[Eeglablist] ICA Debate

Philip Michael Zeman pzeman at alumni.uvic.ca
Tue Sep 14 12:40:17 PDT 2010

Hello Ed


I'll make my response in terms of the ICA (runica) algorithm in EEGLab and I will use a food example.


Using runICA (or a pie knife)

(1) disassemble the pie into pieces.  Each piece does not (except for special circumstances) equate to individual EEG channels. Sometimes it might look like only one or a couple channels are involved.  

(2) Removing any of the pieces removes the rejected piece from the pie entirely. Variance of varied amounts on each channel that relates to the component in question will be removed from all channels.

(3) when you re-assemble the remaining pieces into a new pie, you get a different pie. The new pie does not contain the pieces that were removed.


Hence, you can completely remove ocular artifacts from the EEG using this 3 step process.


For example:


Using runICA

(1) decompose you 32 channel EEG dataset into maximally statistically independent components using runICA. (The rank of the data should be 32.)

(2) reject components identified as ocular artifact

(3) re-combine the remaining components to create a new version of the 32 channel EEG dataset. (The rank of the data will now be less than 32.  If you removed 1 component, the rank will be 31.)


 The Special Circumstance above:

Artefacts such as "electrode pops" that are on single channels will have equivalent components that relate specifically to the channel with the 'pop' artefact.



Philip Michael Zeman B.Eng, Ph.D. (aka: International Data Analysis Guru Mercenary for Hire)
Applied Brain and Vision Sciences Inc.
Brain Function Analysis for Novel Paradigms and Serious Games
Analysis of Pharmaceutical Effects on Brain Function
Latest Brain Research Result:
Email:   pzeman at alumni.uvic.ca
Phone: +1-250-589-4234
LinkedIn Profile: http://ca.linkedin.com/in/philipmichaelzeman

  ----- Original Message ----- 
  From: Modestino, Edward J *HS 
  To: eeglablist at sccn.ucsd.edu 
  Sent: Monday, September 13, 2010 1:26 PM
  Subject: [Eeglablist] ICA Debate

  We are having a heated debate here about EEGLAB and ICA.  When you use ICA to remove components in EEGLAB, does it remove these components only from the channels where the component is present or does it remove it from all channels?




  Edward Justin Modestino, Ph.D.

  Postdoctoral Research Associate

  Ray Westphal Neuroimaging Laboratory

  Division of Perceptual Studies

  Department of Psychiatry and Neurobehavioral Sciences

  University of Virginia

  Email: ejm9f at virginia.edu




  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/20100914/1dac2643/attachment.html 

More information about the eeglablist mailing list