<div dir="ltr">Hi May, Inna, everyone,<div><br></div><div>I'm not sure exactly what Inna meant by the ROI, specifically whether the ROI was applied to source space or to sensors. Applying a ROI to sources will not work: the beamformer result at one location in the brain is the same irrespective of where else you also compute a beamformer result. However, perhaps Inna meant ROI over channels (e.g. left channels for the left auditory cortex). If so, then as she says, it will reduce the problem a bit but it will still be there. See cfg.supchan in ft_sourceanalysis.</div>
<div><br></div><div>Indeed the method by Dalal et al. was initially implemented in the <a href="http://www.nitrc.org/projects/nutmeg/" target="_blank">NUTMEG </a>software (which is compatible with FieldTrip). However, it is also now implemented in FieldTrip by using the cfg.supdip option in ft_sourceanalysis, which allows you to suppress a region (e.g. right auditory cortex) so that you can estimate activity elsewhere (e.g. left auditory cortex).</div>
<div><br></div><div>Cheers,</div><div>Johanna</div>
<div><br></div><div><br></div><div><br></div><div><br></div><div><br></div><div><br></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">2014-03-20 11:47 GMT+00:00 Heng-Ru May Tan <span dir="ltr"><<a href="mailto:Heng-RuMay.Tan@glasgow.ac.uk" target="_blank">Heng-RuMay.Tan@glasgow.ac.uk</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#333300" bgcolor="#FFFFFF">
Hi Inna,<br>
<br>
Thanks for the suggestions re using ROI beamforming separately and
references. I'll give it a go.<br>
<br>
BTW, I wonder if those beamforming algorithms e.g. Dalal et al2006
(IEEE biomed. eng.) might have already been implemented somewhere...
or might it be something that fieldtrip might/could consider adding
to their sourceanalysis options? =)<br>
<br>
Thanks,<br>
May<br>
<br>
<br>
<blockquote type="cite">
<fieldset><legend>ForwardedMessage.eml</legend></fieldset>
<table width="100%" border="0" cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td>
<div style="display:inline">Subject:
</div>
Re: [FieldTrip] Source localizing auditory steady state
responses</td>
</tr>
<tr>
<td>
<div style="display:inline">From:
</div>
"McGowin, Inna" <a href="mailto:mcgoiv0@wfu.edu" target="_blank"><mcgoiv0@wfu.edu></a></td>
</tr>
<tr>
<td>
<div style="display:inline">Date:
</div>
19/03/2014 17:21</td>
</tr>
</tbody>
</table>
<table width="100%" border="0" cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td>
<div style="display:inline">To:
</div>
FieldTrip discussion list <a href="mailto:fieldtrip@science.ru.nl" target="_blank"><fieldtrip@science.ru.nl></a></td>
</tr>
</tbody>
</table>
<br>
<div dir="ltr">
<div>
<div>
<div>
<div>
<div><br>
<div>Hey May,<br>
</div>
Beamorming on correlated sources is my interest too.
The problem is that most beamformers (I deal with SAM
and vector beamformer) are going to suppress or/and
misplace any correlated source. The way I deal with
this problem is by creating a region of interest
(around primary auditory cortex) and beaformer over
each ROI separately. The signal correlation problem
is still there but with a lesser extant (I think).
Here is a great article on this topic:<br>
Beamformer reconstruction of correlated sources using
a modified source model<br>
</div>
and <br>
Modified Beamformers for Coherent Source Region
Suppression<br>
<br>
</div>
Also if I localize evoked responses I always prefer dipole
fitting over other methods, you could try the moving
dipolar fitting method and have your wave former from the
dipole t-course.<br>
</div>
Let me know how it will go for you.<br>
</div>
Cheers,<br>
</div>
Inna<br>
<div>
<div>
<div>
<div>
<div><br>
</div>
</div>
</div>
</div>
</div>
</div>
<br clear="all">
<div>Inna McGowin<br>
</div>
<br>
<br>
On Wed, Mar 19, 2014 at 12:17 PM, Heng-Ru May Tan <span dir="ltr"><<a href="mailto:Heng-RuMay.Tan@glasgow.ac.uk" target="_blank">Heng-RuMay.Tan@glasgow.ac.uk</a>></span>
wrote:<br>
<br>
Hi All,<br>
I am trying to localize frequency specific ASSR but have not been
successful in doing so with the 'mne' method in the
ft_sourceanalysis.<br>
At the sensor-level the topo plots show clear bilateral
generators.<br>
<br>
Is it possible to localize the bilateral sources using
beamforming? And would I need to create a different forward model?<br>
Additionally, I would also like to retrieve the source signals and
ideally comparing between 2 time periods within the same trial.<br>
<br>
I would appreciate if someone could advise or comment on how best
to go about doing so.<br>
<br>
Thanks,<br>
May<span><font color="#888888"><br>
</font></span></blockquote>
<br>
</div>
<br>_______________________________________________<br>
fieldtrip mailing list<br>
<a href="mailto:fieldtrip@donders.ru.nl">fieldtrip@donders.ru.nl</a><br>
<a href="http://mailman.science.ru.nl/mailman/listinfo/fieldtrip" target="_blank">http://mailman.science.ru.nl/mailman/listinfo/fieldtrip</a><br></blockquote></div><br></div>