Dear Tarik,<br><br>thank you very much for your answer. <br><br>Preprocessing of my data before 1st ICA consists of highpass filter and Cleanline line noise removal. Eye and muscle artifacts I hope to detect after 1st ICA when I would remove them (not my removing components, but recognizing bad trials). I am not a medical student (electrical engineering is my area), so I don't know how to recognize bad data just looking at it, which is one of the suggested techniques of preprocessing. <br>
<br>> <font color="#333399">Note that there are many choices on how to drop bad epochs when doing your step 4. <br></font>Yes, I have seen that there are 4-5 methods in this step and all of them requireing some parameters to be entered of course. <br>
- Are there any starting values that you suggest for them with which I can start (each method is using std.dev as a main parameter)? I As a result, I know I shouldn't reject more that approx 10% of the data...Makoto suggested to start with std.dev 8...<br>
<br>I have two more question - I read Arno's suggestion regarding referencing to the average<br><br>"Applying an average reference transformation before applying ICA should
not change the ICA decomposition (since it is a linear operation).
However, we have experienced better quality decompositions if we run ICA
prior to applying average reference, so we advise to apply
average-reference transformations after running ICA (the re-referencing
will be also applied to the ICA weigth matrix)"<br><br>- so I suppose that this would be my step No.6? After baseline correction (on the whole epochs) I should to the re-referencing of the data and in the same time it will re-reference ICA weigth matrix?<br>
- In case I run ICA on continous data that consists of parts of the recordings where the subject performed task, is the mentioned procedure the same? Of course then there would be no step 1 (epoching) and is there something that has to be done instead of baseline removal, since I wouldn't have epochs?<br>
<br>Kind regards,<br><br>Ida<br><br>
<br><div class="gmail_quote">On Tue, Mar 5, 2013 at 11:45 PM, Tarik S Bel-Bahar <span dir="ltr"><<a href="mailto:tarikbelbahar@gmail.com" target="_blank">tarikbelbahar@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><font color="#333399">Hi Ida, </font><div><font color="#333399"><br></font></div><div><font color="#333399">Your steps 1 to 4 seem fine, as long as you are feeding cleaned data to the "ICA 1stTime" step.</font></div>
<div><font color="#333399">After step 4, then you can follow basic steps to get ERP, single-trial, frequency, and dipole info on the extra-clean data,</font></div><div><font color="#333399">as described in the eeglab tutorial, and the many articles that use eeglab and ICA for their analysis (you can find all of these on Google Scholar).</font></div>
<div><font color="#333399">Note that there are many choices on how to drop bad epochs when doing your step 4. </font></div><div><font color="#333399"><br></font></div><div>
<font color="#333399">If you have not done the eeglab tutorial, it is recommended that you do it, rather than using your fresh data to figure things out.</font></div><div><font color="#333399">Once you figure out all basic functions and steps with sample data, then I suggest pushing your data through some well-decided steps.</font></div>
<div><font color="#333399"><br></font></div><div><span style="color:rgb(51,51,153)">Regarding the "baseline removal" based on the Groppe article, you should look more closely at those articles to determine exactly</span></div>
<div><font color="#333399">what you are doing and why you are doing it. you may also want to contact the author directly so you have good clarity.</font></div><div><font color="#333399"><br>
</font></div><div><font color="#333399">Good luck with your process and looking forward to hearing of your success.</font></div><div><font color="#333399"><br></font></div>
<div><font color="#333399">Best wishes,</font></div><div><font color="#333399">Tarik</font></div><div><font color="#333399"><br></font><br><div class="gmail_quote"><div><div>
On Sun, Mar 3, 2013 at 5:29 AM, ida miokovic <span dir="ltr"><<a href="mailto:ida.miokovic@gmail.com" target="_blank">ida.miokovic@gmail.com</a>></span> wrote:<br></div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div><div>
<span style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">Dear eeglablist,</span><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
<br></div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">After reading through many q&a here on eeglab list, I found the recommended literature of Groppe et al regarding baseline removal and went through it.</div>
<div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif"><br></div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
My initial plan was to run ICA on the epoched data 1st time, then to remove bad trials (not the bad components) and then to run ICA second time after which I will hopefully obtain better decomposition. Going through q&a, it wasn't clear to me what is the procedure when ICA is applied 2 times. This is short description of the steps as I understand them at the moment:</div>
<div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">1. data epoching (without any baseline removal)</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
2. ICA 1st time</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">3. bad trials removal</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
4. ICA 2nd time</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">5. baseline correction on the whole epochs(?) </div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">And also one more question, if I am interested into finding a certain frequency presence in channels and ICs, is Tools --> Component spectra and maps good option? In the field of "Freq.to analyze" I would put the one of my interest and obtain the components (for example 5 of them) and the channel with the highest power of frequency of my interest. I have to learn more about dipfit, since I think I will actually need the source of the signal of the certain frequency localization...</div>
<div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif"><br></div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
Thank you very much in advance,</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif"><br></div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
Regards,</div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif"><br></div><div style="color:rgb(34,34,34);font-size:13px;font-family:arial,sans-serif">
Ida</div>
<br></div></div>_______________________________________________<br>
Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" target="_blank">http://sccn.ucsd.edu/eeglab/eeglabmail.html</a><br>
To unsubscribe, send an empty email to <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu" target="_blank">eeglablist-unsubscribe@sccn.ucsd.edu</a><br>
For digest mode, send an email with the subject "set digest mime" to <a href="mailto:eeglablist-request@sccn.ucsd.edu" target="_blank">eeglablist-request@sccn.ucsd.edu</a><br></blockquote></div><br></div>
</blockquote></div><br>