We have made a python script that uses 3dAllineate's "lpc" cost function
for superior EPI to Structural dataset alignment.

The method seems to work very well - even for previously difficult to align
datasets. See test results summarized in the HBM 2008 abstract:
False Sense of EPI-to-Structural Alignment with Common Cross-Modality Registration
Methods (http://afni.nimh.nih.gov/sscc/rwcox/abstracts/)

This new script should make the job of aligning EPI and anatomical datasets
much easier. The script will do:

- slice timing correction,
- time series volume registration to correct for motion,
- align the EPI and anatomical datasets,
- skull strip the data,
- compute a talairach-transformation to a standard template if requested,
- dry clean your winter clothes

Basic Usage:
align_epi_anat.py -anat anat+orig -epi epi+orig -epi_base 5

Another Example (slightly more advanced):
# Instead of aligning the anatomy to an epi, transform the epi
# to match the anatomy. (-epi2anat vs -anat2epi)
# Do other epi datasets at the same time
# Children get the same treatment. Note that
# epi sub-bricks are transformed once in the process.

align_epi_anat.py -anat sb23_mpra+orig -epi epi_r03+orig \
-epi_base 5 -child_epi epi_r??+orig.HEAD \
-epi2anat -suffix al2anat

See align_epi_anat.py -help for detailed information and sample command lines.