[Eeglablist] Opinions on Wallstrom, Kass, Miller, Cohn and Fox (2004)?

[Kolbjørn Brønnick]

Dear EEGLABlist subscribers,

I'm trying to understand the strengths and weaknesses of different
methods for correction of ocular artifacts. I have a material with
different patient groups using a mismatch negativity paradigm and an
oddbal distractor paradigm. As some subjects have very rapid/frequent
eye-blink artifacts, I'd prefer to use ICA to correct the artifacts
rather than rejecting trials with artifacts.

However, the article by Wallstrom, Kass, Miller, Cohn and Fox(2004)
"Automatic correction of ocular artifacts in the EEG: A comparison..."
in International journal of psychophysiology, made me doubt the wisdom
of this approach.

Basically, they state that ICA introduced spectral distortion,
especially in the theta band. They used fastICA and extended ICA on 60
seconds (512 hz sample rate) of 21 channel EEG.

Do you have opinions on this article? Isn't a  60 seconds of 21
channel EEG too short for ICA?

Kolbjørn S Brønnick