[FieldTrip] Computing the system matrix in a the forward model?

[patrick cox]

Hi,

I have just started using feildtrip, and I am trying to do some source
localization using beamforming. I am working with EEG data and would like
to anatomically constrain the forward model, so I am trying out the various
BEM methods in fieldtrip. As a first pass I have been using the SPM T1
template, but I also have anatomical scans of my subjects.

My problem is that when I try to calculate the model for the volume
conductor using most of the methods (openmeeg, dipoli), my vol structure
does not contain a vol.mat field (which I believe is the system matrix).
The one BEM method that actully returns a vol with a system matrix is the
BEMCP method, but the vol.mat file is all NaNs! I have tracked the problem
to the calculation of the Cij's, particularly C12 and C21, in the
ft_headmodel_bemcp.m which contain a large number of NaN entries.

Does anyone have a fix for this, or a advice on how to producing vol
structures with system matrices/vol.mat fields using the other BEM methods?

Thanks!
Patrick Cox
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20120206/bfe6c065/attachment.html>