Karemaker Model

Karemaker, J. M. (1998). Testing the validity of LF/HF as measure of ‘sympathovagal balance’ in a computer model of cardiovascular control. Proceedings of the IX International Symposium on the Autonomic Nervous System.

Theory

A main feature of the model is that besides parasympathetic action through the baroreflex, sympathetic action is added. Sympathetic activity is triggered if the diastolic pressure falls below a reference value and influences heart rate, inotropy, and systemic resistance with some delay. In this way, a 0.1Hz oscillation is induced.

Sympathetic activation

\[\begin{aligned} SympD_n &= 0 \hspace{18mm} \text{if} \hspace{5mm} D_{n} > D_{\text{ref}} \\ SympD_n &= D_{\text{ref}} - D_n \hspace{5mm} \text{if} \hspace{5mm} D_n \leq D_{\text{ref}}\\ Symp_n &= \text{atan} \left(G_{SA}(z) \cdot SympD_{n}\right) \end{aligned}\]

Baroreflex

\[I_n^{\star} = a_1 S_{n-1} + I_0 + w_I + A_{\rho}^I\sin(2\pi f_{\rho}t + \phi) \hspace{5mm} \text{if} \hspace{5mm}I_{n-1} > 700\ ms\]

\[I_n^{\star} = a_2 S_{n-2} + I_0 + w_I + A_{\rho}^I\sin(2\pi f_{\rho}t + \phi) \hspace{5mm} \text{if} \hspace{5mm}I_{n-1} \leq 700\ ms\]

\[I_n = \frac{I_n^{\star}}{1 + \beta Symp_n}\]

with $\beta$ the sympathetic 'force' on $I$

\[D_n = S_{n} \exp\left(\frac{-I_n }{R C (1 + \delta Symp_n)}\right)\]

with $RC$ the diastolic runoff and $\delta$ the sympathetic 'force' on $D$

Starling/Restitution

\[P_n = (\gamma I_n + P_0)(1+ \varepsilon Symp_n) + w_P + A_{\rho}^P\sin(2\pi f_{\rho}t + \phi)\]

with $\varepsilon$ the sympathetic 'force' on $P$ (inotropy by symathetics and respiratory BP modulation)
The next systole is determined by $S_{n+1} = D_n + P_n$

Usage

CardioModels.KaremakerModel — Type

KaremakerModel()

A simple structure that stores a Karemaker-model.

The default initialization was taken from the original paper. When constructiong the model, single properties can be defined as keyword arguments.

Examples

julia> model = KaremakerModel()
KaremakerModel
  c: Float64 350.0
  magn_int: Float64 10.0
  r_int: Distributions.Normal{Float64}
  A_resp: Float64 60.0
  T_resp: Float64 3000.0
  a1: Float64 9.0
  a2: Float64 9.0
  b_I: Float64 0.125
  b_D: Float64 0.125
  b_P: Float64 0.125
  Dref: Float64 80.0
  t_runoff: Float64 2000.0
  offset: Float64 10.0
  starling: Float64 0.016
  magn_pulse: Float64 2.0
  A_BP: Float64 3.0
  w: Array{Float64}((6,)) [0.0, 0.01, 0.03, 0.03, 0.015, 0.005]
  symp_set: Float64 0.5
  symp_add: Array{Float64}((2,)) [-0.05, 0.2]

source

CardioModels.predict — Method

predict(model::KaremakerModel, n::Int; burnIn::Int = 0)

Predicts 'N' values of the cardiovascular variables for the respective model. With 'burnIn' a certain number af values can be dropped in the beginning.

Examples

julia julia> S, D, P, I, Symp, ρ = predict(model, 100) (S = [...], D = [...], P = [...], I = [...], Symp = [...], ρ = [...])

source