Yes - excellent suggestion Andreas!
Also, if that looks a little bit out (or if you don't have the right
sforms/qforms)
then you might like to use Nudge in order to manually refine the
registration.
This is probably the easiest solution at this point, so don't bother
trying
to make a manual mask, just use flirt and/or nudge.
All the best,
Mark
On 11 May 2009, at 16:53, Andreas Bartsch wrote:
> Hi,
>
> in addition to what MJ suggested:
> if you FM-images were recorded in the same session as the images you
> want to register them to and if the subject hasn't moved that much
> you may get a good result by just using the DICOM coordinate /
> positioning entries. Flirt can do that for you (if the s/qform has
> been set properly).
> Cheers-
> Andreas
>
> ________________________________________
> Von: FSL - FMRIB's Software Library [[log in to unmask]] im Auftrag
> von Mark Jenkinson [[log in to unmask]]
> Gesendet: Montag, 11. Mai 2009 17:48
> An: [log in to unmask]
> Betreff: Re: [FSL] Unwarping, Phase but no Magnitude image
>
> Hi,
>
> The order of the steps was as you wrote.
> Highpass filtering should give you an image that
> highlights the areas of noise in the background
> (as these are very "rough" and full of high-frequency
> content). The hope is then that smoothing (after
> taking the abs) generates something where the
> average value in the background is higher than
> inside the brain and you could threshold it.
> However, this is far from guaranteed to work, so
> it might just be that it won't work at all. I've never
> tried it - it was just an idea.
>
> All the best,
> Mark
>
>
> On 11 May 2009, at 14:49, Nikos Green wrote:
>
>> Thanks very much Mark,
>>
>> i will try that. Just one quick question concering the first step:
>> In what order do the proposed steps apply. I tried spatial highpass
>> filtering, abs. values and smoothing but i do not get an image from
>> which I
>> can make a mask easily. Am i missing sth.?
>>
>> Thanks,
>> Nikos
>>
|