[Eeglablist] re-reference to average from Cz

[Jerry Zhu]

Thank you Makoto.

After playing a while and digging into some souce codes and tech notes,
here are some of my experience. Hope they will be helpful to other (present
and future) users:

++++++
size(EEG.data) is independent from channel location information. One can
have 128 channels in the data and have 132 channel information ( in which
case 4 channels do not have actual data).

When exporting from net station as raw using waveform tools/file export
with "export reference channel" (e.g. Cz), the raw file has reference
channel data (flat line, zero) and location information.

When importing to eeglab, eeglab would delete empty data reference channel
(reference channel location is retained: only mark that channel type as
REF, the eeglab main window still shows reference as "unknown").

When eeglab shows the reference is "unknown", one can declare Cz as the
reference channel. But this does not change the data themselves at all; it
just changes the label.

* How to declare a channel as reference channel: (1) insert head model if
necessary (2) you would have the channel location information for your
reference channel from the head model (3) the reference channel is at the
end of the channel list, it is not in the data array (4) set reference for
channel indices, i.e. all channels. once set, in the main window you will
see the reference changes to your specified channels*

Re-reference to average: simply click "Tools | Re-reference | Computer
average", leaving other boxes empty. EEGLAB uses V'i = (Vi-Vref) -
sum(Vi-Vref)/number_of_electrodes, #electrodes = size(EEG.data) = 128 in my
case. See the tech note here
(*ftp://ftp.egi.com/pub/documentation/technotes/AverageReference.pdf
<ftp://ftp.egi.com/pub/documentation/technotes/AverageReference.pdf>*)

"retain old reference channels in data": say re-ref to E55, uncheck this
box, E55 would disappear from the new dataset (it is simply a flat line).
total #channels decreases by 1.

"Add current reference channel back to data": add the data of that channel
back (in case of Cz, 0) and total #channels increases by 1. Computer based
on the changed data (original data + 000...000 from Cz)and increased
#channel (e.g. 129). Use EEG.data(1:8) to see the actual EEG data changes.
++++++

--
Jian Zhu, M.A.
Brain and Cognitive Sciences
Department of Psychology
Southern Illinois University Carbondale
Web: http://zhupsy.com

We have two halves in the brain: left and right. Nothing is right in the
left. Nothing is left in the right.


On Mon, Jun 16, 2014 at 8:49 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
wrote:

> Dear Jerry,
>
> > Should I include Cz when computing average?
>
> No.
>
> > I simply click "Tools | Re-reference | Computer average", leaving other
> boxes empty.
>
> Yes, that's the way to go.
>
> If you want to recover Cz for any reason, do it after above steps. EEGLAB
> allows users to do channel interpolation. In doing so, you need to prepare
> a dataset that has Cz though.
>
> Makoto
>
> On Mon, Jun 16, 2014 at 4:59 PM, Jerry Zhu <jerryzhu at siu.edu> wrote:
>
>> Hi all,
>> My 128-channel data use Cz as the reference channel during collection. I
>> wanna re-reference to average. Should I include Cz when computing average?
>>
>> If not, I simply click "Tools | Re-reference | Computer average", leaving
>> other boxes empty. After that I got a file with 128 channels.
>>
>> If yes, I declare Cz as the reference through the GUI "Edit | Channel
>> location | Set reference". Then in "Tools | Re-reference | Computer
>> average" I also "add current reference channel back to data". After that I
>> got a file with 129 channels (including the Cz).
>> (a side-note, I also found that I can skip the step of declaring Cz as
>> reference and directly add Cz back to data in "Tools | Re-reference |
>> Computer average". So is setting a reference channel necessary to begin
>> with?)
>>
>> Which way is "correct" or "preferred"?
>> --
>> Jian Zhu, M.A.
>> Brain and Cognitive Sciences
>> Department of Psychology
>> Southern Illinois University Carbondale
>> Web: http://zhupsy.com
>>
>> We have two halves in the brain: left and right. Nothing is right in the
>> left. Nothing is left in the right.
>>
>> _______________________________________________
>> 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/20140617/2337d8e2/attachment.html>