抄録
There are two main directions of studies which treat the information obtained from recorded electrocardiograms. One is the study of the waveform of the ECG which contains information about the cardiac activities. The other is the treatment of the time series analysis regarding R-R interval variability.
This paper describes time series analysis on the basis of R-R intervals from the view point of the latter. The R-R intervals were recorded continuously during rest, running load and recovery. We deal with the modeling of the R-R interval variability using autoregressive model. The spectral density function can be calculated from the autoregressive model parameters. The spectra of the R-R interval variability showed significant peaks around 0.10, 0.25, and 0.33 (beat)-1 for healthy subjects at rest. The spectrum has large amount of power at the low frequency band. On the contrary, the periodicities of spectra for patients with ASD and arrythmia, such as AF, VPB, have large amount of power at high frequency band. The spectra of the R-R interval variability in running load also showed significant peak at high frequency of 0.50 (beat)-1 for healthy athletes. The method in this paper is very practicable for the analysis of R-R interval variability.
