Let me add my 2 cents :<br><br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word"><div><div><div class="im"><blockquote type="cite">
<div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><span>I'm back to use fieldtrip after a long period, and now everything seems to work (I used .edf file and I had problem with trigger).<br>
</span></div></div></div></div></div></div></blockquote><div><br></div></div><div>congratulations.</div></div></div></div></blockquote><div><br></div><div>Rodolphe: Congratulations too </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div style="word-wrap:break-word"><div><div><div class="im"><br><blockquote type="cite"><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif">
<div><span>1. I'd like to label my channel to select among EEG type, EMG type etc. because my acquire system labeled them all as EEG. When is it specified in cfg ? I can't find it! And if I can't, how can I exclude EMG,TRIGGER
and EOG channels from ICA computation? <br></span></div></div></div></div></div></div></blockquote><div><br></div></div><div>you must have had a consitent way of plugging in the channels into your amplifier, haven't you? if yes, then you should find out which channel corresponds to which row in your EEG data and you can relabel them within Matlab.</div>
<div><br></div><div>Your EMG, TRIGGER and EOG channels should also be easy to detect when looking at your data. but you should get back to your recording configurations (also data acquisition software) if you are interested in the positions / labels of the other EEG electrodes. You may want to keep your EOG channels if they have a common reference like the EEG channels.</div>
</div></div></div></blockquote><div><br></div><div>Rodolphe: Same answer </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word"><div><div>
<div class="im"><br><blockquote type="cite"><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div>2. I have a lot of channel, but ICA gave me back just 33 of them. How can I proceed to obtain a correct ICA from my 56 channels?</div>
</div></div></div></div></div></blockquote><div><br></div></div><div>you mean components? </div></div></div></div></blockquote><div><br></div><div>Rodolphe: Did you use a PCA to reduce your components first? </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div style="word-wrap:break-word"><div><div><div class="im"><br><blockquote type="cite"><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif">
<div>3. is it wrong if I filter my data before ICA?</div></div></div></div></div></div></blockquote><div><br></div></div><div>no.</div></div></div></div></blockquote><div><br></div><div>Rodolphe: I would tend to say yes. ICA needs a maximum of information to separate the components, </div>
<div>See for example Snyder and Foxe (2010) where they used ICA to get components with a dominant of alpha oscillations. They argued that filtering before is making ICA analysis worse</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div style="word-wrap:break-word"><div><div><div class="im"><br><blockquote type="cite"><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif">
<div>4. If I use ICA probably I can't
use other source separation method (like BSS) to analyze my data, am I wrong?</div></div></div></div></div></div></blockquote><div><br></div></div><div>dunno.</div></div></div></div></blockquote><div>Rodolphe : as ICA is a BSS, what other method did you have in mind? </div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word"><div><div><div><br></div><div>good luck,</div><div>n</div><br><blockquote type="cite">
<div class="im"><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><div><div style="font-size:10pt;font-family:arial,helvetica,sans-serif"><div><br></div></div></div></div></div></div></div></blockquote>
</div></div></div></blockquote><div>Rodolphe Nenert, PhD. </div></div><br>