Dynamic Temporal Relationship Between Autonomic Function and Cerebrovascular Reactivity in Moderate/Severe Traumatic Brain Injury

Figures

• 

  FIGURE 1. The Residuals, ACF, PACF of the PRx and PRx ARIMA Model. The figures demonstrate the minute-by-minute PRx data and an optimal ARIMA model (autoregressive order of 2, integrative order of 0 and moving average order of 0) obtained from the Akaike Information Criteria and Log Likelihood. The reduced significant lags in ACF or PACF of the ARIMA model shows that the ARIMA model in part accounts for the residual relationship between the PRx data. ACF, autocorrelation function; PACF, partial autocorrelation function; PRx, pressure reactivity index; ARIMA, autoregressive integrated moving average.

• 

  FIGURE 2. Granger Causality p-value Comparisons (n = 47). The bar graphs show the number of significant Granger Causality p-values of PRx on an ARV and an ARV on PRx, in this way the impulse that has more significant responses may be considered to have a greater influence on causality. Image (A) shows all BPV and Baroreflex, (B) shows all HRV. ARV, autonomic response variable; BPV_D, standard deviation of diastolic blood pressure variability; BPV_M, standard deviation of mean blood pressure variability; BPV_S, standard deviation of systolic blood pressure variability; HRV, heart rate variability; HRF_HF, heart rate variability high frequency; HRV_LF, heart rate variability low frequency; HRV_LF_HF, heart rate variability ratio between low/high frequency; HRV_RMS, heart rate variability root mean square; HRV_TOT, heart rate variability total; HRV_VLF, heart rate variability very low frequency; PRx, pressure reactivity; SBPV_HF, spectral blood pressure variability high frequency; SBPV_LF, spectral blood pressure variability low frequency; SBPV_TOT, spectral blood pressure variability total.

• 

  FIGURE 3. Example of Hierarchical Clusters of Two Patients. Hierarchy cluster example from two patients, the data is normalized, and the distance between the variables is Euclidean. The distance between two variables shows the height of the dendrogram where the two branches merge into a single branch, thus two variables that diverge on the last branch may be more closely linked. BPV_D, standard deviation of diastolic blood pressure variability; BPV_M, standard deviation of mean blood pressure variability; BPV_S, standard deviation of systolic blood pressure variability; HRF_HF, heart rate variability high frequency; HRV_LF, heart rate variability low frequency; HRV_LF_HF, heart rate variability ratio between low/high frequency; HRV_RMS, heart rate variability root mean square; HRV_TOT, heart rate variability total; HRV_VLF, heart rate variability very low frequency; PRx, pressure reactivity; SBPV_HF, spectral blood pressure variability high frequency; SBPV_LF, spectral blood pressure variability low frequency; SBPV_TOT, spectral blood pressure variability total.