Dynamic modeling of respiratory sinus arrhythmia component from HRV with multivariate Kalman smoother

The estimates of heart rate variability (HRV) low frequency (LF) and high frequency (HF) components with constant frequency bands may distort when the frequency of respiratory sinus arrhythmia induced HF component approaches the LF-HF frequency limit. In this study we present a method for dynamicall...

Full description

Saved in:
Bibliographic Details
Main Authors: Kuoppa, P., Lipponen, J. A., Tarvainen, M. P.
Format: Conference Proceeding
Language:English
Subjects:
Frequency estimation
Heart rate variability
Kalman filters
Mathematical model
Protocols
Resonant frequency
Time-frequency analysis
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The estimates of heart rate variability (HRV) low frequency (LF) and high frequency (HF) components with constant frequency bands may distort when the frequency of respiratory sinus arrhythmia induced HF component approaches the LF-HF frequency limit. In this study we present a method for dynamically estimating the LF-HF limit and dividing the spectrum to LF and HF components that can overlap. The method is based on multivariate autoregressive model which is solved dynamically with Kalman smoother algorithm. The spectra of each individual pole with all the zeros are calculated and then multiplied with a Hanning window on the pole frequency. These spectra are summed to LF or HF components. The method was applied to three subjects whose electrocardiogram and respiration was recorded during a controlled breathing protocol. The results show that the HF component power increases when breathing frequency decreases. Also the component powers obtained with the presented method are reliable even when LF and HF frequencies are close to each other.
ISSN:1094-687X
1558-4615
2694-0604
DOI:10.1109/EMBC.2013.6609842