Dear all,
I am using SPM5 to perform a VBM analysis. In order to maximise the
quality of the segmentation step, and to maintain consistency with data
pre-processing, I am considering manually reorienting all my images to be
perfectly aligned to the grey matter tissue probability map (as far as
human subjectivity allows) using the 'check reg' function.
Having looked at images in both cases - non-reoriented v oriented, the
segmentation result appears visually very similar, however, this cannot be
equated with absolute accuracy, and clearly, I would like to get the very
best out of my data-set.
Given the large volume of images I will be working with(>150), I would
like to know if this is a methodologically robust and sensible approach to
pre-processing the data i.e. would the improvement in segmentation be
significant enough to impact on data quality and hence the final outcome
of the statistical analysis?
Many thanks for your advice,
Kind regards,
Louise Emsell
Clinical Neuroimaging Lab
Dept Psychiatry
NUI Galway
[log in to unmask]
|