---------- Forwarded message ----------
From: rosalyn moran <[log in to unmask]>
Date: Sun, Dec 9, 2012 at 6:45 PM
Subject: ERP/SEP neuronal model contributing states
To: [log in to unmask]
Cc: Vladimir Litvak <[log in to unmask]>
Dear Melissa,
Thank you for the email - I hope I can clarify the choice of contributions.
In early instantiations of the neural mass models we attributed all
dipolar contributions to pyramidal cells due to their apical dendrites
with common tangential arrangement to the cortical surface.
It is recognised that this induces a strong open field arrangement.
However other cells can contribute even if they display more closed
field dendritic properties (eg inhibitory interneurons), the
contribution from these will depend on a strong temporal correlation
among synaptic events. This is why we reorganised the contributions to
be (a priori) shared among pyramidal (60%) excitatory interneuons
(20%) and inhibitory interneurons (20%). This was potentially very
important when dealing with invasive LFP data.
A recent review article outlining the myriad events that may affect
LFPs and non invasive electrophysiological voltage measures is:
http://www.med.nyu.edu/buzsakilab/content/PDFs/BuzsakiKoch2012.pdf
The numbers you mentioned in your email ought to be 20-20-60 - in
spm_L_priors - see below . You can change the prior values yourself,
and indeed compare model evidences with different contributions - eg
100% pyramidal vs the scheme currently there, that would be the most
formal test of where the signal is coming from. These values are
optimised like the other parameters so maybe looking at the posteriors
would also be informative.
Let me know if there's anything else I can help with.
Bests,
Rosalyn
case{'ERP','SEP'}
%------------------------------------------------------------------
pE.J = sparse(1,[1 7 9],[0.2 0.2 0.6],1,9); % 9 states
pC.J = pE.J/16;
case{'CMC'}
%------------------------------------------------------------------
pE.J = sparse(1,[1 3 7],[0.2 0.8 0.2],1,8); % 8 states
pC.J = sparse(1,[3 7],1,1,8)/16;
case{'LFP'}
%------------------------------------------------------------------
pE.J = sparse(1,[1 7 9],[0.2 0.2 0.6],1,13); % 13 states
pC.J = pE.J/16;
|