From: Psytrix (psytrix_at_o...)
Date: 2001-05-30 17:39:57
> From: Lovass Laszlo <prochess_at_e...>
> To: <buildcheapeeg_at_yahoogroups.com>
test-head
>
> > Is there any sophisticated testing device for EEG, what I could call
> > 'test-head', modelling the electric behaviour of a real head?
> > If not, is it worth do develop one?
> > ..
> > E. g. it would be a creane-form object with networked resistances driven
by
> > a special programmable generator at node points,
Very good idea ,but I guess there's no form to model "the" brain , but only
"a" brain in one (or more) specific working pattern , represented for a set
of observations - like EEG, MRi , Spect .
A solid prototype is a difficult task, but a virtual one isn't so , if
we use differential equations
(using the 'inverse problem calculation" or the "principal components
calculation" see http://www.cnl.salk.edu/~jung/ica.html )- considering a
spherical head approximation and a homogeneous "medium" ; in short , we
take a signal from a large number of electrodes and solving equation
system - in which each electrode stays between 2 ( or more) DDPs acquired
(trivial eg. below .knowing values from Channels 1 and 2 -and having a
reference, one can know the voltage distribution of Electrodes 1,2 and three
( and represent them by resistances)- now transpose that for 3D and a graded
shield...
E1 -Ch 1- E2 -Ch2- E3
50uV 100uV
E2 - E1 = 50
E3- E2 = 100
...
Complete Matlab package available in
http://www.cnl.salk.edu/~scott/ica-download-form.html
Good luck, Waldemar
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:30 BST