External Email - Use Caution
This email was sent 22 hours ago, but somehow it haven't appeared on the forum yet. I post it again.
Gabor Perlaki petzinger.gabor@gmail.com ezt írta (időpont: 2021. jún. 2., Sze, 12:11):
The default talairach registration is wrong in ~20% of our subjects and thus eTIV is also quite inaccurate (e.g. <1000 cc in certain cases). I tried -use-mritotal which resulted in much more accurate talairach registration. My questions: Q1, I think it's not a good idea to mix talairach_avi and mritotal registrations in a single study, because they can consequently result in different eTIV values or initialization parameters affecting the final segmentations. Therefore it would be easier to run the whole analysis again with mritotal as default rather than correcting the talairach registration and rerunning the analysis in separate steps for each subject. Is the following command OK for this purpose? recon-all -s SUBJ -i input_dicom_file -use-mritotal -all
Q2, Is mritotal only used to calculate talairach.xfm or are there any other hidden analysis steps which may perform worse (even though talairach.xfm is more accurate) when using it instead of talairach_avi?
Q3, Is it more accurate to use sbTIV instead of eTIV?
Thanks, Gabor