It's difficult to define an anatomically realistic ROI without a template
registration. I know that FSL/Freesurger have ways of defining an area
based on gyri/sulci anatomical lanmarks and some other features. However
these is not straight forward and I don't know how accurate it is
(considering interindividual differences and some image imperfections), but
for your goal sounds like the only way. Well... unless you want to spend a
lot of time manually delineating a structure on several slices using a
piece of software like the ones for volumetry purposes. If you are
anatomically trained and precise it may be a useful approach.
Anyway, good luck and I hope it works for you.
Iwo
On Aug 10 2011, [log in to unmask] wrote:
> Hello,
>
>> I also thought about yet another possibility. If you want to avoid any
>> spatial distortion related to registration, you might want to consider
>> using anatomical landamark registration as implemented in Fressurfer and
>> then you can use different parcellation methods. I don't know much about
>> this method but what I know is that you brains must be in a good state,
>> eg:
>>
>> no lesions or other "natural distortions" ie healthy good quality
>> brains, another thing is that this registration is quite computationally
>> gready and
>>
>> takes a lot of time. I don't know if it could be of any help for you?
>
> yes I think this will help... I'll have a check... I'm just a little
> surprised, that there seems no easy way to create a 3D ROI...
>
>Thanks
>
>Andreas
>
> Arcor empfiehlt: Die TOP 1000 Surftipps zu allem, was im Internet
> interessant und spannend ist - sehen Sie hier:
> http://www.arcor.de/content/pc_technik/internet/surftipps/surftipps/
>
--
Iwo
|