[Eeglablist] Dipfit problem when used on ERP/EEG data

[Arnaud Delorme]

Logan T Trujillo wrote:

>Hi. I have a question about using the dipfit plugin to fit dipoles to
>ERP/EEG data. When I follow the method presented in the eeglab tutorial, I
>get a message saying that the fit has failed. What are the usual reasons
>for this, does this means that the algorithm has not converged properly?
>  
>
Yes, this is correct.

>If so, what is the usual way to try to fix this problem, to adjust the
>shell model settings? If the latter, what is the proper syntax to do this,
>i.e. for statement
>  
>
There is no obvious way to fix the problem. Most likelly your ERP data 
does not fit to a single dipole or a bilateral dipole.

>[ dipole model ] = dipfit_erpeeg(erp(:,pt100), EEG.chanlocs,'settings',X);
>
>where 'settings' activates the parameter option and X is a cell array of
>the model parameters, what is the proper format of this cell array? Could
>someone give a simple example?
>  
>
It is true that it isn't clear. 'Settings' refer to 
pop_dipfit_settings() arguments. I have made that clearer in the 
function header. There is no settings for the algorithm convergence (you 
have to edit the function called eeg_dipole_fit to change them).

>	Also the algorithm does give a dipole fit even though this "fail
>to fit " message occurs. Is this dipole fit acceptable at all?
>  
>
No it isn't (the algorithm did not even reach a local minima). However 
the residual variance it returns should correspond to the current 
position of the dipoles.

Best

Arno

-- 

*Arnaud Delorme, Ph.D.*
Computational Neurobiology Lab, Salk Institute
10010 North Torrey Pines Road
La Jolla, CA 92037 USA

*Tel* : /(+1)-858-458-1927 ext 15/
*Fax* : /(+1)-858-458-1847/
*Web page *: www.sccn.ucsd.edu/~arno <http://www.sccn.ucsd.edu/%7Earno>
*To think upon*:

    I much prefer the sharpest criticism of a single intelligent man to
    the thoughtless approval of the masses.

        /Johann Kepler/