The heart rate variability (HRV) signal represents one of the most promising markers of autonomic activity.
Abstract
The heart rate variability (HRV) signal represents one of the most promising markers of autonomic activity. Frequency domain analysis of HRV provides quantitative and noninvasive measure of autonomic nervous system (ANS) activity .The spectral analysis of HRV has contributed to the understanding of the sympathetic and para-sympathetic activities, development and to the balance between these two systems.
In the last years, researches have shown that from a certain stage of development, fetal HRV shows similar properties to that of adults.
Spectral Analysis of fetal HRV can be of use in monitoring the fetal’s development, and the development of the different systems that modulate the fetal Heart Rate.
Since recent projects in this lab provides us with a reliable tool to measure fetal’s R-R intervals, we will focus on the post acquisition signal processing techniques so that the cardiologist will be provided with a seemingly simple tool for both research and clinical studies. The significant and meaning of the many different measure of HRV are more complex than generally appreciated and there is a potential for incorrect conclusions and for excessive or unfounded extrapolations.
Analysis of HRV
Power spectral density (PSD) analysis provides the basic information of how power (i.e. variance) distributes as a function of frequency. Independent of the method employed, only an estimate of the true PSD of the signals can be obtained by proper mathematical algorithms.
Methods for the calculation of PSD may be classified as non-parametric and parametric. The advantages of the non-parametric methods are the simplicity of the algorithm (Fast Fourier Transform – FFT for example) and high processing speed. On the other hand the advantages of the parametric method are:
1. Smoother spectral components which can be distinguished independently of preselected frequency bands.
2. Easy post-processing of the spectrum with an automatic calculation of low and high frequency power components.
3. An accurate estimation of PSD even on a small number of samples on which the signal is supposed to maintain stationarity.
Because of these advantages we used the Auto Regressive (AR) parametric method.
The following block diagram shows the whole post processing including the PSD estimation which is the core of the process:
A PSD graph and its three band components is shown as an example :
The Case Study
As a preliminary test of the algorithm and the Lab-View application we conducted together with Dr. Taler (Rambam Hospital) a study which involved a group of 38 patients – 30 normal and 8 diabetics. We analyzed the PSD signals for each patient’s fetus during four different periods: A short while before artificially raising sugar levels in the blood, and three successive measurements after .We took each PSD and divided the signal to 10 frequency bands equally spaced between 0-0.5 Hz and calculated the power for each bin (sum integral over the range) we averaged separately for normal and diabetic and plotted the results in four different graphs – one for each period.
Graph 0 for and Graph 2, before and after the injection:
Conclusions
– There is a different in the total VLF average power density between normal and diabetic patients.
– Choosing the correct method for post processing of the HRV signal has a clinical potential in study and investigating fetus HRV.
– Choosing the auto regression parameters must be done with care, and with consideration to the expected model, as the results may change drastically.
Tools
– Matlab for the developing testing calculating and processing the results.
– LAB VIEW as the ‘end user’ application platform and user interface (UI).
Acknowledgments
We would like to thank everybody who assisted us in carrying out our project:
Prof Dan Adam, Dr. Israel Taller, Johanan Erez and the PSPL laboratory stuff.
The Ollendorff Minerva Center Fund supported this project.