 |
 |
Thanks for the replies. Yes, I am referring to b=0 volumes, so there should be no eddy currents, which should make 6 dof the best option. Of course for eddy current correction I am using 12 dof ;) Dustin ________________________________________ From: FSL - FMRIB's Software Library [[log in to unmask]] On Behalf Of Andreas Bartsch [[log in to unmask]] Sent: Friday, April 20, 2012 5:39 PM To: [log in to unmask] Subject: Re: [FSL] b=0 averaging Hi Stam, oh yeah - sure. I was still hanging a bit in the previous query...;) Cheers, Andreas Am 20.04.12 23:35 schrieb "Stamatios Sotiropoulos" unter <[log in to unmask]>: >Hi Andreas, > >I thought that Dustin was referring to b=0's, so no eddy currents ;) >Obviously registration won't be perfect due to EPI distortions, but it is >safer to try 6 dofs rather than 12, if such a global within-subject >transformation is attempted. > >Cheers >Stam > > > >----- Original Message ----- >From: "Andreas Bartsch" <[log in to unmask]> >To: <[log in to unmask]> >Sent: Friday, April 20, 2012 9:51 PM >Subject: Re: [FSL] b=0 averaging > > >>6 dof maybe be good but it assumes that there are no eddy currents >> involved and that the B0 volumes can be registered by a rigid body >> transformation. >> cheers, >> andreas >> >> Am 20.04.12 17:20 schrieb "Stamatios Sotiropoulos" unter >> <[log in to unmask]>: >> >>>Hi Dustin, >>> >>>correct, apart from the SNR gain. It is sqrt(N), so in your case >>>sqrt(6). >>> >>>Cheers >>>Stam >>> >>> >>>----- Original Message ----- >>>From: "Cunningham, Dustin" <[log in to unmask]> >>>To: <[log in to unmask]> >>>Sent: Friday, April 20, 2012 4:07 PM >>>Subject: [FSL] b=0 averaging >>> >>> >>>If I have say 6 b=0 images and I want to average them I would simply >>>coregister them (6 dof) then add and divide by 6, correct? >>> >>>...and this will give me a 6*sqrt(2) increase in SNR? >>> >>>Just want to make sure Im not missing anything. >>> >>>Dustin >>> >>> >>>________________________________________ >>>From: FSL - FMRIB's Software Library [[log in to unmask]] On Behalf Of >>>Andreas Bartsch [[log in to unmask]] >>>Sent: Thursday, April 19, 2012 5:47 PM >>>To: [log in to unmask] >>>Subject: Re: [FSL] DTI: Order of Eddy_correction vs. Unwarping >>> >>>Hi Pete, >>> >>>this is good! >>>Cheers, >>>Andreas >>> >>>PS: You don't really need 2.c, do you? >>> >>>Am 19.04.12 15:38 schrieb "Peter Fried" unter <[log in to unmask]>: >>> >>>>Hi FSL Community, >>>> >>>>Can someone confirm the appropriate order of pre-processing steps for >>>>DTI >>>>data? >>>> >>>>Specifically, does it matter whether the unwrapping (using the >>>>pre-stats >>>>option in FEAT) occurs before or after the eddy_correction? >>>> >>>>Here is what I have been doing and I just want to make sure it's >>>>correct: >>>> >>>>#1. Eddy Correction >>>>#2. Feat Pre-stats >>>> a. Unwarp >>>> b. bet >>>> c. Registration >>>>#3. dtifit >>>>#4. bedpostx >>>> >>>>Thanks! >>>> >>>>Cheers, >>>>Pete Fried >>>> >>>>Department of Anatomy & Neurobiology >>>>Boston University School of Medicine >> >>