As Andreas said, I would do it after. The purpose of BET in this case is to
give you a brain mask for calculations and isn't important for registration.
Peace,
Matt.
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf
Of Torsten Ruest
Sent: Friday, September 10, 2010 5:28 PM
To: [log in to unmask]
Subject: Re: [FSL] DTI pipeline questions
Hi Matt,
thanks for your answer. I did this before, I mean run eddy_current first,
and then split up the 4D image and average the corresponding acquisitions
and then put the averaged volumes back together to a 4D image. Just wanted
to make sure that this is common practice.
Regarding my 2nd question: I am doing the BET after eddy_current correction
at the moment - I am just not sure whether it is wise thinking eddy_current
correction may give better results using the brain only, but then the BET
results may not be the same for each volume (using 1st volume, ie b0, as
reference) in case the subject has moved.
Would you use BET before or after eddy_current correction?
Thanks,
Torsten
|