Back to the main page.

Bug 1314 - consider incorporating MNI source grids into ft_prepare_sourcemodel

Reported 2012-02-07 17:43:00 +0100
Modified 2012-07-07 20:25:07 +0200
Product: FieldTrip
Component: forward
Version: unspecified
Hardware: PC
Operating System: Mac OS
Importance: P3 normal
Assigned to: Jan-Mathijs Schoffelen
Depends on: 1122
See also:

Robert Oostenveld - 2012-02-07 17:43:35 +0100


Johanna - 2012-02-08 10:02:03 +0100

Not sure exactly the plan or what's needed/required/assumed, but maybe nut_makeMNIvoi can help with ideas? It takes a requested voxel spacing, assuming the subject's MRI has been normalised to MNI, and creates the grid in MNI coords, inside the bounds defined by the MNI brain and *_sn.mat mapping.

Robert Oostenveld - 2012-02-08 10:31:17 +0100

(In reply to comment #1) something along these lines was indeed what I was thinking of, and something like The thought was inspired by a discussion with Vladimir, in which we looked at the SPM source reconstruction pipeline. SPM makes a lot of use of templates and warps them back and forth. We actually do the same, but not in a user-friendly way. My idea is that instead of a long recipe as in the example script on the wiki, we would enhance the ft_prepare_sourcemodel function to allow for a short recipe that involves - ft_normalisevolume - ft_prepare_sourcemodel

Johanna - 2012-02-08 10:49:26 +0100

Created attachment 226 Nutmeg file to create MNI voi I've attached the file in case it helps ideas. It needs some sub-functions, let me know if I should post those too.

Jan-Mathijs Schoffelen - 2012-02-10 10:37:39 +0100

first implementation in revision 5270 use as cfg = []; cfg.mri = ... cfg.grid.warpmni = 'yes'; cfg.grid.resolution = 10; grid=ft_sourcemodel(cfg); TO DO: update documentation, and check the template grids

Jan-Mathijs Schoffelen - 2012-04-11 10:31:12 +0200

implemented nonlinear warping in revision 5604. To do: update documentation

Jan-Mathijs Schoffelen - 2012-04-11 11:15:46 +0200

Documentation updated: