With the refactoring of T26350 the test for HotspotMaskGenerator are broken. It is not clear why but also as the state of the generator is unclear, they where just deactivated for now until we have a final decission.

Status:

• The generator is currently not used in our code base
• The purpose of it is e.g. in the context of PET imaging, where Hotspot masking is used
  • But I am not sure if for PET applications and alike, this implicit usage is the best. Problem: when used as a generator, you don't get any feedback on the part of the image your statistics where actually generated. More transperent (but may be not so slick in terms of UX) would be to add it as a seg tool that allows you to generate a hot spot mask. That can then be selected in the statistics.
• The implementation is very hand crafted. I guess at least one could use itk::ShapedNeighborhoodIterator or alike to need to craft the double loop for the sliding window on its one.

We need to decide if we:
a) completly remove the code
b) or fix the code and the test
c) or make it a seg tool (this can also be done in combination with a, remove the code and make a ticket for later introduction of c)