Print

Print
Normalising via the MRI should be better, although this may depend on your images.  I'd suggest estimating the spatial transformations via segmentation.  One of the reasons for this is that the PET template was constructed from blood flow images, which may be slightly different from FDG.  There's also a bit more useful information in MRI that should allow more accurate alignment to be achieved.

Generally, the higher the degree of interpolation, the better the results should be.  Usually, the reasons for using lower degree interpolation just involve computational time.

Best regards,
-John

On 6 December 2012 14:33, Pieter Vandemaele <[log in to unmask]> wrote:
Dear SPM'ers,

We are starters in using PET and MRI images. In our study we would like to coregister FDG PET and MRI and normalise them to MNI.
I've searched the archives and the web, but I do not find recent comments on wich method is the best:
normalizing MRI to MRI template and PET to PET template or normalizing MRI to MRI template and normalize coregistered pet with MRI normalization transformation matrix.
Does anyone of you have experience with this? Are there other approaches? What are good parameters in terms of interpolation?
All help will be greatly appreciated

Pieter
--


Pieter Vandemaele, MSc-Ing
GIfMI Site Manager
Ghent Institute for Functional and Metabolic Imaging
MR Department -1K12
Ghent University Hospital
De Pintelaan 185
9000 Ghent - BELGIUM
[log in to unmask] tel: +32 (0)9 332 48 20
http://gifmi.ugent.be fax: +32 (0)9 332 49 69