Hi there,
I have gotten good results from eddy with 12 diffusion dirs and even with
just 3.
It may be worth to try. Just check the outputs.
Cheers,
Andreas
Am 02.11.15 13:59 schrieb "FSL - FMRIB's Software Library on behalf of
Jesper Andersson" unter <[log in to unmask] on behalf of
[log in to unmask]>:
>Dear Marek,
>
>> I would like to ask a general question about FSL, one specific about
>>eddy_correction and the one about new eddy script.
>> The general question:
>> I have a data set of 1 b0 volume and 15 diffusion weighted volumes
>>(DWI) of human spinal cord. Which scripts are suitable to use on this
>>data set (I mean which are applicable to general data and which are
>>applicable specifically only to brain? E.g. BET works obviously on brain
>>only...)?
>
>most of the diffusion tools can be used also outside the brain.
>eddy_correct, eddy, dtifit, bedpost etc should all work fine. Many of
>them take masks that you will have to make manually (because bet wonąt
>work).
>
>> Eddy_correct:
>> I used eddy_correct on my data set and generally it works right but
>>after application on same data random DWI volumes are all-zero (without
>>any error massage) which produce wrong FA image. I use same magnet, same
>>protocol, same everything, but sometimes eddy_correct works wrong and I
>>do not know why. Can you help me?
>> Eddy:
>> You gave an example on FSL web site on data set with 5 b0 and 59 DWI
>>images but it is possible to use eddy to my data set (1 b0 and 15 DWI
>>images)? Could eddy gives better results than eddy_correct script? I
>>mean if eddy can fix the problem with zero volumes which eddy_correct
>>makes.
>
>I suspect that 15 dwi might be pushing it a bit for eddy, but there is no
>harm in trying.
>
>The problem with all zero sounds strange though. Have you looked at the
>resulting .mat files to see if you have some very large translations that
>have translated the object outside the FOV?
>
>Jesper
>
>
>> Thanks
>> Marek
|