dipole model

[Robert Oostenveld]

Hi Punita,

On 5 Jun 2006, at 19:05, Punita Christopher wrote:
> Has anyone seen any literature that deals with a spherical multiple
> layer
> head model in which the dipole spans across layers i.e. the source and
> sink of the dipole are in separate layers.

In the dipole model the source and sink are infintely close together.
The difference on the scalp between a dipole and a source+sink model
with only a few mm in between will be negligible. A source+sink model
with more distance between them can be modelled as such, it requires
the solution of the monopole in a sphere (i.e. the desired source
+sink consists of the superposition of two monopoles). I am not sure
whether an analytical solution for that exsists in a 3- or 4-shell
model, but I would would guess that it does (*). For a BEM model it
is trivial to replace the dipole infinite medium potential with a
monopole infinite medium potential, i.e the BEM formulation
(separation of source and volume conduction) allows the specification
of arbitrary sources.

best
Robert

*) I don't know literature for a monopole in a multi-shell sphere,
but I would expect it to be in either textbooks as a didactical
example or in older articles (70s or 80s). You may want to search for
multipole expansions, which are described in more recent literature,
the monopole is the first term (0th order) of that, the dipole the
second term. Perhaps "Phys Med Biol. 2001 Aug;46(8):2255-66. A study
of equivalent source techniques for high-resolution EEG imaging. Yao
D, Zhou Y, Zeng M, Fan S, Lian J, Wu D, Ao X, Chen L, He B." could
provide more insight.