I hiave some questions that I i've been thinking about on a technical level for some time, but have never considered in the abstract. I'm using the fsl_anat sequence these days which is significantly improving the quality of my brain extractions.....but
Even with the sequence the final extracted brain is missing the very top of the brain in some areas and including small bits of meninges in others. I'm working in s1 s2 so I'm particularly interested in the gray matter at the vertex of the brain.
The brain mask however is expanded slightly covering the brain AND some meninge.
1. What sort of consequences will this have on subsequent registrations such as for my .functional data? Should I take the effort to hand correct these issues, or does fnirt blur the image enough to render these small differences unimportant?
2. Although well within the norms of human cariable, some of my participants have heads that are quite different than the standard. The head to head registration pulls their heads into unexpected shapes (the bunp posterior to the parital/occipital suture boundary Is muchi morepronounced in men then women). Some of my female oarticipants are getting n effect where the middle of their head is dragged upward. Should I refrain from the fnirt head registration step when this happens?
|