[Eeglablist] dipole fit parameter for children

[arno]

Zeynep, a post-doc in Scott's lab found out this reference below (end of this email). To use it in EEGLAB (dipfit 2), you need to edit the model file (it is a standard Matlab file). 

>> load('-mat', 'eeglab5.03/plugins/dipfit2.0/standard_BESA/standard_BESA.mat'); <mailbox:///C%7C/Documents%20and%20Settings/Arnaud%20Delorme/Application%20Data/Mozilla/Profiles/default/9wegoinz.slt/Mail/pop3.sccn.ucsd-1.edu/Drafts?number=304124&part=1.2&filename=standard_BESA.mat>
>> vol % contains volume information
>> vol.r % contains the 4 spheres diameters

For 4 year old for instance, enter the radii for the brain, cerebrospinal fluid, skull and scalp (see measures at the end of this email).

>> vol.r = [head-scalp-skull-fluid   head-scalp-skull  head-scalp    head]
>> vol.r = [136.1-3.6-3.1-2     136.1-3.6-3.1   136.1-3.6    136.1]/2;

above I used 1 mm (x2) for the fluid thickness as in the standard model.
Then resave the model

>> save('-mat', 'eeglab5.03/plugins/dipfit2.0/standard_BESA/standard_BESA_child4.mat'); <mailbox:///C%7C/Documents%20and%20Settings/Arnaud%20Delorme/Application%20Data/Mozilla/Profiles/default/9wegoinz.slt/Mail/pop3.sccn.ucsd-1.edu/Drafts?number=304124&part=1.2&filename=standard_BESA.mat>

And use this file as a custom model in the dipole settings graphic interface.

Note that electrode radii in the EEGLAB dataset should be 136.1/2 (unless they are scanned).

To display the result,

1) If you can find an average child MRI (or individual child MRI) that is coregistered with the MNI model (same unit etc...), then you may simply use it as input into the dipplot function and you are done. 

2) If you cannot find that, then you need coregister with the spherical model with your MRI. For this, you may use the coregister function. Enter the child MRI mesh (I have a Matlab function I could share to extract the mesh from the MRI or you may use brainvisa or freesurfer to do that) along with electrode position in spherical coordinates. Use the function to align the fiducials with the head. The transformation matrix returned by coregister may be input into dipplot ('transform' input) which will convert spherical coordinates to mri coordinates for the dipplot function.

Arno

------------------------------------------------------------------------

There is a study on SAR levels due to RF emissions from mobile phones on
children and adults (Anderson, 2003). They have used head modes reported
in Farkas (1994), based on an average of male and female data. The head
models are as below:

model     head diameter (mm)  scalp thickness (mm) skull thickness (mm)
4 yo              136.1            3.6                 3.1
8 yo              140.6            4                   3.5
12 yo             143.3            4.4                 3.9
16 yo             147.3            4.8                 4.3
adult             147.6            5.2                 4.7

Is it possible to define a spherical head model in eeglab (by entering
only the radii of the spheres)?

Zeynep.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20070128/ca5d6f04/attachment.html>