[Eeglablist] How to compute Grand ERP on data with different number of channels

[Stephen Politzer-Ahles]

I'm not sure about the details of the ERPLAB functions (others in the list
might be able to give more specific feedback on that), but in general there
are three simple ways to do this.

1) Interpolate the missing channels so that all the channels in all the
sessions line up correctly again.
2) If one channel is removed from one session, remove it from all sessions
(so that all the channels in all sessions line up correctly again)
3) Replace the missing channel with a placeholder (e.g. a row of all
`NaN`s) so that all the channels in all sessions line up correctly again,
and then average while ignoring the missing channels

All of these are pretty straightforward to do in MATLAB basic functions,
regardless of whether ERPLAB has built-in functions for them or not.

As for ICA, in a case like this you generally would run separate ICA
decompositions on separate sessions. Concatenating all the sessions and
then running ICA on that will generally wind up giving you multiple ICs for
each real "component" (e.g., if you have three sessions, you might end up
with three slightly different "blink" components) since the cap is probably
at a slightly different angle each time you put it on again.



---
Stephen Politzer-Ahles
The Hong Kong Polytechnic University
Department of Chinese and Bilingual Studies
http://www.mypolyuweb.hk/~sjpolit/
<http://www.nyu.edu/projects/politzer-ahles/>

On Tue, Jul 18, 2017 at 1:18 AM, mahsa shalchy <mahsa.shalchy at gmail.com>
wrote:

> Dear Experts,
>
> I would like to compute the Grand ERP of my data using ERPLAB, however,
> I'm encountering an error since after bad channel rejection, I ended up
> with different number of channels for different sessions. Also, the
> channels that I rejected, are not the same channels across sessions, i.e.
> one session might lose T7 whereas the other may lose PO7.
>
> How could I get the ERPLAB to compute Grand average ERP (without doing any
> channel interpolation )?
>
> Finally, I was wondering if it is meaningful to run ICA on this dataset,
> since I am lacking different channels for different sessions which yields
> slightly different components.
>
> Many thanks,
> Mahsa
>
> _______________________________________________
> 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
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170719/658d16ab/attachment.html>