Dear Iris,
I would suggest leaving the maps in their native resolution, and using a more recent version of SPM (preferably SPM5) to realign the maps to T1 images, then normalize them just as you describe.
-Matt Senjem
-----Original Message-----
From: SPM (Statistical Parametric Mapping)
To: [log in to unmask]
Sent: 2/15/2006 8:23 PM
Subject: [SPM] normalizing metabolic maps
Hello everybody,
We are trying to do an ROI correlation analysis on perfusion and
metabolic maps. We already coregister (using high res, 3D, T1 images)
and spatially normalize perfusion images using SPM99.
The metabolic maps are the outcome of a spectroscopy software we have
developed. They correspond to spectra acquired on (packages of) slices
overlaid on the high resolution T1 image of the subject.
To get these maps into the same template as the perfusion images, we are
thinking of doing the following: 1) zero-fill, realign and
sinc-interpolate the metabolic maps to match the resolution of perfusion
maps. 2) Transform them into the standard brain space (SPM99) using the
matrix obtained from the spatial normalization of the high res, 3D, T1
images.
Does this make sense? Are we missing something important in our
rationale? (We are new to using SPM and far from a detailed
understanding of all the steps involved).
Any input/advice would be greatly appreciated,
Iris Asllani
Hatch Center for MR Research
Columbia University
Tel. 212 342 0183
<<Blank Bkgrd.gif>>
|