Relationships Between Autonomic Nervous System Indices Derived from ECG Signals

Abstract

We investigated the validity of heart rate variability (HRV) as an autonomic nervous system (ANS) assessment method and also investigated the correlations between all pairs of several indices obtained from ECG signals during a task performance. We recorded ECGs from 18 healthy participants during two 30-min mental tasks, i.e., a word processor typing task and a mental arithmetic task, and calculated five indices of HRV spectral analysis and five parameters from the Poincaré plot. In the results, significant correlations were found among the participants between LF and SD2 and between HF and SD1. There were no significant differences in the correlation coefficients between the two tasks, and no subjective scores correlated with physiological indices. The correlation coefficients between the sympathetic nervous system activity indices such as LF/HF ratio and CSI, however, were not very high, and these results were different from previous studies. Based on these results, we investigated the correlation coefficient time series in each participant and found that there were fluctuations in correlation coefficients between two indices which had been reported in other studies to indicate the same ANS activity. It was clarified that weak correlations among participants occurred by the fluctuations of correlation coefficients obtained in each participant.