[FieldTrip] Predefined Source Orientations for Source Localization

[손준호]

Dear fieldtrip community,

I am currently trying to get source time series form my MEG data using
beamformer LCMV.

I have run ft_sourceanalysis with fixed orientation option, and I found
that the estimated source orientations do not perpendicular to source
surface.

I thought that dendrites of pyramidal cells point perpendicular directions
to the cortical surface, but the result was different from what I expected.

Then I have tried to provide predefined orientations so that I use
the provided orientations rather than estimate them.
However, I also failed to get orientations that I expected.

So, here is my question,
1) Why estimated source orientation does not perpendicular to source
surface?
2) Are there any methods or options that use predefined source orientations
for source localization?
3) If it is natural to get sources that are not perpendicular to
the cortical surface,  how can I explain those source activities in terms
of neuronal structure and neuronal activity?

Thank you,
Junho
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20181127/bd6c1574/attachment-0001.html>