Dear list,
I'm currently seeing funky changes of laterality for some contrast of
interest when comparing cluster permutation solutions between sensor space
and source space. Does this ring a bell fo anybody? Input would be
appreciated.
Cheers,
Denis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20140824/d35308dc/attachment.html
Hi Denis,
Is this MEG or EEG data? Perhaps sone examples plots of what you mean by changes in laterality would be useful...
Sometimes our intuition (or at least mine) about what field patterns correspond to what source patterns can be quite off. This is particularly true when looking at fields that are not classic dipole like...
That said, sometimes p-values can be unpredictably high in source regions where the projected signal is small (and hence low across permutation variance)..e.g., deep sources like the insula or the cingulate often tend to show up...
Hari
Hi Hari,
I'm personally not too surpised as well. I often see cluster solutions with
bilateral activaiton where only one of the clusters survives the
permutation test. The connectivity constrain that's currently implemented
in MNE-Python would disconnect such bilateral activations easily.
Especially for weak effects that lead to 1-cluster solutions. That being
said it's actually surprising that I do see any results in sensor-space
since no position-correction was available for the system + data-set I'm
currently working with.
But I'm trying to collect some arguments that you can use to explain this
phenomenon to fMRI experts that haven't used MEG. And of course trying to
rule out stupid mistakes on my end.
In this case we're talking about visual activation at the border between
BA18 and BA37.
- Denis
2014-08-24 20:51 GMT+02:00 Hari Bharadwaj <hari at nmr.mgh.harvard.edu>:
Hi Denis,
Is this MEG or EEG data? Perhaps sone examples plots of what you mean
by changes in laterality would be useful...Sometimes our intuition (or at least mine) about what field patterns
correspond to what source patterns can be quite off. This is particularly
true when looking at fields that are not classic dipole like...That said, sometimes p-values can be unpredictably high in source regions
where the projected signal is small (and hence low across permutation
variance)..e.g., deep sources like the insula or the cingulate often tend
to show up...Hari
>
> Dear list,
>
> I'm currently seeing funky changes of laterality for some contrast of
interest when comparing cluster permutation solutions between sensor space
and source space. Does this ring a bell fo anybody? Input would be
appreciated.
>
> Cheers,
> Denis
> _______________________________________________
> Mne_analysis mailing list
> Mne_analysis at nmr.mgh.harvard.edu
> Mne_analysis Info Page_______________________________________________
Mne_analysis mailing list
Mne_analysis at nmr.mgh.harvard.edu
Mne_analysis Info PageThe information in this e-mail is intended only for the person to whom it
is
addressed. If you believe this e-mail was sent to you in error and the
contains patient information, please contact the Partners Compliance
HelpLine at
MyComplianceReport.com: Compliance and Ethics Reporting . If the e-mail was sent to you in
error
but does not contain patient information, please contact the sender and
properly
dispose of the e-mail.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20140824/dc9e1a94/attachment.html
Hmm.. BA18/19/37 on the left and right hemispheres are not that far from each other... Funny things could happen if the activation is bilateral, especially if there are phase/latency differences between hemispheres...
Perhaps you could look for arguments along the lines presented here:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2797557/#!po=2.17391
Hari
You could use Euclidean distance to identify neighbors on the medial wall,
too. I have some code somewhere experimenting along those lines if you want
it.
Eric