Skip to main content

Analog computer model of the liver

Short answer

An analog computer uses voltage to represent variables and solves differential equations via integrators, summers, and potentiometers. The three-compartment liver model: dx1/dt = -k1 x1 + k2 x2, dx2/dt = k1 x1 - (k2+k3) x2, dx3/dt = k3 x2, implemented with interconnected integrators. Pros: parallel, fast, interactive; cons: limited accuracy (about 0.1%).

What is an analog computer / analog method? Design the analog computer model of the liver and explain it (with necessary equations and figures; pros and cons of analog computer).

Analog computer and analog method

An analog computer is a computing device that represents each variable by a continuously varying physical quantity (usually voltage) and solves equations by connecting electronic units that behave like the mathematical operations. The analog method solves a differential equation by building a circuit whose voltages obey the same equation (an analogue of the system) and recording the output.

The main units are:

UnitFunction
SummerAdds input voltages, output is the negative sum
IntegratorIntegrates the sum of its inputs with respect to time, with sign inversion
Coefficient potentiometerMultiplies a voltage by a constant between 0 and 1
InverterChanges the sign of a voltage
Multiplier / function generatorProduct of two variables, nonlinear functions
Recorder / scopeDisplays the output against time

Liver model

The liver removes substances from the blood and passes them into bile. A common test injects a small amount of a radioactive-tagged dye into the blood and measures how the amount in blood, in liver and in bile varies with time. The system is modelled by three compartments (a compartmental model):

  • x1(t): amount of tracer in the blood.
  • x2(t): amount of tracer in the liver.
  • x3(t): amount excreted into bile (and intestine).

Assumptions: each flow is proportional to the amount in the compartment it leaves (first-order kinetics); the compartments are well mixed; the tracer injected at t = 0 gives x1(0) = X0, x2(0) = x3(0) = 0.

  Blood x1 --k1--> Liver x2 --k3--> Bile x3
           <--k2--

Equations

dx1/dt = -k1 x1 + k2 x2
dx2/dt =  k1 x1 - (k2 + k3) x2
dx3/dt =  k3 x2
  • k1: rate of uptake from blood by the liver.
  • k2: rate of return from liver to blood.
  • k3: rate of excretion from liver to bile.

The total x1 + x2 + x3 = X0 stays constant (mass balance), which is a useful check on the model.

Analog computer model

Each integrator gives the output with sign inverted, so the integrator outputs are -x1, -x2, -x3. Coefficient potentiometers are set to k1, k2, k3, and inverters supply the positive signs where needed.

 +x2 -[k2]--\
             +-->[Int 1]--> -x1
 -x1 -[k1]--/
 +x1 -[k1]--\
             +-->[Int 2]--> -x2
 -x2 -[k2+k3]/
 +x2 -[k3]---->[Int 3]--> -x3

(+x1 and +x2 are obtained from -x1 and -x2 by inverters.)

Connections:

  1. Integrator 1 input = -k1 x1 + k2 x2 (the first from the output -x1 through pot k1, the second from +x2 through pot k2). Output -x1; its initial condition is set to -X0.
  2. Integrator 2 input = k1 x1 - (k2 + k3) x2. Output -x2; initial condition zero.
  3. Integrator 3 input = k3 x2. Output -x3; initial condition zero.
  4. Outputs are plotted on a recorder against time.

The recorded curves show x1 falling, x2 rising to a peak and then falling, and x3 rising towards X0. The values of k1, k2, k3 are adjusted till the curves match clinical data, and a diseased liver is shown by a smaller k3.

Pros and cons of the analog computer

Advantages of the analog computer

  • Operates in parallel, so the solution is fast, even real time.
  • Continuous solution; no step size or numerical round-off error.
  • Interactive: parameters can be changed by a knob and the result seen immediately.
  • Shows the physical meaning of the equation; good for control studies.

Disadvantages

  • Accuracy is limited (about 0.1 to 1 percent) by component tolerance, noise and drift.
  • Amplitude scaling and time scaling are needed, and are tedious.
  • The size of the problem is limited by the number of units available.
  • Hard to implement nonlinearities, logical decisions and data storage.
  • Re-patching for a new problem takes time; poor repeatability and documentation.

✓ Answer checked by Claude Opus 5.5 (high effort)

All Simulation and Modeling chapter 3 questions →