<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=utf-8"><meta name=Generator content="Microsoft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
p.msonormal0, li.msonormal0, div.msonormal0
{mso-style-name:msonormal;
mso-margin-top-alt:auto;
margin-right:0cm;
mso-margin-bottom-alt:auto;
margin-left:0cm;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
span.E-mailStijl18
{mso-style-type:personal;
color:#1F497D;}
span.E-mailStijl19
{mso-style-type:personal;
font-family:"Calibri",sans-serif;
color:windowtext;}
span.E-mailStijl21
{mso-style-type:personal-reply;
color:#993366;}
.MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;}
@page WordSection1
{size:612.0pt 792.0pt;
margin:70.85pt 70.85pt 70.85pt 70.85pt;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]--></head><body lang=NL link=blue vlink=purple><div class=WordSection1><p class=MsoNormal><span lang=NL-BE style='color:#993366;mso-fareast-language:EN-US'><o:p> </o:p></span></p><div><div style='border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm'><p class=MsoNormal>><br><b>Subject:</b> RE: [Eeglablist] ICA misinformation<o:p></o:p></p></div></div><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>Dear Johanna<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>The model of volume conduction would be completely correct if brains were a kind of balloons filled with saline and sources would be dipoles suspended in it at fixed places with recording electrodes placed on the outside.<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>Although of course volume conduction is a reality in brains (Nunez) we are especially interested in communication transfert between nodes in several well defined networks and delays in the communication from a node A to a node B will be reflected in a non zero phase lag between the signals. Of course one must use a measure of information transfert (such as Nolte’s imaginary part of the coherence or other equal or better parameters to extract the phase diff NEQ zero cf Pasqual Marqui) <o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>So If one does find a phase difference between two scalp electrodes that is different from zero that cannot be caused by volume conduction, can it?<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>I agree that analysis in source space will provide even better fine grained insight in network communication patterns but I am not so convinced that ICA is the best way to pinpoint those network hub sources especially in low channel recordings where there is a real possibility of overrepresentation (more sources then recording electrodes). We know a lot of those network graphs by extensive DTI and fMRI studies (cf Catani, De Schotten) an,d while I agree that ICA is indeed a very elegant statistical technique I think we must also admit that it is subject to limiting conditions that perhaps are more prominent in low grade (19 ch) recordings.<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>A lot of people (I include myself) have been very impressed by the power of ICA to separate sources in the toy model (cocktail party model) where all sources are nicely independent and stationary with well behaving non gaussian distributions but in brains things are seldom that simple and synchronization and network communications can result in lumping of sources into a same component.<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>I prefer swLoreta for source analysis as it more in relation to neurophysiological reality we know from DTI “neuroanatomy”.<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>Again, I absolutely admire the statistical elegance of ICA but feel that we -as a scientific community- must try to resolve all errors and pitfalls that analysis algorithms such as ICA may present to improve robustness and clinical validity of our methods.<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>I find the input to this list very positive and very rewarding.<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'>Sincerely<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='font-size:14.0pt;color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><span lang=EN-US style='color:#1F497D;mso-fareast-language:EN-US'><o:p> </o:p></span></p><p class=MsoNormal><b><span lang=EN-US>From:</span></b><span lang=EN-US> <a href="mailto:eeglablist-bounces@sccn.ucsd.edu">eeglablist-bounces@sccn.ucsd.edu</a> [<a href="mailto:eeglablist-bounces@sccn.ucsd.edu">mailto:eeglablist-bounces@sccn.ucsd.edu</a>] <b>On Behalf Of </b>Johanna Wagner<br><b>Sent:</b> Wednesday, June 14, 2017 11:07 PM<br><b>To:</b> </span><a href="mailto:tarikbelbahar@gmail.com"><span lang=EN-US>tarikbelbahar@gmail.com</span></a><span lang=EN-US>; Brunner, Clemens (</span><a href="mailto:clemens.brunner@uni-graz.at"><span lang=EN-US>clemens.brunner@uni-graz.at</span></a><span lang=EN-US>) <</span><a href="mailto:clemens.brunner@uni-graz.at"><span lang=EN-US>clemens.brunner@uni-graz.at</span></a><span lang=EN-US>>; Tim Mullen <</span><a href="mailto:mullen.tim@gmail.com"><span lang=EN-US>mullen.tim@gmail.com</span></a><span lang=EN-US>>; </span><a href="mailto:billinger.martin@gmail.com"><span lang=EN-US>billinger.martin@gmail.com</span></a><span lang=EN-US><br><b>Cc:</b> eeglablist <</span><a href="mailto:eeglablist@sccn.ucsd.edu"><span lang=EN-US>eeglablist@sccn.ucsd.edu</span></a><span lang=EN-US>><br><b>Subject:</b> Re: [Eeglablist] ICA misinformation<o:p></o:p></span></p><p class=MsoNormal><span lang=EN-US><o:p> </o:p></span></p><div><p class=MsoNormal>I think there is a <span style='font-size:9.5pt'>basic misunderstanding here on what EEG is and what channels actually measure</span> as Clemens wrote <o:p></o:p></p><div><p class=MsoNormal><o:p> </o:p></p><div><p class=MsoNormal>Connectivity should never be computed between EEG channels. Due to volume conduction, signals recorded at all channels are a linear mixture of sources. Signals at all channels are therefore highly correlated. To use No source separation before estimating connectivity is simply wrong from my understanding. <o:p></o:p></p></div><div><p class=MsoNormal>If you reject artifacts with ICA and then go back to the channel space you have the same problem since all brain sources are still mixed at the channel level....<o:p></o:p></p></div></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>see also Clemens Brunners paper (<a href="http://journal.frontiersin.org/article/10.3389/fncom.2016.00121/full" target="_blank"><span style='font-size:9.5pt'>http://journal.frontiersin.org/article/10.3389/fncom.2016.00121/full</span></a>)<o:p></o:p></p></div><div><p class=MsoNormal>and <o:p></o:p></p></div><div><p class=MsoNormal><a href="http://journal.frontiersin.org/article/10.3389/fninf.2014.00022/full">http://journal.frontiersin.org/article/10.3389/fninf.2014.00022/full</a><o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>Johanna<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p><div><p class=MsoNormal>2017-06-14 13:30 GMT-07:00 Tarik S Bel-Bahar <<a href="mailto:tarikbelbahar@gmail.com" target="_blank">tarikbelbahar@gmail.com</a>>:<o:p></o:p></p><blockquote style='border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0cm;margin-bottom:5.0pt'><p class=MsoNormal>Thanks everyone for this cool ongoing conversation. Although not an<br>expert in these issues, it’s exciting for me to know that we will soon<br>very likely have new clear reports from multiple labs that will speak<br>to these general issues and assumptions. Arno’s proof of concept is<br>straightforward, and the article passed on by Clemens certainly lays<br>out some important points. However, to date there is simply not enough<br>recent published work directly examining these issues. In fact, top<br>journals such as Neuroimage regularly publish articles using<br>ICA-cleaned connectivity data. Overall, the field is dependent of<br>valid/robust methods, and of course it’s important to test any/all<br>assumptions as specifically as possible in a replicable/empirical<br>manner.<br><br><br><br>Possible dimensions/constraints to consider?<br><br>Number of channels (as mentioned by Rob Lawson)<br><br>Channel density and relative total-head coverage<br><br>Number/type of artifact ICs removed<br><br>Clarity/robustness of artifacts (e.g., ICs that are mixed vs. ICs that<br>are mixed (containing both artifact and neural info)<br><br>Channel-level vs. ICA-level vs. source-level connectivity metrics<br><br>Length of epochs/trials<br><br>Type of source analysis<br><br>Type of reference (e.g., Chella et al., 2016)<br><br>Type of ICA/blind source separation (e.g., Bridwell et al., 2016;<br>Brain Topography)<br><br>Event-related or resting data<br><br>Signal quality (e.g., gel versus saline, very noisy vs. quite clean)<br><br>Reliability of particular metric<br><br>Type of connectivity metric (e.g., various kinds of phase measures,<br>and the plethora of other connectivity and graph theoretical measures)<br><br>MEG vs. EEG ?<br><br>Sampling rate?<br><br><br><br>Sample articles that seem to use ICA in relation to connectivity<br>metrics are listed below.<br>Moving forward, it may be beneficial to survey these authors and their findings.<br><br>de Pasquale, F., Della Penna, S., Sporns, O., Romani, G. L., &<br>Corbetta, M. (2015). A dynamic core network and global efficiency in<br>the resting human brain. Cerebral Cortex, bhv185.<br><br>Lai, M., Demuru, M., Hillebrand, A., & Fraschini, M. (2017). A<br>Comparison Between Scalp-And Source-Reconstructed EEG Networks.<br>bioRxiv, 121764.<br><br>Colclough, G. L., Woolrich, M. W., Tewarie, P. K., Brookes, M. J.,<br>Quinn, A. J., & Smith, S. M. (2016). How reliable are MEG<br>resting-state connectivity metrics?. NeuroImage, 138, 284-293.<br><br>Kuntzelman, K., & Miskovic, V. (2017). Reliability of graph metrics<br>derived from resting‐state human EEG. Psychophysiology, 54(1), 51-61.<br><br>Siems, M., Pape, A. A., Hipp, J. F., & Siegel, M. (2016). Measuring<br>the cortical correlation structure of spontaneous oscillatory activity<br>with EEG and MEG. NeuroImage, 129, 345-355.<br><br>Toppi, J., Astolfi, L., Poudel, G. R., Innes, C. R., Babiloni, F., &<br>Jones, R. D. (2016). Time-varying effective connectivity of the<br>cortical neuroelectric activity associated with behavioural<br>microsleeps. NeuroImage, 124, 421-432.<br><br>Farahibozorg, S. R., Henson, R. N., & Hauk, O. (2017). Adaptive<br>Cortical Parcellations for Source Reconstructed EEG/MEG Connectomes.<br>bioRxiv, 097774.<br><br>Rueda-Delgado, L. M., Solesio-Jofre, E., Mantini, D., Dupont, P.,<br>Daffertshofer, A., & Swinnen, S. P. (2016). Coordinative task<br>difficulty and behavioural errors are associated with increased<br>long-range beta band synchronization, NeuroImage.<br><br>Cooper, P. S., Wong, A. S., Fulham, W. R., Thienel, R., Mansfield, E.,<br>Michie, P. T., & Karayanidis, F. (2015). Theta frontoparietal<br>connectivity associated with proactive and reactive cognitive control<br>processes. Neuroimage, 108, 354-363.<br><br>Nayak, C. S., Bhowmik, A., Prasad, P. D., Pati, S., Choudhury, K. K.,<br>& Majumdar, K. K. (2017). Phase Synchronization Analysis of Natural<br>Wake and Sleep States in Healthy Individuals Using a Novel Ensemble<br>Phase Synchronization Measure. Journal of Clinical Neurophysiology,<br>34(1), 77-83.<br><br>Vecchio, F., Miraglia, F., Piludu, F., Granata, G., Romanello, R.,<br>Caulo, M., ... & Rossini, P. M. (2017). “Small World” architecture in<br>brain connectivity and hippocampal volume in Alzheimer’s disease: a<br>study via graph theory from EEG data. Brain imaging and behavior,<br>11(2), 473-485.<br><br>Ranzi, P., Freund, J. A., Thiel, C. M., & Herrmann, C. S. (2016).<br>Encephalography Connectivity on Sources in Male Nonsmokers after<br>Nicotine Administration during the Resting State. Neuropsychobiology,<br>74(1), 48-59.<br><br>Vecchio, F., Miraglia, F., Curcio, G., Della Marca, G., Vollono, C.,<br>Mazzucchi, E., ... & Rossini, P. M. (2015). Cortical connectivity in<br>fronto-temporal focal epilepsy from EEG analysis: a study via graph<br>theory. Clinical Neurophysiology, 126(6), 1108-1116.<br><br>Smit, D. J., de Geus, E. J., Boersma, M., Boomsma, D. I., & Stam, C.<br>J. (2016). Life-span development of brain network integration assessed<br>with phase lag index connectivity and minimum spanning tree graphs.<br>Brain connectivity, 6(4), 312-325.<br><br>Chung, Jae W., Edward Ofori, Gaurav Misra, Christopher W. Hess, and<br>David E. Vaillancourt. "Beta-band activity and connectivity in<br>sensorimotor and parietal cortex are important for accurate motor<br>performance." NeuroImage 144 (2017): 164-173.<br><br>Shou, G., & Ding, L. (2015). Detection of EEG<br>spatial–spectral–temporal signatures of errors: A comparative study of<br>ICA-based and channel-based methods. Brain topography, 28(1), 47-61.<br><br>Kline, J. E., Huang, H. J., Snyder, K. L., & Ferris, D. P. (2016).<br>Cortical Spectral Activity and Connectivity during Active and Viewed<br>Arm and Leg Movement. Frontiers in neuroscience, 10.<br><br>van Driel, J., Gunseli, E., Meeter, M., & Olivers, C. N. (2017). Local<br>and interregional alpha EEG dynamics dissociate between memory for<br>search and memory for recognition. NeuroImage, 149, 114-128.<br><br>Castellanos, N.P., Makarov, V.A., 2006. Recovering EEG brain signals:<br>Artifact suppression with wavelet enhanced independent component<br>analysis. J. Neurosci. Methods 158, 300–312.<br>doi:10.1016/j.jneumeth.2006.05.033<br><br>Mehrkanoon, S., Breakspear, M., Britz, J., & Boonstra, T. W. (2014).<br>Intrinsic coupling modes in source-reconstructed<br>electroencephalography. Brain connectivity, 4(10), 812-825.<o:p></o:p></p><div><div><p class=MsoNormal>_______________________________________________<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">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">eeglablist-request@sccn.ucsd.edu</a><o:p></o:p></p></div></div></blockquote></div><p class=MsoNormal><br><br clear=all><o:p></o:p></p><div><p class=MsoNormal><o:p> </o:p></p></div><p class=MsoNormal>-- <o:p></o:p></p><div><div><div><div><div><div><div><div><div><div><p class=MsoNormal><br>Johanna Wagner, PhD<o:p></o:p></p><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal><a href="http://scholar.google.at/citations?user=vSJYGtcAAAAJ&hl=en" target="_blank">http://scholar.google.at/citations?user=vSJYGtcAAAAJ&hl=en</a><o:p></o:p></p></div><p class=MsoNormal><o:p> </o:p></p></div></div></div></div></div></div></div></div></div></div></div></div></div></body></html>