<font class="Apple-style-span" color="#333399">cheers. see also this recent article if you haven't yet from Makeig et al that supports the idea that ICA is better than PCA at finding realistic eeg sources. </font><span class="Apple-style-span" style="color:rgb(51,51,153)">Just search on the web at PLos if the link is scrubbed.</span><div>
<div><div><font class="Apple-style-span" color="#333399"><br></font></div><div><font class="Apple-style-span" color="#333399"><span class="Apple-style-span" style="color:rgb(34,34,34);line-height:15px;font-family:arial,sans-serif"><h3 class="r" style="font-size:medium;font-weight:normal;padding-top:0px;padding-right:0px;padding-bottom:0px;padding-left:0px;margin-top:0px;margin-right:0px;margin-bottom:0px;margin-left:0px;display:block;overflow-x:hidden;overflow-y:hidden;text-overflow:ellipsis;white-space:nowrap">
<a href="http://www.plosone.org/article/authors/info%3Adoi%2F10.1371%2Fjournal.pone.0030135;jsessionid=3834349B9EE3031E0764F6666B59FEC2" class="l" style="color:rgb(17,34,204)"><em style="font-weight:bold;font-style:normal">PLOS ONE</em>: Independent EEG Sources Are Dipolar</a></h3>
<div class="vspib" tabindex="0" style="background-image:initial;background-color:transparent;height:auto;margin-top:0px;margin-right:0px;margin-bottom:0px;margin-left:0px;min-height:40px;padding-left:9px;padding-right:4px;width:28px;background-repeat:initial initial">
<div class="vspii" style="border-top-width:1px;border-bottom-width:1px;border-left-width:1px;border-top-style:solid;border-bottom-style:solid;border-left-style:solid;border-top-color:transparent;border-bottom-color:transparent;border-left-color:transparent;border-top-left-radius:2px;border-top-right-radius:2px;border-bottom-right-radius:2px;border-bottom-left-radius:2px;border-right-style:none;border-right-width:initial;border-right-color:initial">
<div class="vspiic" style="background-image:url(http://www.google.com/images/nav_logo117.png);background-color:initial;height:13px;margin-left:6px;width:15px;background-repeat:initial initial"></div></div></div><div class="s" style="max-width:42em;color:rgb(34,34,34)">
<div class="f kv" style="color:rgb(102,102,102);display:block;margin-bottom:1px"><cite style="color:rgb(0,153,51);font-style:normal;display:inline-block;margin-bottom:1px">www.<b style="color:rgb(0,125,42)">plosone</b>.org/.../info%3Adoi%2F10.1371%2Fjournal.p<b style="color:rgb(0,125,42)">on</b>...</cite></div>
</div></span></font>
<br><br><div class="gmail_quote">On Wed, Feb 20, 2013 at 5:10 AM, Veerle ROSS <span dir="ltr"><<a href="mailto:veerle.ross@uhasselt.be" target="_blank">veerle.ross@uhasselt.be</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div lang="NL-BE" link="blue" vlink="purple"><div><p class="MsoNormal"><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Dear Tarik<u></u><u></u></span></p><p class="MsoNormal">
<span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Thank you for your quick response.<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">I will look into the ADJUST plugin and the other points you mention.<u></u><u></u></span></p>
<p class="MsoNormal"><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Best Veerle<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p>
<div style="border:none;border-top:solid #b5c4df 1.0pt;padding:3.0pt 0cm 0cm 0cm"><p class="MsoNormal"><b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> Tarik S Bel-Bahar [mailto:<a href="mailto:tarikbelbahar@gmail.com" target="_blank">tarikbelbahar@gmail.com</a>] <br>
<b>Sent:</b> woensdag 20 februari 2013 2:01<br><b>To:</b> Veerle ROSS<br><b>Cc:</b> <a href="mailto:eeglablist@sccn.ucsd.edu" target="_blank">eeglablist@sccn.ucsd.edu</a><br><b>Subject:</b> Re: [Eeglablist] running ICA a second time<u></u><u></u></span></p>
</div><div><div class="h5"><p class="MsoNormal"><u></u> <u></u></p><div><p class="MsoNormal"><span style="color:#333399">Greetings Ross:</span><u></u><u></u></p></div><div><p class="MsoNormal"><u></u> <u></u></p></div><div>
<p class="MsoNormal"><span style="color:#333399">Some brief responses to your questions below, I hope they are helpful. I've numbered the responses according to your questions.</span><u></u><u></u></p></div><div><p class="MsoNormal">
<u></u> <u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">0. if you have note please search eeglab list archives where you will definitely find some past discussions.</span><u></u><u></u></p></div><div>
<p class="MsoNormal"><span style="color:#333399">PLease read through the articles about cleaning with eeglab, which you can easily find on Google Scholar.</span><u></u><u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">Check out the eeglab tutorial and wiki, particular guidelines. </span><u></u><u></u></p>
</div><div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">1a. You can use PCA (see post within last 2 months on this subject based on a "should I PCA and if so what is best method")</span><u></u><u></u></p>
</div><div><p class="MsoNormal"><span style="color:#333399">Reducing by PCA may decrease the "validity" or "truthfullness" of your ICs. </span><u></u><u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">Check Joseph Dien's PCA toolkit for some principled methods to do PCA.<span> </span></span><u></u><u></u></p>
</div><div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">1c. You should find similar ICs after 2 ICAs, as the second ICA decomposition. The ones in the second ICA should be cleaner and more accurate.</span><u></u><u></u></p>
</div><div><p class="MsoNormal"><span style="color:#333399">Some people just publish their first ICA results, as it is not always the case that you gain much from a second ICA.</span><u></u><u></u></p></div><div><p class="MsoNormal">
<u></u> <u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">2a. I assume you have properly cleaned your data before the first ICA. </span><u></u><u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">My recommendation is, if you want to do a second ICA, first consider doing what eeglab documentation recommend: </span><u></u><u></u></p>
</div><div><p class="MsoNormal"><span style="color:#333399">after your first ICA, then do artifactual epoch rejection using ICA-based rejection, then do a second ICA.</span><u></u><u></u></p></div><div><p class="MsoNormal">
<u></u> <u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">2b. it's not clear what your question is, perhaps there is a word missing in your question.</span><u></u><u></u></p></div><div><p class="MsoNormal">
<span style="color:#333399">You could remove noisy or blink or other artifactual components if you want to, go ahead</span><u></u><u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">and do a second ICA, and compare the results to doing it via the method suggested in the response to 2a above.</span><u></u><u></u></p>
</div><div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal"><span style="color:#333399">also:</span><u></u><u></u></p></div><div><p class="MsoNormal">Check out the the ADJUST plugin (amongst others for a variety of cleaning techniques). <u></u><u></u></p>
</div><div><p class="MsoNormal">&<u></u><u></u></p></div><div><p class="MsoNormal">Note there are least three camps:<u></u><u></u></p></div><div><p class="MsoNormal">a. those who use ICA just to deblink or otherwise clean their data, and then they reconstruct the EEG and do their analyses outside of ICA space. [ergo, ICA is a cleaning tool]<u></u><u></u></p>
</div><div><p class="MsoNormal">b. those who use ICA to get ICs that reflect brain dynamics and (often) established ERP components, they do their analyses on ICs [ergo, ICA gives real brain dynamics]<u></u><u></u></p></div>
<div><p class="MsoNormal">c. those who use PCA (with or without ICA) to decompose ERPs into spatial (and/or) temporal components, they do their analyses on PCs<u></u><u></u></p></div><div><p class="MsoNormal"><u></u> <u></u></p>
</div><div><p class="MsoNormal"><u></u> <u></u></p></div></div></div></div></div></blockquote></div><br></div></div></div>