ReducedWongWang models

[JESUS CABRERA ALVAREZ]

Dear all,

Amazing job you do in here supporting the use of TVB. Thanks and congrats. 

I wanted to ask about ReducedWongWang models for those of you that may have used them. I have implemented both ReducedWongWang and ReducedWongWangInhExc, along with their parametrizations described in the original papers: Deco 2013 and Deco 2014, respectively. 

I was expecting to see the models oscillating when integrated with a deterministic scheme but I cannot achieve such a thing. I have been exploring the interactive phase plane and I couldn't find the way. 

I wonder if this is actually right, and these models' interest is based on the oscillations obtained given a specific noise level around an attractor (a fixed point; in fact around the synaptic gating activity level I see for each node):

Deterministic integration

Stochastic integration

Thats my question: should I expect ReducedWongWang models to oscillate with a deterministic integration scheme?

Thanks in advance,

Cheers,

Jesús Cabrera-Álvarez

[Michael Schirner]

Dear Jesús,

thanks for the positive feedback.

Indeed, depending on the model parameters and the network it is embedded
in, the phase space of a ReducedWongWang(InhExc) population shows one or
several fixed points.

For a simulation without noise input this means: if a population reached a
fixed point it will stay there unless there is input from connected
populations that is able to drive it out again.

Consequently, in an uncoupled ReducedWongWang(InhExc) population that
receives no noise there will be no oscillations. Also, if the coupling is
too low there will be no oscillations.

Oscillations result from noise and/or coupling.

Noise will produce oscillations on a fast time scale, which results in
power spectra that resemble the characteristic 1/f slope of
electromagnetic brain signals.

Furthermore, if the large-scale brain network coupling is "just right"
("at the edge of criticality") then slow oscillations on the time scale of
fMRI resting-state oscillations emerge in the network. Furthermore, the
correlation pattern of these slow oscillations (functional connectivity)
would then resemble empirical functional connectivity (depending on the
used structural connectome).

Best wishes,
Micha

> --
> You received this message because you are subscribed to the Google Groups
> "TVB Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to tvb-users+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/tvb-users/550583b6-d0df-4fdd-a84e-70208c26eaabn%40googlegroups.com.
>

[WOODMAN Michael]

Hi

The RWW model itself is one dimensional and will not oscillate alone, though two common exceptions would be stochastic resonance and network setup which creates an oscillator out of two or more nodes.  The IE variant of the model might oscillate alone it the right parameter regime.  

In the TVB the phase plane tool is the right way to examine this. 

Cheers

Marmaduke Woodman, TVB Engineer, INS AMU; +33 4 91 32 42 38 +33 7 67 77 84 72

On 15 Jan 2021, at 07:33, Michael Schirner <m.sch...@fu-berlin.de> wrote:

Dear Jesús,

To view this discussion on the web visit https://groups.google.com/d/msgid/tvb-users/7556.94.134.177.133.1610692411.webmail%40webmail.zedat.fu-berlin.de.

[JESUS CABRERA ALVAREZ]

Thanks for the quick answer.

Ok, my intuition was that it -reducedWongWangInhExc model- should oscillate, as the wilsonCowan model with similar mathematical form. I will keep exploring the phase plane to see where I am getting it wrong. 

Best,

Jesús Cabrera-Álvarez

[WOODMAN Michael]

Hi

If you prefer working with the math directly, you can do a stability analysis at the fixed point and determine which parameters lead to positive real part of the eigenvalues.  This doesn’t guarantee the oscillation in a network but can be useful for setting initial parameters as well. 

Cheers

Marmaduke Woodman, TVB Engineer, INS AMU; +33 4 91 32 42 38 +33 7 67 77 84 72

On 15 Jan 2021, at 09:19, JESUS CABRERA ALVAREZ <jesc...@ucm.es> wrote:

Thanks for the quick answer.

To view this discussion on the web visit https://groups.google.com/d/msgid/tvb-users/c6b3275d-06c4-4320-b364-70817f2b2102n%40googlegroups.com.

[JESUS CABRERA ALVAREZ]

I couldn't get a working point for the reducedWongWangIE model to oscillate: neither working with math nor exploring the phase plane. If anyone worked it out before and could give me a clue or some further advice, I'd be very thankful!

Best,