There's no actual code within SPM that allows you to transform the Template_0-6.nii to MNI space. If you really want to do this, then you can do it with a little bit of MATLAB: matfile = 'Template_6_2mni.mat'; imfile = 'Template_6.nii'; t = load(matfile); Nii = nifti(imfile); mat = t.mni.affine; Nii.mat = t.mni.affine; Nii.mat_intent = t.mni.code; create(Nii) Note that for Template_6_2mni.mat to exist, you have to have run a Normalise to MNI space first. Yes, the rigid body transform is based on aligning with spm12\tpm\TPM.nii , although it is actually extracted from the nonlinear deformation that is estimated during segmentation. The rigid bod transforms are encoded in the headers of the rc*.nii and u_*.nii files. Best regards, -John On 20 April 2017 at 11:53, FENG LIU <[log in to unmask]> wrote: > Dear SPM experts, > > > > I performed the DARTEL analysis using SPM12. I have three questions in my > experiment. > > > > First, I want to create a study specific template using DARTEL in MNI > space. However, when I finish the steps “Run Dartel (create template)” and > “Normalize to MNI space”, I found that the only the gray matter and white > matter images normalized to the MNI space, the Template_0-6.nii are still > in the dartel (group averaged) space. How could I transform these Templates > to the MNI space? > > > > Second, in the step of “Run Dartel (create template)”, I inputted the > rc1*.nii and rc2*.nii which generated by the previous “Segment” step. The > rc1 and rc2 are the roughly aligned (via a rigid-body) gray matter and > white matter images. Which image was used as the template for the > rigid-body transformation? Is this image “spm12\tpm\TPM.nii”? > > > > Finally, after the “Run Dartel (create template)” step, a series of > u_rc1*.nii were generated which parameterized the deformations. In the > following step of “Normalize to MNI space”, I inputted the templated_6.nii, > u_rc1*.nii and c1&2*.nii, and finally obtained a bunch of smwrc1&2*.nii > files. Here, my question is that the u_rc1*.nii was obtained based on > rc1&2, so the u_rc1*.nii I think is the deformation from the rigid-body > images (rc1&2) to the dartel space. When I run “Normalize to MNI space”, > why I need to input the c1&2 rather than rc1&2? Have the rigid body > transformation parameters already imbedded into the u_rc1*.nii? When? > > > > Any help would be greatly appreciated! > > > > Best, > > Feng >