An atrioventricular node model incorporating autonomic tone

Figures

• 

  FIGURE 1. A schematic representation of the AV node model. Note that retrograde conduction is possible within the AV node model. For simplicity, only a subset of the ten nodes in each pathway is shown.

• 

  FIGURE 2. Illustration of the distribution-based sensitivity analysis. (A) RR¯ plotted against one of the model parameters ΔR SP . The samples that are used to estimate the conditional CDFs Fym|xn(ym|xn∈Ic) are illustrated as circles and the conditioning intervals Ic are illustrated with vertical dotted lines. The samples that are used to estimate the unconditional CDF Fym(d)(ym) are illustrated as diamonds. (B) Fym|xn(ym|xn∈Ic) are illustrated as black lines, where the leftmost line corresponds to Fym|xn(ym|xn∈I1) with the lowest ΔR SP values and the rightmost line corresponds to Fym|xn(ym|xn∈I15) . The 1,000 Fym(d)(ym) lay all within the area illustrated by the red patch. (C) Each of the 15 squares correspond to mediand=1,…,DKS(Fym(d)(ym),Fym|xn(ym|xn∈Ic)) .

• 

  FIGURE 3. Schematic illustrating how the clinical and simulated RR series characteristic trends are computed.

• 

  FIGURE 4. Tilt recordings of 24 patients divided into supine (blue), HDT (red) and HUT (yellow). The bars represent the length of the recorded ECG data. All recordings are centered along the middle of the HDT section.

• 

  FIGURE 5. Distribution-based sensitivity indices describing the influence of changes in the 14 model parameters to changes in the three RR series characteristics. A model parameter is assumed to have influence on the RR series characteristics if the sensitivity coefficient is above the threshold of the dummy parameter (horizontal black line), otherwise it is not influential and illustrated with a white bar. The black vertical line illustrates the 95% confidence interval of the t bootstrapping iterations of the sensitivity coefficient. The ranking of the four most influential model parameters for each RR series characteristic is shown with the numbers above the bars.

• 

  FIGURE 6. Averaged mean and standard deviation of the AA series estimated from ECG recordings of 24 patients. The first 5 min were during supine position, followed by 5 min of HDT and 5 min of HUT. The horizontal dotted line illustrates the average of the μ trend during the first 5 min.

• 

  FIGURE 7. Average clinical RR series characteristics (A) RR¯C(t) (B) RRVC(t) and (C) RRIC(t) (yellow) and average simulated RR series characteristics for the original model (A) RR¯O(t) (B) RRVO(t) and (C) RRIO(t) (red) and average simulated RR series characteristics for the extended model (A) RR¯E(t) (B) RRVE(t) and (C) RRIE(t) (blue). The dashed black lines mark the transition between the supine and HDT, and HDT and HUT, respectively. Horizontal black lines show 5-min averages of RR¯E(t) , RRVE(t) and RRIE(t) during HDT or HUT with A R as indicated and A D = 1. Arrows show the impact of perturbing A D by +0.1 (arrow pointing up) or −0.1 (arrow pointing down).