[Eeglablist] Trouble with dipole locations using standard MNI channel locations: No co-registration produces dipoles that appear to be rotated 90 degrees.

[James Schaeffer]

Dear eeglablist,

 I am source localizing with dipfit2 in eeglab 11.0.3.1b with a 128 biosemi
system and have renamed my channel labels to their closest match in the
10-20 system. After re-naming, I loaded the standard MNI channel locations
through 'look up locs' under 'edit:channel locations' in the GUI so I could
use them in dipfit. I thought that if I did this, there would be no need
for me to co-register my channels with the dipfit head model (as the
tutorial suggests). However, if I do not co-register, (by selecting 'no
coreg' in the head model settings) my dipole locations appear to be rotated
90 degrees clockwise (when viewed from above), relative to their scalp
maps. If I do co-register the standard MNI channel locations, and warp the
montage, the dipole locations appear where they are expected (I think). I
tested this by fitting dipoles to what I believe to be an eye artifact in a
few datasets. Without co-registration, the dipoles appeared in the right
inferior temporal lobe; with co-registration (and warping), the dipoles
appeared in the left eye (as I expected).

 Has anyone else encountered this? Am I missing an important step when
using standard MNI locations without co-registration? Is co-registration
and warping required even when using the standard MNI locations? Any help
would be greatly appreciated.

 Thanks in advance,

James
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20120910/bbcd69b8/attachment.html>