Evaluation of techniques for estimating the power spectral density of RR-intervals under paced respiration conditions

Abstract

Heart rate variability (HRV) analysis is increasingly used in anaesthesia and intensive care monitoring of spontaneous breathing and mechanical ventilated patients. In the frequency domain, different estimation methods of the power spectral density (PSD) of RR-intervals lead to different results. Therefore, we investigated the PSD estimates of fast Fourier transform (FFT), autoregressive modeling (AR) and Lomb-Scargle periodogram (LSP) for 25 young healthy subjects subjected to metronomic breathing. The optimum method for determination of HRV spectral parameters under paced respiration was identified by evaluating the relative error (RE) and the root mean square relative error (RMSRE) for each breathing frequency (BF) and spectral estimation method. Additionally, the sympathovagal balance was investigated by calculating the low frequency/high frequency (LF/HF) ratio. Above 7 breaths per minute, all methods showed a significant increase in LF/HF ratio with increasing BF. On average, the RMSRE of FFT was lower than for LSP and AR. Therefore, under paced respiration conditions, estimating RR-interval PSD using FFT is recommend.

Consciously Controlled Breathing Decreases the High-Frequency Component of Heart Rate Variability by Inhibiting Cardiac Parasympathetic Nerve Activity

“…To avoid the respiratory impact on HRV, consciously controlled breathing has been used in experiments and clinical trials to standardize HRV measurements (Grossman et al 1991;Piccirillo et al 2004;Sin et al 2010;DeBeck et al 2010;Schaffer et al 2013). However, it is possible that HRV is influenced by changes in f and V T during the transition from spontaneous to consciously controlled breathing, which might lead to inaccurate estimations that deviate from the true autonomic nervous activity in the heart.…”

The European Society for Computing and Technology in Anaesthesia and Intensive Care (ESCTAIC): A special issue of full papers (Erlangen meeting 2011) and conference abstracts (Timisoara, meeting 2014)

“…These reflect the activities of ESCTAIC meeting participants spanning a wide range of interests from clinical workflow [4], data representation and management [5], application of data and models in decision support [6], to detailed analysis of individual continuous measurements [7]. Each of these papers illustrates novel technological contribution in the fields of anaesthesia and intensive care.…”

“…Four full papers [4][5][6][7] were accepted following peer review for publication in this issue. These reflect the activities of ESCTAIC meeting participants spanning a wide range of interests from clinical workflow [4], data representation and management [5], application of data and models in decision support [6], to detailed analysis of individual continuous measurements [7].…”

“…The paper describes evaluation of the system in 19 patients showing that the system could achieve safe settings with reduced respiratory load. In the final paper of the issue Schaffer et al [7] address the appropriate method of estimation of power spectral density (PSD) when performing frequency domain analysis of heart rate variability. In 25 healthy subjects they found that fast Fourier transform was the most appropriate method for estimating RR-interval PSD during paced respiration.…”

Performance analysis of low-complexity welch power spectral density for automatic frequency analyser

“…There are many methods of spectral density calculation been proposed, namely as Robust Spectral Density Estimation (RSDE), Independent Component Analysis (ICA), Multitaper Power Spectral Density Estimation (MPSDE), Low-complexity Welch Power Spectral Density (LCWPSD), B-Spline Windows Power Spectral Density Estimation and etc. [1][2][3][4][5][6][7][8][9][10][11][12][13]. Among the method listed, LCWPSD is the most suitable for biosensor processing.…”