Hi SPM'ers
I'm grappling with the VBM method using SPM.
From my understanding, I should be able to accomplish the three basic
steps of spatial normalization, segmentation and smoothing by following
these steps using the SPM GUI:
1. Segmenting each T1 image, answering "No" to the prompt, "Are the
images spatially normalised?" and then specifying the modality (T1). I
assume this spatially normalises the images to the SPM T1 template prior
to segmentation.
2. Smoothing the segmentation outputs with a 12 mm kernel.
Is this simple approach even valid ?
When I run the optimized VBM script (from
http://dbm.neuro.uni-jena.de/vbm.html ) on the same data using the
default templates, the t-test results are not the least bit similar. (I
also smoothed the segmented outputs with a 12 mm kernel). I gather that
the script normalizes to SPMs GM template in this case. Could this be a
reason for the huge differences between the two?
Thanks for your help
VBM Newbie
|