Dear Felix,

I talked to John today and he said that the problem is due to some changes he introduced into how inverse normalisation is done. He will think of a fix. Yes, you do need to find fiducials for every new structural but you only have to do it once. If you have a headshape or EEG sensors you could also use some standard fiducials to get rough alignment and then fit the headshape.

Best,

Vladimir 


On Mon, Jun 2, 2014 at 8:27 PM, Felix Nettersheim <[log in to unmask]> wrote:
Dear Vladimir,

thank you very much for your quick reply.

When using "type" rather than "select" to specify the fiducial's position there is indeed no error occuring.

I also discovered, that the error only appears, if specifying the positions of the two preauricular fiducials by pressing "select" rather than "type".
In other words, there is no error occuring, when using "select" for the position of the nasal fiducial  and then using "type" for the positions of the preauricular fiducials.
Furthermore I figured out, that there is also no problem, when using the template rather than an individual MRI.

Since this really confused me, I had a look at one of our MRI-images  (pressing "display" after loading it):
While SPM8 is showing all the fiducials that should be hard-coded in SPM (nas, lpa,rpa, FIL_CTF_L, FIL_CTF_R) in SPM12b there are always some fiducials missing (it depends on the MRI-image which ones are missing (mostly there is only "nas" displayed and sometimes also "lpa" or "rpa").

Do you have an idea why SPM12 does not display the fiducials correctly, whereas SPM8 does or what we can do to solve our problem?
Is it right, that if we "type" in the positions of the fiducials we have to get their MINI-coordinates from every single MRI, or did I get it wrong?

Thanks in advance.

Kind regards,

Felix Nettersheim