Hi,
Leaving some eye/optic nerve in the images will have a negligible
effect on the registration generally. We only use the brain extracted
images for linear registration and that is pretty robust. The non-linear
registration uses the images without brain extraction. Hence I would
not worry about the eye and optic nerve unless you are doing SIENA.
All the best,
Mark
On 28 Jan 2011, at 01:28, Amy wrote:
> Hello,
>
> I'm running the bet on some T1 weighted images, but every now and then I still have the eye/optic nerve left on one side after the brain extraction. I'm not sure why it's not removing them bilaterally. I've tried the standard brain extraction, also the robust brain centre estimation, along with playing with the thresholds, but still, I do not get satisfactory results, unless I run the Eye & optic nerve cleanup, but that usually takes 2-3 hours per image. Does anyone know what else I could try to optimize the brain extraction while not taking a couple of hours?
>
> Another question--would leaving an eye in the brain extracted image affect much of the registration results?
>
> Thanks,
> Amy
>
|