How to register two images that do not have the same information?

I have two MRI images, but one image has the complete head and the other has some parts of the head cut, including some parts of the brain. So how can I register these images?

PS: My initial investigation was firstly do an affine registration, then create a mask that separate the parts in the fixed image that do not appear in the moving image, and then do a deformable registration. I’ve done this, but I don’t know if this approach is correct.

Can someone give some ideas, or indicate me some article in this subject?


Sounds fine to me. But what do you want to do/calculate with the registered images?

You could also try applying the resulting transform to the original (full extent) image.

1 Like

I’m trying just to segment some areas of the brain. I will register a “partial volume” with an atlas, and after that by investigating the overlaying areas, segment the brain.

Specifically in this case I can’t apply the transformation in the whole image because the “entire image” is the fixed volume. So my attempt is just trying that my algorithm do not “see” the areas that are not included in the “partial moving volume”.