[Eeglablist] TMS artifact removal with constrained ICA

[Zhilin]

Hi, Martin,

I have a matlab code for temporally constrained ICA, which can be downloaded at:
http://dsp.ucsd.edu/~zhilin/cICA.zip

You can apply it on your data by designing a reference signal
according to your TMS artifacts, which is the traditional way when
using constrained ICA. However, as stated in the paper included in the
zip file, you can use another auxiliary semi-blind source extraction
algorithm to automatically obtain a reference signal, and then run the
temporally constrained ICA. Both auxiliary algorithm and the
constrained ICA are included in the zip file.

If you have any question, feel free to contact me.

Best regards,
Zhilin




On Mon, Jun 13, 2011 at 10:43 AM, Martin Wiener
<wimartin at psych.upenn.edu> wrote:
>
> Hello,
> Has anyone out there analyzed data in EEGLAB with a simultaneous
> transcranial magnetic stimulation (TMS) and EEG paradigm?  There are a
> number of methods out there that I've seen for removal of the artifact
> generated by TMS, but I'm curious if anyone has implemented them
> specifically in EEGLAB.  I'm currently analyzing data from a study I
> conducted using TMS/EEG, but I've found EEGLAB to be less ideal for artifact
> removal than other software alternatives; however, I prefer to use EEGLAB
> for my analyses and so would like to get this to work.  The reason for
> EEGLAB being less ideal is that running ICA doesn't sufficiently remove the
> artifact; the amplitude of the TMS artifact is generally so large that it is
> difficult to parse out TMS-induced artifacts from neural sources when
> looking at the ICA components.  An alternative method that I've found is to
> conduct a  temporally constrained ICA, wherein a subset of the continuous
> data is marked as artifact and another subset is marked as clean; ICA can
> then be conducted on both sets by maximizing components that are similar to
> the artifact-marked dataset and minimizing components similar to the
> clean-marked dataset, then applying a spatial filter to the continuous data.
>  I've seen this technique written by Mark Pfleiger at Source Signal Imaging
> and have used it to remove the TMS artifact in EMSE, but I'm wondering if
> there is a way I can do this in EEGLAB, either with a plugin that conducts a
> similar method or some other way.  Any help would be great, thanks!
>
> --
> Martin Wiener, M.S.
> Doctoral Candidate in Psychology
> Department of Psychology
> University of Pennsylvania
>
> _______________________________________________
> 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
>