which can ensure that the correct rate of training and detection and recognition of the whole neural network is 93.96% and the stability is good. The artificial neural network algorithm is applied to the HRV analysis, which provides a new research field and application space for ECG intelligent analysis.