Interesting question, but it's fairly complicated and likely not worth it. We don't have a program or script to do this now. Your data is 1x1x5mm over two orientation directions. The combination of the dataset still will have holes (and fairly large ones) where you don't have data. Add to that the fact this data doesn't cover the same intensity ranges and spatial extents, and it makes this problem very difficult because you will also require good volume registration and intensity normalization between the two. You will be better off reacquiring your structural data at 1x1x1mm.