Dear Marc,
You can find some answers at https://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=SPM;6b20e904.1004
Also you can look at the functions of the Beamforming toolbox in SPM
for an example of how to get the results into MNI coordinates. A
convenient option is to transform everything to MNI coordinates right
at the beginning of your analysis (as is done for instance in
spm_eeg_ft_beamformer_freq) with just a rigid body transform. Also if
you have an individual MRI, start with coregistering it to the
template. Then you have less chances of getting things wrong.
Best,
Vladimir
On 15 Jul 2010, at 17:52, Marc Recasens wrote:
> Hi everybody,
>
> I'm doing MEG source analysis with Fieldtrip, but I use spm8 Source
> Reconstruction in order to use the coregistration algorithm, so I
> can use a digitized headshape to match the individual brain
> segmentation.
> Within the Source Reconstruction tool I do all the steps, including
> the Forward solution and applying a Single-Shell head model, but not
> the inverse solution, which I do with Fieldtrip.
>
> Then I take the generated single-shell volume from:
>>> D.others.inv{1,1}.forward.vol
> and the sensors and headshape from:
>>> D.sensors.meg
>>> D.fiducials
> Everything seem nice and the matching between sensors, headshape and
> the generated volume is perfect.
>
> However, I need to find an MRI transformation matrix which
> coordinates correspond with the volume generated.
> As far as I know (and reach to understand), the headshape and
> sensors I use are in head-coordinates (i.e. in the coordinate system
> of the 4-D MEG machine); and the generated volumen fits into those
> coordinate system.
>
> I don't know if it it possible to obtain a transformation matrix
> that converts the typical MRI coordinate system into the coordinate
> system “D.others.inv{1,1}.forward.vol" works with.
> Could anyone tell me if that's possible and what would be the
> transformation matrix? I tried many solutions but no one seems to
> work.
>
>
> Bests,
>
> Marc
|