[Eeglablist] Problem processing ​EEG signals

Tarik S Bel-Bahar tarikbelbahar at gmail.com
Fri Mar 3 12:45:20 PST 2017


Hello Tiago, Interesting project! Some notes below that should help you
debug.

0. It's not clear whether you are visually rejecting epochs or
automatically. It does not make too much of a difference, except if it is
visual, then it could be based on differences between researchers, but I
think you have controlled for that already.

0. Replicate your epoch rejection steps both with the GUI and with your
script. Double check that the eegh output is the same across all three
computers, and that the "working" settings from doing it via the gui really
match the script settings.

1. Reinstall fresh matlab and eeglab on each computer. Alternatively kill
all matlab paths, and load up new clean folder of eeglab - same one to each
of the three computers. Don;t use your old eeglab folder. Repeat all checks.

2. Check all settings in eeglab, and force-change them via command line for
each computer. Check that all plugins are the same for each eeglab/computer.

3. Review the epoching step on all three computers. How many epochs?, how
many boundary events?, review command window output [how many samples
dropped if any, etc..?]
4. On all three computers, count number of boundary events within one file,
before and after epoching, and after epoch dropping.
5. Review/compare the specific trial numbers (not how many) that are being
dropped across the three computers.

6. One solution for you is this:
If you can confirm that you truly have the same exact epochs in the file
for which epoch rejection is to occur.
Then all you have to do is generate, one one computer, the numbers of the
epochs that should be dropped, as a list.
Then just use that information whenever you have to repeat the epoching for
that file.
Just have one researcher decide on all the epochs to reject.

7. Google eeglablist to check if your topic has been mentioned before.
8. Check and make an entry in eeglab Bugzilla if the problem is replicating.









On Fri, Mar 3, 2017 at 7:30 AM, tiago lopes <tiago.lopes56 at yahoo.com> wrote:

> Dears EEGLAB list,
> ​Hope you all are doing well.
> I am having a problem processing ​EEG signals of individuals with chronic
> pain​. That's why I would like to know if you ​ever​ had​ a similar issue
> at your laboratory.
>
> ​I first run a script to: establish the sampling rate (600 Hz), filter the
> data (.5 - 40 Hz) and re-ref them to the mean of all 30 electrodes we used,
> define channel locations, divide files in epochs (1280 ms)​, run ICA, and
> remove components 1 and 2.
>
> When I run a second script to remove epochs by amplitude (+/- 750 microV),
> I have the problem of rejecting different number of epochs for some files
> when compared to my colleague analyzing the same data.
>
> My colleague (Researcher A) and I are using the same version of Matlab​
> (2014a)​ and EEGlab​ (13.6.5b)​. We are also using the same script​s​. Note
> that when I say the same script, it really is the SAME in EVERYTHING. We
> check​ed​ line by line and confirm​ed​ that the commands​ written in one
>  script is written exactly in the ​same way on the ​other.​ The only
> difference is the path to open the files (same files, of course).​
>
> However, when we reject epochs, ​there are differences in some of the
> files. For instance, ​for ​one ​file, ​while the Researcher A gets 1 epoch​
> rejected, I get 5 epochs​ rejected​. W​hat ​would be ​the explanation for
> this?
>
> We thought it was an error in data manipulation. Then we run the script in
> DEBUG mode to find possible difference​s​ in the variables generated by the
> scripts. W​e did that simultaneously and ​did not find any differences
> among our variables.
>
> Then I ran th​at same file on the third computer with MATLAB 2015 and it
> rejected a different number of epochs than we had rejected​ (for that file
> that she rejected 1 and I rejected 5, third computer rejected 6)​.
>> Other examples of inconsistency for the number of epochs rejected:
> A) Researcher A rejects 27 epochs, I reject 26 epochs, third ​COMPUTER
> ​rejects ​27 epochs.
> B) Researcher A rejects 4​ epochs, I reject 5 , third ​COMPUTER ​rejects 6.
> B) Researcher A rejects 40 ​epochs, I reject 45, ​third ​COMPUTER ​rejects
> 45.
> D)Researcher A rejects 60 epochs, I reject 60​, third COMPUTER 60 epochs.​
> (for several files we all reject the same number of epochs)​
>
> ​Have you ever had this problem at your lab?
> If yes, how did you solve it?​ If not, would you have any insights about
> what could be happening? What ​would be your advice​ ​about this problem?
>
> Tiago Lopes, MS
> Health Sciencies Institute - Federal University of Bahia
>
> Att. Tiago Lopes
>
> _______________________________________________
> 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/20170303/a834346b/attachment-0001.html>


More information about the eeglablist mailing list