[Eeglablist] Polarity after ICA

[Arnaud Delorme]

Dear Stefano,

I apologize for the late reply.

> We are using fastICA to extract sources of somatosensory EPs and use the
> Dipfit2 function in EEGLab to localise the extracted sources.
>
> We import data and the ICA weight matrix using File->Import Data.
>
> Given that ICA weight matrix has undetermined polarity (sign), how does
> dipfit2 in EEGLAB handle the estimation of the orientation of the extracted
> dipolar source in this particular context ?
>   

For component with inverted polarities, Dipfit will find dipoles 
pointing in the opposite direction. This is that simple.
Hope this helps,

Arno