ESTIMATION OF CENTRAL NERVOUS SYSTEM ACTIVITY BY DATA MINING NORMAL SINUS-ECG RHYTHMS

摘要

In this study, we propose a method to assess central nervous system activity in terms of activity-degree in sympathetic and parasympathetic state by estimating the cardiac oscillator-parameters of Integral Pulse Frequency Modulation (IPFM) model in which the artificial heart rhythms are generated by modulating sinusoid signal with applying the threshold level for resolving R-peaks. With this aim, we proposed a modified IPFM model by the empirical method with applying preset-threshold of unity. The artificial R-R interval data are analyzed by the time and frequency-domain features to describe Heart Rate Variability (HRV) under the effects of cardiac oscillator constants. The benchmarking MIT/BIH database consisted of Normal Sinus Electrocardiogram rhythms (NSR-ECG) are utilized to estimate the sympathetic and parasympathetic constants by comparing HRV measures on MIT/BIH NSR with those on the data generated by our modified IPFM model. Based on our experimental results on estimating the modulatory parameters of central nervous system activity, we can conclude that IPFM parameters on the real ECG data can be effectively estimated to assess cardiac-sympathetic and parasympathetic activity.