<div dir="ltr"><div><div>Dear Giovanni, and your topographies seem dipolar? seem neuronal? If you have bad topographies you will have high residual variance</div></div><div><br></div><div><a href="http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0030135">http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0030135</a><br></div><div><br></div><div>Best,</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Sep 21, 2016 at 7:30 AM, Giovanni Vecchiato <span dir="ltr"><<a href="mailto:giovanni.vecchiato@gmail.com" target="_blank">giovanni.vecchiato@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div lang="IT" link="blue" vlink="purple"><div><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Dear all,<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">I’m performing an ICA followed by a dipole source localization (DIPFIT) and I’m not satisfied by the outcome of the analysis because it returns too high values of residual variance (RV%).<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Here the details of the EEG signal acquisition and processing (eeglab v13.6.5b):<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>1.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Data acquisition with EGI (srate = 500 Hz; 129 channel): <u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>a.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">two blocks of an execution task (EX) + two blocks of motor imagery task (IM) (overall, around 17 minutes of recording)<u></u><u></u></span></p><p style="margin-left:72.0pt"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>2.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Signal processing<u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>a.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">PREP pipeline with line frequency removed = [50 100 150 200]; the other parameters are the default ones<u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>b.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">high pass filtering @ 0.1 Hz<u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>c.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">data segmentation in EX and IM conditions ([-1, 6] seconds), 40 trials per condition<u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>d.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">PCA on the IM dataset (35 trials remained after trial rejection)<u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>e.<span style="font:7.0pt "Times New Roman"">    </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Extended ICA with 27 PCs to retain (which explain the 99% of the variance)<u></u><u></u></span></p><p style="margin-left:72.0pt"><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><span>f.<span style="font:7.0pt "Times New Roman"">     </span></span></span><u></u><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">DIPFIT analysis using the MNI template and manual co-registration (the actual EGI channel positions are correctly loaded) returning an average residual variance of (0.5 +/- 0.2) which values seem too high to be taken into account for a following analysis of ICs.<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">I have tried several modification to the above processing chain (e.g. low pass filtering @ 45 Hz; ICA on a larger number of PCs and on a larger number of trials (80); ICA performed with AMICA; ICA performed with Brain Vision Analyzer – Extended ICA) but with no improvement of the results. I have also tried on different datasets and tasks with the same results.<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Do you all have any suggestion to achieve lower DIPFIT residual variance which are plausible from the neurophysiological point of view?<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">I can also share a sample dataset if you consider it useful.<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Thanks in advance,<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Best,<u></u><u></u></span></p><p class="MsoNormal"><span lang="EN-US" style="font-size:10.0pt;font-family:"Verdana","sans-serif"">Giovanni <u></u><u></u></span></p><p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Verdana","sans-serif""><u></u> <u></u></span></p></div></div><br>______________________________<wbr>_________________<br>
Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" rel="noreferrer" target="_blank">http://sccn.ucsd.edu/eeglab/<wbr>eeglabmail.html</a><br>
To unsubscribe, send an empty email to <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu">eeglablist-unsubscribe@sccn.<wbr>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">eeglablist-request@sccn.ucsd.<wbr>edu</a><br></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature" data-smartmail="gmail_signature">John Ochoa<br>Docente de Bioingeniería<br>Universidad de Antioquia<br></div>
</div>