[Eeglablist] Problem with ICLabel and Bilateral dipoles (Makoto's Preprocessing Pipeline)

[Luca Pion-Tonachini]

Hi Mahjabeen,

ICLabel neither computes dipoles, nor uses them when labeling. It is up to you to compute dual dipole fits on any ICs that you think require it if you wish to see such a fit in the IC properties plot. If you did a two dipole fit but are not seeing that fit in the properties window, then there may be a bug in the code. Could you share what you did?

Again, none of this will affect how ICLabel classifies components. 

Luca

> On Apr 2, 2020, at 09:53, Delorme, Arnaud <adelorme at ucsd.edu> wrote:
> 
> Dear Mahjabeen,
> 
>> 1.  I found this from SCCN tutorial's DIPFIT section
>> "In case your channel location are the same as the one of the model. If all your electrode locations are within the International 10-20 System, you may use the standard channel coordinates associated with the head model. In this case, no co-registration is required. "
>> My device is 10-10 configuration, so do I still need to co-register ?
> 
> 10-10 and 10-5 are extension of the 10-20 system (https://www.ncbi.nlm.nih.gov/pubmed/17207640) so in your case, you may use template locations.
> 
>> 2.  When I run IClabel I see only single dipoles, even though I followed the instructions for both single and then searching the bilateral dipoles. So If I use single dipoles only, is it giving me wrong labels ?
> 
> This is a question for Luca copied.
> 
>> 3.  I saw there are some updates in dipfit, iclabel and fieldtrip plug-ins. I am using MATLAB R2019b and EEGLABv14.1.2. By any chance, are these problems happening for not updating ? I did not update yet, cause I am not sure how compatible are those updates with my version of EEGLab.
> 
> Yes, ICLabel is more thoroughly tested with EEGLAB 2019.1. All updates should be fully backward compatible (except for fixing bugs). In other words, your scripts should work in EEGLAB 2019 with zero modification.
> 
> Arno
>