[Eeglablist] study ideas

[Robert Thatcher]

Rob,

     ICs are not sources they are mathematicalconstructs used to decompose a mixture of signals into independent factor orcomponents.  It is important to recognizethat ICA is based on two invalidphysiological assumptions:

1.    The source signals are independent of eachother.

2.    The values in each source signal havenon-Gaussian distributions.

ICAdecomposition is a useful academic method to evaluate components or factorsthat are created based on these assumptions but ICs are not themselveselectrical sources of the EEG.  ICA-R orreconstruction of a new time series after removing ICs distorts the original electricalfield produced by the brain and therefore source localization based on the reconstructedtime series is a violation the Lead Field and Poisson’s equation and Green’sfunction which are at the foundation of the inverse solution.

 

Bob



On Monday, August 14, 2017, 1:08:02 PM EDT, Rob Coben <drcoben at gmail.com> wrote:

<!--#yiv9537296241 _filtered #yiv9537296241 {font-family:"Cambria Math";panose-1:2 4 5 3 5 4 6 3 2 4;} _filtered #yiv9537296241 {font-family:Calibri;panose-1:2 15 5 2 2 2 4 3 2 4;}#yiv9537296241 #yiv9537296241 p.yiv9537296241MsoNormal, #yiv9537296241 li.yiv9537296241MsoNormal, #yiv9537296241 div.yiv9537296241MsoNormal {margin:0in;margin-bottom:.0001pt;font-size:11.0pt;font-family:"Calibri", sans-serif;}#yiv9537296241 a:link, #yiv9537296241 span.yiv9537296241MsoHyperlink {color:#0563C1;text-decoration:underline;}#yiv9537296241 a:visited, #yiv9537296241 span.yiv9537296241MsoHyperlinkFollowed {color:#954F72;text-decoration:underline;}#yiv9537296241 span.yiv9537296241EmailStyle17 {font-family:"Calibri", sans-serif;color:windowtext;}#yiv9537296241 .yiv9537296241MsoChpDefault {font-family:"Calibri", sans-serif;} _filtered #yiv9537296241 {margin:1.0in 1.0in 1.0in 1.0in;}#yiv9537296241 div.yiv9537296241WordSection1 {}-->
We have conducted a study assessing resting eeg in adults that were traumatized as children and wish to compare them to control subjects without such problems in their history. Our primary data to analyze is based on 64 ch eeg sampled at 2000 c/s over two separate recordings of 10 minutes in duration. This is resting eeg not erp data.

  

We wish to analyze these data for two primary questions. First, analyze ic’s and determine sources and compare the groups for differences in regions/sources. What would you suggest using for this? Study function? MPT? Other thoughts? 

  

Next, we want to measure the difference between the groups in source derived connectivity. We focus on granger causality using PDC as our primary measure. We often use, on an individual level, a SIFT like application that does this. Suggestions would be welcome. Use SIFT for this group level analysis? Other ideas?

  

Alternatively, we have thought of using graph theory measures but would prefer to do at the source level not channel. Any thoughts?

  

Thanks,

  

Rob Coben, PhD 
_______________________________________________
Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170814/33491d4a/attachment.html>