Dear SPMers,
I'm using VBM8 and I think the manual is quite helpful!
However I still have a couple of questions on which I would like to
get some help:
- As far as I understand, the segmentation procedure does not include
the bias correction. It is rather a separate option of "Estimate and
Write" module. Wouldn't it be recommended to do the segmentation into
GM, WM and CSF of previously bias corrected images? I mean, perform
and save bias corrected images and subsequentely give these images as
input for segmentation. In this case, in which space should we write
the bias corrected images? Normalized or native space?
- For VBM the partial volume effect (PVE) is only needed to perform a
better segmentation of tissues. It does not output any image that will
be needed later for the next steps of VBM, right?
- The difference in writing normalized or modulated normalized tissue
images. When writing only 'normalized' images we can study
concentration of gray matter. When writing 'modulated normalized'
images images we can study volume of gray matter. What is the
difference, i.e. how do we interpret the results from the two options?
What does it mean 'concentration' of gray matter?
- When performing normalization and segmentation the non-brain (the
skull) tissue is automatically removed if 'clean up any partitions' is
chosen. It is not necessary thus to apply any brain mask, correct?
- In any step, is there any change to the intensity values of the
images, in relation to the native space? I mean, for example in the GM
images, what does it mean a higher intensity value? Does it mean
higher amount of GM tissue? And how does this change when writing the
GM image 'normalized' or 'modulated normalized'?
Sorry for the long e-mail.
Thanks in advance!
Joćo Duarte
|