When you coregister a pair of images, and want to carry an additional image
along, then you can specify the additional image as other. In the example
below, there were two ways of doing the coregistration:
1) Target: Image 1
Object: Image 2
Other: Image 3
Changes the .mat file of image 2 so that it matches image 1, and applies
the same change to image 3.
2) Target: Image 2
Object: Image 1
Other: none
Updates the mat file of image 1 so that it matches image 2. Image
2 is already in register with image 1.
Because the voxels of image 2 are aligned with those of image 3, then all
you need to do is ensure that image 3 has the same .mat file as image 2.
In this case, you can do this by simply copying the .mat file of image 2
to that of image 3. Use the <Check Reg> button to make sure everything
is OK.
Best regards,
-John
| I have a question regarding co-register.
|
| First some background-
|
| I have three T1-weighted volume datasets,
|
| 1) 3d volume scan, 256x256x124 voxel dimensions, 0.781x0.781x1.6mm size.
| Full brain coverage, coronally acquired.
|
| 2) scan 1, manually reoriented to form isotropic voxels of side
| 0.781, in a sagittal plane orthogonal to the bicommisural plane. I do
| not have the transformation matrix from scan 1 --> scan2
|
| 3) Scan 2, parcellated to form a dorsal and an orbital 'subfield' of the
| prefrontal cortex. All voxels not deemed to be prefrontal cortex are
| given a value of zero.
|
| Now, I am interested in ascertaining prefrontal GM and WM values for
| these prefrontal subfields, so have segmented scan 1 using SPM99b, and
| wish to get GM/WM values for my prefrontal subfields.
|
| I have successfully coregistered scan 2 to scan 1 using the coregister
| function in spm (after rotating scan 2 through 90 degrees to put them
| in an approximate frame of reference, ie roughly coronal).
|
| I am sure that the parameters that describe the transformation from scan
| 2 into scan 1 are hidden within an spm .mat file somewhere. If so, as I
| understand it it applying that transformation should put the parcellated
| images of scan 3 into the space of scan 1.
|
| Is there a way I can do this? Am I missing something straightforward?
| I tried to use the .mat file created in the coregistration of scans 1->2
| within the normalize function of SPM, but it fails- I guess that there
| is a difference between the formats of the _sn3d.mat and other .mat
| files.
|
|
| Can anyone give me some assistance?
|
| Cheers
|
| Matt
|
|
|
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Matt Howard
| Magnetic Resonance and Image Analysis Research Centre
| University of Liverpool
| Liverpool
| L69 3BX
|
| Tel 0151 794 5634
| Fax 0151 794 5635
|
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|