[Eeglablist] Pop_regica and conducting ICA on dataset with EOG channels

Tyler Grummett tyler.grummett at flinders.edu.au
Tue Jul 4 01:24:53 PDT 2017

Dear eeglabers,

In light of the discussion that you should do an ICA on a dataset with EOG channels, assuming they are not bipolar referenced, I was wondering whether if it's potentially sensible to hack the code so that the EOG channels don't get taken out, and the EOG artefact correction is then performed on components that could potentially include EOG components. I originally hacked the code so that I could use my AMICA-produced ICs, I was wondering if I should continuing hacking and prevent it from removing EOG channels.

I assume taking them out is done on purpose, so I'll gladly accept that you shouldn't leave them in for the EOG correction. I just thought I'd ask :)


Sent from my iPhone

More information about the eeglablist mailing list