Dear Gabor,
The training data that FIRST is based on has standard T1-weighted contrasts. If your data does not have this type of contrast then it will not match the training data and FIRST is unlikely to work well. If you cannot use FIRST then you might want to try a registration-based solution, although this will require your own training data (or at least a single atlas).
You can use your own mask for the registration steps, just look at how this is done in the script.
All the best,
Mark
On 16 Jan 2014, at 11:41, Gabor Perlaki <[log in to unmask]> wrote:
> Dear all,
>
> Does FSL-FIRST work on a contrast enhanced T1 image? Is it possible to use an own mask for the stage 2 registration? For instance, I'm interested only in the hippocampus and I'd like to optimize the registration for that structure.
>
> Thanks,
> Gabor
|