If you are using SPM5, then simply creating .mat files in MATLAB may not help.
SPM5 modifies the image headers when writing such information - unless you
have a 4D image time series, which requires an additional .mat file to accout
for time series realignment (no space in the .hdr for multiple matrices).
Try using spm_get_space to read and write such matrices.
If your scans were acquired within the same scanning session, then the
transmission and emission scans should be in fairly good alignment anyway.
By knowing which voxel of the images is at the centre - as well as voxel
sizes and image orientations (L-R), it should be possible to create matrices
that do a pretty good job of aligning the volumes. I'm not sure how accurate
the transmission/emission coregistration is likely to be if done via mutual
information coregistration.
Best regards,
-John
On Wednesday 19 September 2007 00:30, Wardak, Mirwais wrote:
> Dear all,
>
>
>
> I have a question regarding transformation matrices and their
> application to other images of different zooms. I did a co-registration
> between an emission (EM) and transmission (TX) scan (both of zoom = 3.5)
> and would like to apply this transformation to a transmission image of
> zoom = 1.0. I have the following images with their respective
> dimensions:
>
>
>
> 1) EM1_zoom_3.5.img (2.06 x 2.06 x 2.43 mm)
>
> 2) TX1_zoom_3.5.img (2.06 x 2.06 x 2.43 mm)
>
> 3) EM2_zoom_1.0.img (5.14 x 5.14 x 2.43 mm)
>
> 4) TX2_zoom_1.0.img (5.14 x 5.14 x 2.43 mm)
>
>
>
> TX1_zoom_3.5.img (source) ---(Coreg&Reslice)---> EM1_zoom_3.5.img
> (target) (a .mat file was generated, called 'TX1_zoom_3.5.mat')
>
>
>
> I tried to now "Reslice only" by using the .mat file generated above and
> placing it the same folder as TX2_zoom_1.0.img & EM2_zoom_1.0.img. I
> changed the prefix of the .mat file name so that it is the same as the
> name of the transmission image (i.e., TX1_zoom_3.5.mat =
> TX2_zoom_1.0.mat). Therefore, now I have:
>
>
>
> TX2_zoom_1.0.img (source) ---(Reslice Only/TX2_zoom_1.0.mat)----->
> EM2_zoom_1.0.img (target image space)
>
>
>
> However, SPM did not do the registration correctly as it did not realize
> the different voxel sizes between source and target images from that
> described in the transformation matrix, which was generated in dealing
> with images of zoom 3.5. Please note that the TX image has to be in
> zoom=1.0 for the reconstruction program to work when given a sinogram
> and that direct registration of TX2_zoom_1.0.img to EM2_zoom_1.0.img
> does not work using Co-Register because of the small zoom, hence why we
> had to do the registration in a larger zoom. Normalized mutual
> information is applied in the registration process.
>
>
>
> How do we modify "TX2_zoom_1.0.mat" so that this registration will work
> in zoom 1.0 for our case? Any help would be greatly appreciated. Thank
> you for your time.
>
>
>
> Cheers,
>
> Mir
>
>
> ----------------------------------------------------------
> IMPORTANT WARNING: This email (and any attachments) is only intended for
> the use of the person or entity to which it is addressed, and may contain
> information that is privileged and confidential. You, the recipient, are
> obligated to maintain it in a safe, secure and confidential manner.
> Unauthorized redisclosure or failure to maintain confidentiality may
> subject you to federal and state penalties. If you are not the recipient,
> please immediately notify us by return email, and delete this message from
> your computer. ----------------------------------------------------------
|