[ITK] Smoothing under mask

[Matt McCormick]

Hi Andras,

Perhaps replacing the masked region with the average brain value for
with a dilated value [1] will yield better results.

HTH,
Matt

[1] http://www.itk.org/Doxygen/html/classitk_1_1GrayscaleDilateImageFilter.html

On Mon, Sep 14, 2015 at 1:31 PM, András, Osztroluczki
<andras.osztroluczki at gmail.com> wrote:
> Matt, and Bradly, thanks for your answers. Unfortunately the suggested
> solutions are not good for me...:(
> My mask basically the brain, without the skull. If I set the non-masked
> pixels to 0, those 0-s still will be used during the smoothing --> the
> boundary of the brain will be "darker". If i smooth the whole image and then
> crop --> the boundary of the brain will be "lighter", because of the high
> bone intensities. That's why i want that the computation only use the masked
> values. Any suggestions for this particular problem ?
>
> Regards,
> Andras
>
>
>
> 2015-09-14 16:34 GMT+02:00 Bradley Lowekamp <blowekamp at mail.nih.gov>:
>>
>> Hello,
>>
>> Why is cropping the masked region from the smoothed not good?
>>
>> What do you expect he boundary condition around the mask to be? You could
>> set the non-masked pixels to zero, and then run the smoothing on the whole
>> image or a crop/ROI of the image?
>>
>> Brad
>>
>> On Sep 14, 2015, at 3:12 AM, András, Osztroluczki
>> <andras.osztroluczki at gmail.com> wrote:
>>
>> Dear ITK users,
>>
>> Is there a way to perform smoothing (for example with:
>> RecursiveGaussianImageFilter) under a masked region only (e.g.: the
>> computation only use those pixel values which are marked by the mask).
>> Smoothing the whole image and crop the masked region from the smoothed
>> image is not good for me unfortunately.
>>
>> Thanks,
>> Andras
>> _______________________________________________
>> Community mailing list
>> Community at itk.org
>> http://public.kitware.com/mailman/listinfo/community
>>
>>
>
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/mailman/listinfo/community
>