<div dir="ltr">Dear Eric,<div><br></div><div><div>> If I calculate the ICA weights using data filtered at 1 Hz, by definition this solution will <i>not</i> give maximally independent components when it is applied to 0.1 Hz filtered data. In the 0.1 Hz filtered data there is low frequency information in the data that ICA was not trained on and therefore doesn't "know" anything about. </div><div><br></div><div>Correct.</div><div><br></div><div>> That low frequency information must, mathematically, end up somewhere in the IC activations when I multiply the unmixing matrix with the 0.1 Hz filtered data. What are the consequences and pitfalls of this? Is it something to worry about? Under what circumstances?</div><div><br></div><div>Let's focus on two scenarios I brought up in my previous response.</div><div>1. If the < 2Hz activity is temporally correlated with > 2Hz activity -> will be decomposed altogether with no conflict.</div><div>2. If the < 2Hz activity is independent of > 2Hz activity -> depending on the source of <2Hz activity, the ICs localized nearby got affected.</div><div><br></div><div>Imagine the Scenario 2 in this way: You are in the dinner party. On a buffet table, there are big four plates of salad, steak, salmon, and desert. You want to take some salad to your place. The moment you reached your arm to the salad plate--power outage happened. All the lights are off. It's pitch dark, but you decided to take the salad anyway. You know where it the salad plate is, and you have a tong and a plate in your hand, so why not? So you take the salad to your plate in the darkness, enjoying the unexpected inconvenience. Now the light is back. Guess what you find in your plate--you have the salad, but it is soaked with mushroom soup! Probably someone spilled his bowl of mushroom soup over the salad plate by accident. Now you eat salad unexpectedly mixed with mushroom soup.</div><div><br></div><div>This is what happens when > 2 Hz ICA is applied back to > 0.1 Hz data. The unexpected mushroom sour contamination represents < 2 Hz activities. The 2-Hz highpassed ICA could only see the four plates, and did not know about the mushroom soup contamination. If you apply ICA again with full data including the mushroom soup contamination, most likely the mushroom-soup contaminated foods are distinguished as a new independent component.<br></div><div><br></div><div>> My guess is that the low frequency information gets divided across the ICs in a somewhat a priori unpredictable way depending on the scalp distribution of the low frequency information and the scalp distribution of the ICs. </div><div><br></div><div>In the example above, it depends on which plate the mushroom soup contaminate. It may hit only the salad plate, or it may hit three plates, etc.</div><div><br></div><div>> If I then remove one of the ICs, I may remove some of this low frequency information with it.</div><div><br></div><div>If you replace the salad plate with a new one, then the mushroom-soup-contaminated salad is also gone.</div><div><br></div><div>> If the low frequency information in question is purely noise, this could lead to some pattern of noise/artifact in the data that is hard to interpret or move noise to electrodes that didn't originally include it.</div><div><br></div><div>Correct.</div><div><br></div><div>> One of the reasons for using the 0.1 Hz filter is that part of the effects I am interested in (e.g., later ERP components) contain information below 1 Hz (see Tanner et al., 2015). </div><div><br></div><div>ERP indeed seems a broadband phenomenon. Broad toward the lower end.</div><div><br></div><div>> Can I be confident that ICA does a good job of isolating artifact from neural activity of interest if part of that activity of interest was not present in the training dataset?</div><div><br></div><div>To determine this, we need a study to determine SNR in the infraslow EEG. We SCCN people tend to say 'there is more noise in infraslow' but apparently there are convincing infraslow EEG studies. But I don't know how much signal and how much noise are in the infraslow range. Another reason why we SCCN people cut infraslow range is because ICA is biased toward power. EEG signal has 1/f power spectral density, which means the lowest frequency tends to have highest impact to ICA results, although we are *usually* more interested in alpha and theta which are main contributors to ERPs.</div><div><br></div><div>One suggestion you may want to try is that you separate the data into <2Hz and >2Hz, and perform ICA separately. You have to adjust data rank when in doing so, the filtering can affect data rank, particularly for < 2Hz data. Then you can see what you are excluding. You can also see if any components found in < 2Hz ICA is similar/dissimilar to > 2Hz ICA.</div><div><br></div><div>Makoto</div></div><br><div class="gmail_quote"><div dir="ltr">On Tue, Feb 27, 2018 at 9:14 AM Eric Fields <<a href="mailto:eric.fields@bc.edu">eric.fields@bc.edu</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div>Hi Jumana and others,<br><br></div>To be clear, my question isn't about implementation (I know it is relatively easy to calculate ICA weights from one dataset and apply them to another). My question was about how this effects the data.<br><br></div>If I calculate the ICA weights using data filtered at 1 Hz, by definition this solution will <i>not</i> give maximally independent components when it is applied to 0.1 Hz filtered data. In the 0.1 Hz filtered data there is low frequency information in the data that ICA was not trained on and therefore doesn't "know" anything about. That low frequency information must, mathematically, end up somewhere in the IC activations when I multiply the unmixing matrix with the 0.1 Hz filtered data. What are the consequences and pitfalls of this? Is it something to worry about? Under what circumstances?<br><br>My guess is that the low frequency information gets divided across the ICs in a somewhat a priori unpredictable way depending on the scalp distribution of the low frequency information and the scalp distribution of the ICs. If I then remove one of the ICs, I may remove some of this low frequency information with it. If so:<br><ol><li>If the low frequency information in question is purely noise, this could lead to some pattern of noise/artifact in the data that is hard to interpret or move noise to electrodes that didn't originally include it.</li><li>One of the reasons for using the 0.1 Hz filter is that part of the effects I am interested in (e.g., later ERP components) contain information below 1 Hz (see Tanner et al., 2015). Can I be confident that ICA does a good job of isolating artifact from neural activity of interest if part of that activity of interest was not present in the training dataset?</li></ol><br></div><div>Have these issues been addressed anywhere in the literature or does anyone have recommendations?<br></div><div><br></div><div>Eric<br></div><div class="gmail_extra"><br clear="all"><div><div class="m_-7645931956761137640m_-3505749243944693018gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><span>-----<br>Eric Fields, Ph.D.<br>Postdoctoral Fellow<br><a href="https://www2.bc.edu/elizabeth-kensinger/" target="_blank">Cognitive and Affective Neuroscience Laboratory</a>, Boston College<br><a href="http://www.brandeis.edu/gutchess/" target="_blank">Aging, Culture, and Cognition Laboratory</a>, Brandeis University<br><a href="mailto:eric.fields@bc.edu" target="_blank">eric.fields@bc.edu</a></span></div></div></div></div></div></div></div>
<br><div class="gmail_quote">On Wed, Feb 21, 2018 at 1:00 PM, Ahmad, Jumana <span dir="ltr"><<a href="mailto:jumana.ahmad@kcl.ac.uk" target="_blank">jumana.ahmad@kcl.ac.uk</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div dir="ltr">
<div id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910divtagdefaultwrapper" style="font-size:12pt;color:#000000;font-family:Calibri,Helvetica,sans-serif" dir="ltr">
<p style="margin-top:0;margin-bottom:0">It depends if you want to examine component activity below 1Hz. Most artifacts of interest, such as blinks and saccades should be higher frequency etc. </p><span>
<p style="margin-top:0;margin-bottom:0"><br>
</p>
<div id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910Signature">
<div id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910divtagdefaultwrapper" style="font-size:12pt;color:rgb(0,0,0);background-color:rgb(255,255,255);font-family:Calibri,Arial,Helvetica,sans-serif,EmojiFont,"Apple Color Emoji","Segoe UI Emoji",NotoColorEmoji,"Segoe UI Symbol","Android Emoji",EmojiSymbols">
<p></p>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><b>------------------------------------------</b></span></font></span></font></div>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><b><span style="background-color:rgb(255,238,148)">Jumana
Ahmad</span></b></span></font></span></font></div>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB">Post-Doctoral Research Worker in Cognitive Neuroscience </span></font></span></font></div>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><i>EU-AIMS Longitudinal European Autism Project (LEAP)
& SynaG Study</i></span></font></span></font></div>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB">Room <a href="https://maps.google.com/?q=M1.26&entry=gmail&source=g" target="_blank">M1.26</a>.Department of Forensic and Neurodevelopmental
Sciences (PO 23) | Institute of Psychiatry, Psychology & Neuroscience | King’s College London | <a href="https://maps.google.com/?q=16+De+Crespigny+Park+%7C+London+SE5+8AF&entry=gmail&source=g" target="_blank">16 De Crespigny Park | London SE5 8AF</a></span></font></span></font></div>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> </span></font></span></font></div>
<div style="color:rgb(0,0,0);font-family:Helvetica;margin:0px"><font face="Calibri,sans-serif" size="2"><span style="font-size:11pt"><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><b>Phone:</b></span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> 0207
848 5359| </span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><b>Email:</b></span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> <span style="background-color:rgb(255,238,148)">jumana.ahmad</span></span></font><a href="mailto:antonia.sanjose@kcl.ac.uk" id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910LPNoLP" target="_blank"><font color="purple" size="2"><span style="font-size:10pt" lang="en-GB">@kcl.ac.uk</span></font></a><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> | </span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><b>Website:</b></span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> </span></font><a href="http://www.eu-aims.eu/" id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910LPNoLP" target="_blank"><font color="purple" size="2"><span style="font-size:10pt" lang="en-GB">www.eu-aims.eu</span></font></a><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> | </span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"><b>Facebook:</b></span></font><font color="#1F497D" size="2"><span style="font-size:10pt" lang="en-GB"> </span></font><a href="http://www.facebook.com/euaims" id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910LPNoLP" target="_blank"><font color="purple" size="2"><span style="font-size:10pt" lang="en-GB">www.facebook.com/euaims</span></font></a></span></font></div>
<br>
<p></p>
</div>
</div>
</span></div><span>
<hr style="display:inline-block;width:98%">
<div id="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910divRplyFwdMsg" dir="ltr"><font style="font-size:11pt" face="Calibri, sans-serif" color="#000000"><b>From:</b> eeglablist <<a href="mailto:eeglablist-bounces@sccn.ucsd.edu" target="_blank">eeglablist-bounces@sccn.ucsd.edu</a>> on behalf of Eric Fields <<a href="mailto:eric.fields@bc.edu" target="_blank">eric.fields@bc.edu</a>><br>
<b>Sent:</b> 21 February 2018 03:45:30<br>
<b>To:</b> EEGLAB List<br>
<b>Subject:</b> [Eeglablist] Baselining and filtering for ICA with epoched data</font>
<div> </div>
</div>
</span><div><div class="m_-7645931956761137640m_-3505749243944693018h5"><div>
<div dir="ltr">Hi,<br>
<br>
I know there have been other threads related to this, so I apologize if this has been addressed directly and I missed it.<br>
<br>
Groppe et al. (2009) showed that ICA gives more reliable results if you use the full epoch instead of the prestimulus period to baseline. The reason generally given for this is that baseline correction changes the scalp distribution of sources depending on
what is happening in the baseline period. By this logic, using the full epoch should improve ICA (because longer periods are less affected by random variations), but no baseline correction at all should be even better.<br>
<br>
Meanwhile, Winkler et al. (2015) have suggested that ICA works best on data high pass filtered at 1-2 Hz.<br>
<br>
Assuming I prefer to use a 0.1 Hz high pass filter (because of distortions 1 Hz filters can cause in the ERP: Tanner et al., 2015), I have two questions:<br>
<br>
<ol>
<li>Does the removal of additional low frequency noise you get from using a full epoch baseline (vs no baseline) outweigh the downsides of baseline correction for ICA?</li><li>Alternatively, is it appropriate to apply a 1 or 2 Hz filter to the data used for ICA training, and then apply the ICA solution to an EEGset filtered at 0.1 Hz? Winkler et al. suggest this, but what happens to the low frequency information in the data when
the ICA solution that has been learned without it is applied? Can this cause problems?<br>
</li></ol>
<br>
Thanks!<br>
<br>
Eric<br>
<br clear="all">
<div>
<div class="m_-7645931956761137640m_-3505749243944693018m_2974847851183948910x_gmail_signature">
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr"><span>-----<br>
Eric Fields, Ph.D.<br>
Postdoctoral Fellow<br>
<a href="https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww2.bc.edu%2Felizabeth-kensinger%2F&data=01%7C01%7Cjumana.ahmad%40kcl.ac.uk%7C60790510144d4d17f8f608d579548d03%7C8370cf1416f34c16b83c724071654356%7C0&sdata=Vmoa8t0Q6S95V85lnDxKRWHj4Wt%2B2Pw87O2v5nSZE%2BM%3D&reserved=0" target="_blank">Cognitive
and Affective Neuroscience Laboratory</a>, Boston College<br>
<a href="https://emea01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.brandeis.edu%2Fgutchess%2F&data=01%7C01%7Cjumana.ahmad%40kcl.ac.uk%7C60790510144d4d17f8f608d579548d03%7C8370cf1416f34c16b83c724071654356%7C0&sdata=KkZ5TQyb7BlOduU%2BXU66q2I5qJJXDvjk%2BPguBy3veQI%3D&reserved=0" target="_blank">Aging,
Culture, and Cognition Laboratory</a>, Brandeis University<br>
<a href="mailto:eric.fields@bc.edu" target="_blank">eric.fields@bc.edu</a></span></div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div></div></div>
</blockquote></div><br></div></div>
_______________________________________________<br>
Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" rel="noreferrer" 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></blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr">Makoto Miyakoshi<br>Swartz Center for Computational Neuroscience<br>Institute for Neural Computation, University of California San Diego<br></div></div></div>