Dear Martin, I would recommend using vitamin E capsules and the 'type' option. You should not normalize your individual structurals, just coregister them to the template, write down the coordinates of the capsules in those coregistered images and then use the same images with the 'type' option to build and coregister the head model. Regarding headshape I would say that precisely measured 3 points are sufficient for coregistration whereas headshape usually has some 'wiggle room' and can only spoil things if you have such a precise measurement of the fiducials. Best, Vladimir On Fri, Nov 25, 2011 at 1:02 PM, Martin Dietz <[log in to unmask]> wrote: > Dear Vladimir, Rik & MEG users, > > Given that we aquire subject-specific structural MRIs for the MEG head models, what would be the preferred way of specifying our digitized fiducial points for nasion, lpa and rpa in structural MRI space ? Would you recommend the 'select' option and use SPM's hardcoded locations or use more precise information by including e.g. vitamin E capsules in the structural image and use the 'type' option to select the MNI coordinates in the normalized MRIs. > > Also, would you recommend using head shape point ? Our system is the new Elekta Neuromag Triux. > > Best wishes, > Martin > > Martin Dietz > [log in to unmask] >