Meridian ECG information transmission system modeling using NARX neural network

• Published in: 2016 IEEE/ACIS 15th International Conference on Computer and Information Science (ICIS) pp. 1 - 6
• Main Authors: Hong He, Xiaowen Yan, Wei Wei
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2016
• Subjects: Cardiovascular diseases; Computational modeling; Data models; ECG signal; Electrocardiography; Information processing; information transmission system; meridian; modeling; Pollution measurement; Predictive models
• DOI: 10.1109/ICIS.2016.7550775

With the increasing death number of cardiovascular disease, it is significant to study ECG signals at meridian acupoints for developing new alternative and complementary therapies for chronic cardiovascular diseases. Therefore, an ECG measuring experiment at acupoints of the human meridian is firstly carried out for obtaining information transmission data of the meridian system. Then according to the nonlinear characteristics of the meridian, a nonlinear autoregressive with exogeneous inputs (NARX) network is established for the modeling of meridian information transmission system. The analysis results of acupoint ECG data of ten subjects show that the NARX model outperforms the autoregressive with exogenous input (ARX) model and autoregressive and moving average model with exogenous input (ARMAX) in the ECG signal prediction at meridian acupoints. The prediction accuracy of the NARX neural network model for meridian ECG signal is larger than 0.98.