You have made a valiant effort. It seems you may be doing a MEMRI experiment, and alignment can be tricky because of the variation in intensities across time. The problem becomes more difficult especially if you also have partial coverage too. Check one solution using align_epi_anat.py and the "edge" method. See an example here in this presentation:
[
afni.nimh.nih.gov]
If you can not find something that works for you, you may upload the data, and I can take a look.