Differential effects of the blood pressure state on pulse rate variability and heart rate variability in critically ill patients

Figures

• 

  Fig. 1. Linear relationship between pulse rate variability and heart rate variability. Spearman correlation coefficients (ρ) were measured between a time-domain indices, b absolute and entropy indices from the frequency domain, c relative indices from the frequency domain, d centroid-related indices from the frequency domain, e Poincaré plot indices, f entropy-related indices, g phase-related indices, and h indices resulting from the detrended fluctuation analysis. All indices were obtained from pulse rate variability and heart rate variability in each blood pressure state (hypotension, normotension, and hypertension).

• 

  Fig. 2. Bias measured from the Bland–Altman analysis comparing indices extracted from pulse rate variability and heart rate variability. The bias was obtained between measured indices when comparing pulse rate variability and heart rate variability in each blood pressure state (hypotension, normotension, and hypertension) using the Bland–Altman analysis. Indices analysed were a time-domain indices, b absolute and entropy indices from the frequency domain, c relative indices from the frequency domain, d centroid-related indices from the frequency domain, e Poincaré plot indices, f entropy-related indices, g phase-related indices, and h indices resulting from the detrended fluctuation analysis.

• 

  Fig. 3. Differences between upper (LoAU) and lower (LoAL) limits of agreement obtained from Bland–Altman analysis comparing indices measured from pulse rate variability and heart rate variability. The limits of agreement were obtained from the Bland–Altman analysis performed with the extracted indices in each blood pressure state (hypotension, normotension, and hypertension). Indices analysed were a time-domain indices, b absolute and entropy indices from the frequency domain, c relative indices from the frequency domain, d centroid-related indices from the frequency domain, e Poincaré plot indices, f entropy-related indices, g phase-related indices, and h indices resulting from the detrended fluctuation analysis.

• 

  Fig. 4. Signal quality assessment algorithm. This algorithm was applied for discarding low-quality arterial blood pressure signals.

• 

  Fig. 5. Arterial blood pressure signal analysis. Example of the analysis of a 5-min arterial blood pressure (ABP, gray line) signal, with the trends for systolic (SBP, continuous line) and diastolic (DBP, dotted line) blood pressure, as well as the determination of blood pressure state (BP state, dashed line). Hypo hypotension, Normo normotension, Hyper hypertension.

• 

  Fig. 6. Electrocardiography and photoplethymography analysis for the extraction of heart rate variability and pulse rate variability, respectively. Example of a an electrocardiography (ECG) and b a photoplethysmography (PPG) signal. R peaks (black circles on the ECG signal) were detected from ECG signals to measure heart rate variability (HRV) as the time interval between consecutive R peaks (RR intervals). Onsets (black circles on the PPG signal) were detected from PPG signals to measure pulse rate variability (PRV) as the time interval between consecutive onsets (PP intervals).