Dear Catarina,
> Dear Jesper,
> Thank you very much for the quick reply. Indeed when I set it to zero for all iterations I don’t observe shifts in the through-slice direction.
> I want to assume this because I’m taking the two volumes consecutively (blip up then blip down) in a very high-res acquisition, ~7 slices and small FoV. Do you think the motion correction algorithm that is used won’t create overmodelling problems in this particular case?
I think it would still probably estimate the parameters well. But it is likely you would loose the top or bottom slice, which would be a pity if you only have 7. So maybe best to turn movement estimation off.
Jesper
> Thank for the help.
> Best wishes,
> Catarina
>
>
>> On 22 Aug 2016, at 18:58, Jesper Andersson <[log in to unmask]> wrote:
>>
>> Dear Catarina,
>>
>>> Dear FSL experts,
>>> I’m using the distortion correction tool, topup, for distortion correction of GRE-EPI data.
>>> I have a particular question about the setup of parameters in the config file.
>>> I acquire data with two phase encoding directions and I merge the two volumes then feed into topup as it is advised in the wiki.
>>> However, I want to assume that there was no head-motion between the two volumes and so I wanted to do a simple constrained in-plane estimation of the distortion field, so strictly slice by slice. Is it enough to set the estmov parameter to 0 in all iterations and use other default values that are on the b02b0.cnf file?
>>
>> yes, I think that should work.
>>
>> But why do you want to make that assumption? There is always the risk of movement, and my experience of topup is that the estimation of the field is not negatively affected by also estimating subject movement,
>>
>> Jesper
>>
|