indii::fmri::hemodynamic::LogNeuralBalloonModel

LogNeuralBalloonModel Class Reference

Inherits BalloonModel.

Inheritance diagram for LogNeuralBalloonModel:

Inheritance graph
[legend]
List of all members.

Detailed Description

Balloon model, driven by neural activity, and working in log space.

Author:
Lawrence Murray <lawrence@indii.org>
Version:
Rev
285
Date:
Date
2007-07-20 17:25:40 +0100 (Fri, 20 Jul 2007)
This class modifies NeuralBalloonModel so that $f$, $v$ and $q$ vary in log space:

\begin{eqnarray*} \frac{d \ln f}{dt} &=& \frac{\dot{f}}{f} \\ \frac{ds}{dt} &=& \dot{s} \\ \frac{d \ln q}{dt} &=& \frac{\dot{q}}{q} \\ \frac{d \ln v}{dt} &=& \frac{\dot{v}}{v} \end{eqnarray*}

References to $f$, $q$ and $v$ in biophysical functions are substituted with the identities $f = e^{\ln f} $, $q = e^{\ln q} $ and $v = e^{\ln v}$.

Working in log space ensures that $f$, $q$ and $v$ are always positive. This is particularly important in a stochastic setting, where noise may otherwise cause them to become negative, such as in the case of the Wiener process or discretised Gaussian noise. Introducing such noise into the log space instead ensures positivity.

See also:
FlowBalloonModel for more information on the usage of this model.

Definition at line 44 of file LogNeuralBalloonModel.hpp.

Public Types

enum  StateVariable { LN_Q, LN_V, LN_F, S }
 State variable indices. More...
enum  BiophysicalParameter {
  V_0, E_0, TAU_0, ALPHA,
  EPSILON, INV_TAU_S, INV_TAU_F
}
 Parameter indices. More...
enum  BiophysicalFunction { F_IN, F_OUT, E, U }
 Function indices. More...

Public Member Functions

 LogNeuralBalloonModel ()
 Create new model with default biophysical parameters and functions as defined in LogNeuralBalloonModelDefaults, LogFlowBalloonModelDefaults, NeuralBalloonModelDefaults and FlowBalloonModelDefaults.
virtual ~LogNeuralBalloonModel ()
 Destructor.
virtual indii::ml::aux::vector suggestInitialState ()
 Provide a suggested initial state.
virtual void calculateDerivatives (double t, const double y[], double dydt[])
 
See also:
indii::ml::ode::DifferentialModel

Member Enumeration Documentation

enum StateVariable

State variable indices.

Enumerator:
LN_Q  $\ln(q) = \ln\left(\frac{Q(t)}{Q_0}\right)$; natural logarithm of the normalised dHb content at the current time.
LN_V  $\ln(v) = \ln\left(\frac{V(t)}{V_0}\right)$; natural logarithm of the normalised volume at the current time.
LN_F  $\ln(f) = \ln\left(\frac{F_{in}(t)}{F_0}\right)$; natural logarithm of the normalised inflow at the current time.
S  $s$; flow inducing signal.

Definition at line 49 of file LogNeuralBalloonModel.hpp.

enum BiophysicalParameter

Parameter indices.

These match NeuralBalloonModel for consistency.

Enumerator:
V_0  $V_0$; volume of the venous compartment at rest.
E_0  $E_0$; oxygen extraction rate of the venous compartment at rest.
TAU_0  $\tau_0 = \frac{V_0}{F_0}$; mean transit time through the venous compartment at rest.
ALPHA  $\alpha$; Grubb's exponent.
EPSILON  $\epsilon$; efficacy with which neuronal activity causes an increase in signal.
INV_TAU_S  $\frac{1}{\tau_s}$; time constant for signal decay.
INV_TAU_F  $\frac{1}{\tau_f}$; time constant for autoregulatory feedback from blood flow.

Definition at line 78 of file LogNeuralBalloonModel.hpp.

enum BiophysicalFunction

Function indices.

These match NeuralBalloonModel for consistency.

Enumerator:
F_IN  $f_{in}(t) = \frac{F_{in}(t)}{F_0}$; normalised flow into the venous compartment.
F_OUT  $f_{out}(t) = \frac{F_{out}(t)}{F_0}$; normalised flow out of the venous compartment.
E  $E(t)$; oxygen extraction rate of the venous compartment.
U  $u(t)$; neural activity.

Definition at line 122 of file LogNeuralBalloonModel.hpp.

Constructor & Destructor Documentation

LogNeuralBalloonModel (  ) 

Create new model with default biophysical parameters and functions as defined in LogNeuralBalloonModelDefaults, LogFlowBalloonModelDefaults, NeuralBalloonModelDefaults and FlowBalloonModelDefaults.

Definition at line 15 of file LogNeuralBalloonModel.cpp.

~LogNeuralBalloonModel (  )  [virtual]

Destructor.

Definition at line 33 of file LogNeuralBalloonModel.cpp.

Member Function Documentation

aux::vector suggestInitialState (  )  [virtual]

Provide a suggested initial state.

This can be passed to the constructor of the ordinary differential equation solver being used on the model, such as indii::ml::ode::AdaptiveRungeKutta.

Returns:
Suggested initial state for the system.

Implements BalloonModel.

Definition at line 49 of file LogNeuralBalloonModel.cpp.

void calculateDerivatives ( double  t,
const double  y[],
double  dydt[] 
) [virtual]

See also:
indii::ml::ode::DifferentialModel

Definition at line 37 of file LogNeuralBalloonModel.cpp.

The documentation for this class was generated from the following files:
Generated on Tue Oct 9 22:02:12 2007 for fmrii fMRI Modelling Library by  doxygen 1.5.2