[Eeglablist] ICA artifact removal for simultaneous EEG/fMRI

[Luca Finelli]

Hi Jed,

> Is there a reason a priori to expect that it would not
> work?  Has anyone out there tried it?

We successfully applied ICA to EEG data acquired
in the magnet and could identify several independent components
related to signal sources of artifactual nature, in
particular, the ballistocardiogram you mention.

Subtraction of MR pulse-related artifacts is possible,
although  requires the EEG to be sampled adequately enough
to get the high-frequency, high-amplitude signals observed during the
pulse without aliasing/saturation.

Hope this helps
Luca

______________________________________
Luca A. Finelli, Ph.D.

 Computational Neurobiology Laboratory
 The Salk Institute
 10010 N. Torrey Pines Rd
 La Jolla CA 92037 - USA

 http://www.cnl.salk.edu

 Tel. + 858 - 458 1927 x14
  or  + 858 - 453 4100 x1455
 Fax. + 858 - 587 0417

On Wed, 15 Jan 2003, Jed Meltzer wrote:

> Hello all,
>
> Here's a question for open discussion.  What do you
> think about the possibility of using the ICA-based
> artifact removal procedure outlined in the EEGLAB
> tutorial, for removing the gradient and ballistocardio
> artifacts from EEG aquired inside an fMRI scanner?  I
> know these artifacts are huge compared to the EEG
> signal, but they are quite predictable and stable
> across time, and so it seems to me that they might
> lend themselves quite nicely to removal by subtraction
> of their independent components from the data.  Not
> that I have tried yet though.  I am surprised that the
> Salk group has posted that book chapter "Having your
> voxels and timing them too" dealing with simultaneous
> EEG/fMRI, but they did not elect to use this method.
> Is there a reason a priori to expect that it would not
> work?  Has anyone out there tried it?
>
> -Jed Meltzer
> Yale University
>
> __________________________________________________
> Do you Yahoo!?
> Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
> http://mailplus.yahoo.com
> _______________________________________________
> Eeglablist mailing list