[Eeglablist] removing artifacts with ICA using alternative reference.

[Tero Hakala]

Hi,
I have some EEG data referenced to CMS (biosemi cap) and then 
re-referenced to electrode at the nose tip. It seems that the data 
becomes contaminated by muscular artifacts at nose (eye blinks, 
movements etc)  This is evident from constant amplitude jumps that is 
seen across all the channels. (the subjects were kids, so artifacts are 
abundant.)

For comparison purposes the nose reference is preferred.  However, 
artifact removal using ICA does not seem to work too well in this 
situation.  Does it make sense to re-reference to average, run ICA and 
do artifact rejection using this reference to try to clean the nose and 
then re-reference back for further analysis?

  -t