[FieldTrip] EEG Source Reconstruction - Interpolation by Trial?

[Roey Schurr]

Dear Jörn,

Thank you for your kind answer!



>From what we saw on several links it is customary to perform the
interpolation before normalisation, both before statistical analysis (e.g.
http://fieldtrip.fcdonders.nl/tutorial/introduction) or just before
plotting as you mentiones (e.g.
http://fieldtrip.fcdonders.nl/tutorial/beamformer).



The reason we wanted *to interpolate before normalising* is that
interpolation relies on anatomical data, hence it is not possible to use
ft_volumenormalise on pure "source" structure (yet it is possible using it
on "interpolated source" structure).



The reason we wanted to do the *normalisation* *per trial* is the
following: we want to compare the source structure for each condition. In
principle, we could have performed the source analysis over all trials and
get the AVERAGE source per subject per condition, and then compare these
two groups (source condition 1, source condition 2). However, we have too
few subjects, and so we thought it could be best to compare the source
reconstructions of ALL trials over ALL patients in both conditions. Does
this make any sense to you?



Best,

Aia and Roey


On Tue, Sep 3, 2013 at 1:46 PM, "Jörn M. Horschig" <
jm.horschig at donders.ru.nl> wrote:

>  Dear Roey,
>
> of course there is a way, actual several. I don't see a reason why you
> would want to do that though :) The interpolation purpose is mainly for
> plotting, not for performing fancy statistics or other analyses. In case
> you want to go to a finer spatial resolution, I'd do the reconstruction on
> a finer grid (and there, you will be faced with just physical and
> mathematical constraints in terms of spatial accuracy). That way, your
> reconstruction will be at least based on a physical model, not on some
> 'weird' linear interpolation.
>
> Anyway, to answer your question:
> First, there is a cfg.parameter field, with which you can specify the
> parameter. If that does not work, the easiest way for you would be to go to
> the code of ft_sourceinterpolate at line 171, copy the lines up to 184 and
> put these into your own function. You would need to change them
> appropriately, so that you loop over trials. As I said, there are other
> ways, but that would be the most elegant, fastest (computationally) and
> easiest to understand.
>
> Best,
> Jörn
>
>
> On 9/2/2013 4:54 PM, Roey Schurr wrote:
>
>  Hi all,
>
>  We are interested in comparing two source reconstruction calculated
> using EEG records taken in two different conditions (not ERP), from
> different subjects, across the group.
>
>  As we see it, what we must do is the following:
>
>  1) ft_sourceanalysis (for each condition in each subject, in all trials)
> 2) ft_sourceinterpolate (the same)
> 3) ft_volumenormalise (the same)
> 4) ft_sourcegrandaverage
> 5) ft_sourcestatistics
>
>  The problem is ft_sourceinterpolate will not interpolate each trial
> individually, as there is no way to give it the "pow" field per trial.
>
>  Is there a way to performd the interpolation over the anatomycal MRI per
> trial?
>
>  Thank you so much!
> Aia and Roey
>
>
> _______________________________________________
> fieldtrip mailing listfieldtrip at donders.ru.nlhttp://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
>
>
> --
> Jörn M. Horschig
> PhD Student
> Donders Institute for Brain, Cognition and Behaviour
> Centre for Cognitive Neuroimaging
> Radboud University Nijmegen
> Neuronal Oscillations Group
> FieldTrip Development Team
>
> P.O. Box 9101
> NL-6500 HB Nijmegen
> The Netherlands
>
> Contact:
> E-Mail: jm.horschig at donders.ru.nl
> Tel:    +31-(0)24-36-68493
> Web: http://www.ru.nl/donders
>
> Visiting address:
> Trigon, room 2.30
> Kapittelweg 29
> NL-6525 EN Nijmegen
> The Netherlands
>
>
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20130910/f5bf2f55/attachment.html>