[FieldTrip] about leadfield correction following ICA component rejection

[Schoffelen, J.M. (Jan Mathijs)]

Hi Sanghyun,

Although the ‘perceived SNR’ may change when applying the leadfield correction, it is simply biophyisically incorrect, if you DON’T do it.

The reason is the following:
If you remove a component from your channel data, what happens is that each of the resulting EEG/MEG channels reflect the potential difference/magnetic field strength as measured by a referenced electrode or a magnetic field pickup coil at a specific location in space, minus a weighted combination of potential differences/magnetic field strengths measured at all other locations (where the weighting is given by the spatial topography of the component). In other words, you apply an intricate re-referencing scheme/create a complicated synthetic gradiometer sensor. When building your forward model (a.k.a. leadfield), the purpose is to build an accurate model of the spatial distribution of potential differences/magnetic field strengths given what your channel leveld data represents. Since it now represents this re-rereferenced data/synthetic gradients, it needs to be accounted for in the model.

Best,
Jan-Mathijs



On Feb 15, 2016, at 3:57 AM, Sanghyun Lim <santagravity at gmail.com<mailto:santagravity at gmail.com>> wrote:


Dear fieldtripers,


I recently read a hipp's 2015 paper which asserts that do not modify leadfields after ICA artifact rejection since it could decrease the estimated source SNR.
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0121048

But by seeing ft_rejectcomponent, fieldtrip recommends modifying leadfields after ICA.
Is there a specific reason to do this?


thanks in advance
Sanghyun.
[https://mailfoogae.appspot.com/t?sender=ac2FudGFncmF2aXR5QGdtYWlsLmNvbQ%3D%3D&type=zerocontent&guid=8d602c05-59e3-43c4-876d-df591e738182]ᐧ
_______________________________________________
fieldtrip mailing list
fieldtrip at donders.ru.nl<mailto: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/20160217/ad84214b/attachment.html>