marcniethammer

I am not sure what you mean by 3D images with two different dimensions, but the method itself is currently geared at working with 3D patches. See: https://arxiv.org/abs/1703.10908

This is not what is it designed for. However, since it works based on patches there may be some hope that it could work. This would, of course, requires some...

Looks like this is a multi-modal test case. Can you run a case without instance optimization? If this works then it is possible that switching the similarity measure from 1-LNCC...