[Mne_analysis] mne_make_cor_set - transform problem

Krieger, Donald N. kriegerd at upmc.edu
Mon Apr 18 11:53:22 EDT 2016
Search archives:

Hi Matti,

Thanks for responding so quickly.
I tried what you suggest for a scan for which I have dicoms.
I created 2 .fif’s, one with dicom_access and one with mne_make_cor_set.

Here are the two transforms grabbed with show_fiff:
# dicom_access
# 222 = transform         head -> MRI
#         0.999356 0.028735 -0.021495
#         -0.035189 0.902169 -0.429945
#         0.007038 0.430425 0.902599
#         -4.798210 8.148447 -50.913933 (inv. 5.440169 14.701207 49.355110)
# mne_make_cor_set
# 222 = transform         head -> MRI
#         0.999505 0.024229 0.020096
#         -0.014078 0.915056 -0.403080
#         -0.028155 0.402598 0.914944
#         -3.010607 -9.352709 -35.697933 (inv. 1.872374 23.003098 28.952219)

I guess the rotations look pretty much the same and if I apply the tkr2scanner transform to the displacement, -3.010607 -9.352709 -35.697933, I get
-5.14 6.311 -52.083 which is also pretty good.

That’s encouraging actually.  Do you see any way to get it to kick out the correct displacements?


From: mne_analysis-bounces at nmr.mgh.harvard.edu [mailto:mne_analysis-bounces at nmr.mgh.harvard.edu] On Behalf Of Matti Hamalainen
Sent: Monday, April 18, 2016 11:36 AM
To: Discussion and support forum for the users of MNE Software
Subject: Re: [Mne_analysis] mne_make_cor_set - transform problem

Hi Don,

You can read the fif file mne_make_cor_set made in mrilab and do the coordinate transformation there as you have used to and produce a new fif file where the transformation is in place.
This does not deal with the different coordinate systems between the one derived from the DICOM files and the mgz files, though.

- Matti

On Apr 18, 2016, at 11:31 AM, Krieger, Donald N. <kriegerd at upmc.edu<mailto:kriegerd at upmc.edu>> wrote:

This is a follow-on to a previous post to which Jon Houck helpfully suggested using mne_make_cor_set.
I am using this to convert .mgz files to .fif .
I normally start with the original dicom set and convert them to .fif using ELEKTA’s dicom_access.
But for these MEG data sets, I only have a defaced (via freesurfer) .mgz files.

In either case I then use the .fif files to do MEG coregistration with ELEKTA’s Mrilab tool.
Using Mrilab is part of my standard pipeline and I would like to stick with that if possible.

When I convert a standard dicom scan series to fif using ELEKTA’s dicom_access, the .fif is in MR RAS coordinates.
But a .fif file obtained by conversions from .mgz with mne_make_cor_set appears to be in tkr RAS coordinates instead.
So the transform matrix that I get when I do the coregistration is: Head --> tkrRAS  rather than Head --> RAS  .

I tried spoofing mne_make_cor_set with –talairach SomeOtherTransformation  but Mrilab doesn’t appear to care.
In addition, it’s not clear to me what transform to use even if it does work; maybe tkr2scanner .

Any thoughts would be welcome and thanks.

Best - Don

Mne_analysis mailing list
Mne_analysis at nmr.mgh.harvard.edu<mailto:Mne_analysis at nmr.mgh.harvard.edu>


Matti Hamalainen, Ph.D.
Athinoula A. Martinos Center for Biomedical Imaging
Massachusetts General Hospital

msh at nmr.mgh.harvard.edu<mailto:msh at nmr.mgh.harvard.edu>
mhamalainen at partners.org<mailto:mhamalainen at partners.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20160418/81743ab5/attachment-0001.html 

More information about the Mne_analysis mailing list