[Eeglablist] ICA: include linked mastoids?

[Makoto Miyakoshi]

Dear Clemens,

Even though I do have a case I am investigating regarding the same
re-referencing issue after adding an additional channel, the rule of thumb
is that 1) do not add any additional channel 2) after re-referencing,
reject one channel to match the rank (or use PCA option when running ICA to
specify channel-1 rank).

> Perhaps not the two mastoids, thus 63 channels for the ICA?

If you say this because 65-2 == 63, you reason in a wrong way. The two
channels you think here must be the mastoids, but for any re-referencing,
rank reduction is just by 1. In your case, if there are 64 channels for
data recording, after re-referencing you should reject any one channel (or
PCA to reduce rank) to run ICA so that your rank is 63, hence 64-1 == 63 is
correct (as I said, this reasoning could be changed after the investigation
I mentioned).

Makoto

On Wed, Mar 16, 2016 at 7:27 AM, Clemens DICKHUT <clemens.dickhut at uni.lu>
wrote:

> Dear all,
>
>
>
> I am using a 64 channel acticap, online reference FCz. To recover FCz I
> append a 65th channel and rereference to the average. Then, I rereference
> to linked mastoids (TP9, TP10).
>
> When running an ICA now, I try to include all 65 channels and a window
> popped up saying the rank computation may be inaccurate (?). I changed the
> channel number from 64 to 65 and ran the analyses.  All the resulting
> components where the same (kind of mastoid topgraphy). I am therefore
> asking which electrodes I have to include for the ICA... Perhaps not the
> two mastoids, thus 63 channels for the ICA?
>
>
>
> I am including my code for the rereferencing here
>
>
>
> % Get channel locations, append FCz, set as reference
>
> EEG = pop_chanedit(EEG, 'lookup','C:\\Program
> Files\\MATLAB\\eeglab13_5_4b\\plugins\\dipfit2.3\\standard_BESA\\standard-10-5-cap385.elp'
> ,'append',64,'changefield',{65 'labels' 'FCz'},'lookup','C:\\Program
> Files\\MATLAB\\eeglab13_5_4b\\plugins\\dipfit2.3\\standard_BESA\\standard-10-5-cap385.elp'
> ,'setref',{'1:65' 'FCz'});
>
> [ALLEEG EEG] = eeg_store(ALLEEG, EEG, CURRENTSET);
>
> EEG = pop_chanedit(EEG, 'setref',{'1:65' 'FCz'});
>
>
>
> % Average reference
>
> EEG = pop_reref( EEG, [],'refloc',struct('labels',{'FCz'},'type',{''},
> 'theta',{0},'radius',{0.12662},'X',{32.9279},'Y',{0},'Z',{78.363},
> 'sph_theta',{0},'sph_phi',{67.208},'sph_radius',{85},'urchan',{65},'ref',{
> 'FCz'},'datachan',{0}));
>
> [ALLEEG EEG CURRENTSET] = pop_newset(ALLEEG, EEG, 1,'setname',
> 'average_ref','overwrite','on','gui','off');
>
> EEG = eeg_checkset( EEG );
>
> % Re-reference to linked mastoids (TP9 TP10) and retain the two channels
>
> EEG = pop_reref( EEG, [17 22] ,'keepref','on');
>
> [ALLEEG EEG CURRENTSET] = pop_newset(ALLEEG, EEG, 1,'setname',
> 'ref_linked_mastoids','overwrite','on','gui','off');
>
> [ALLEEG EEG] = eeg_store(ALLEEG, EEG, CURRENTSET);
>
>
>
>
>
> I am looking forward to your ideas.
>
>
>
> Best,
>
> Clemens
>
>
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>



-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20160404/f5af4508/attachment.html>