|
Jeff_Browndyke
Junior Member
35 Posts |
 Posted - Aug 25 2006 : 12:29:20 AM
|
SPM appears to do a pitiful job coregistering anatomical to functionals collected via the 4T inverse spiral sequence. My problems are further magnified by ventricle size and generalized atrophy. Any thoughts on how to the tackle coregistration problems with the inverse spiral data?
Anatomicals collected axial-oblique w/ .94 x .94 x 1.9
Functionals collected axial-oblique w/ 3.75 x 3.75 x 3.8
The plan is to eventually establish my own study-specific segmentation priors and apply the grey matter seg image to study-specific grey template normalization parameters to bring the functionals into the same space. This seems to be the most sensible method given the atrophy in the subject data and poor similarity with the demographics of the standard SPM templates.
I've run through a few subjects and with standard coregistration flags (medium or light regularization), and I'm getting poor registration in the inferior occipital region, which then results in functional voxels in this region pushed down into cerebellum on the "coregistered" T1.
Also, we're planning on running the same data w/ ROI analyses in native space, so good-to-better coregistration is going to be key.
Any help or pointers would be most appreciated.
Jeff |
|
4T inverse spiral to T1 coreg problems
