History of AFNI updates  

|
August 25, 2014 05:31PM
Thanks for the suggestions, Rick and Peter. I tried -twopass and adding extra zeropadding but neither improved the results. I also tried uber_align_test.py and had some success. Using the script below, I was able to get the EPI to align with the anatomical underlay "anat_al_lpc+ZZ." This is all in orig space. I'll try adding the lpc cost to afni_proc.py to see if it improves the outcome in tlrc space.

# created by uber_align_test.py: version 0.7 (September 22, 2011)
# creation date: Mon Aug 25 17:05:46 2014

# --------------------- set processing variables ---------------------

# top data directory
set top_dir = /raid10/studies/varenicline/subjects/subject_results/group.Varenicline_Faces/post_alc/jjl.results1

# input dataset options (ebase is EPI index)
set in_anat = $top_dir/mprage_series010+orig.HEAD
set in_epi = $top_dir/pb00.jjl.r01.tcat+orig.HEAD
set in_ebase = 4
set center_base = TT_N27+tlrc

# main options
set cost_main = lpc
set cost_list = ( lpc+ZZ lpa nmi )

# all other align_epi_anat.py options
set align_opts = ( -tshift off \
-volreg off \
-giant_move \
)

# ------------------- test and create results dir -------------------

# note directory for results
set results_dir = align.results

# make sure it does not yet exist
if ( -e $results_dir ) then
echo "** results dir '$results_dir' already exists"
exit
endif

# create results directory, where the work will be done
mkdir $results_dir

# ---------------------------- copy data ----------------------------

# copy dataset to processing directory
3dbucket -prefix $results_dir/anat $in_anat
3dbucket -prefix $results_dir/epi $in_epi"[$in_ebase]"

# enter the processing directory
cd $results_dir

# -------------------------- align centers --------------------------

# since altering grid, remove any oblique transformation
3drefit -deoblique anat+orig epi+orig

# align volume centers (we do not trust spatial locations)
@Align_Centers -no_cp -base $center_base -dset anat+orig
@Align_Centers -no_cp -base $center_base -dset epi+orig

# ---------------------------- align data ----------------------------

# test alignment, using variables set above
align_epi_anat.py -anat anat+orig -epi epi+orig -epi_base $in_ebase \
-cost $cost_main -multi_cost $cost_list $align_opts
Subject Author Posted

poor alignment after align_epi_anat.py

jgowin August 22, 2014 02:26PM

Re: poor alignment after align_epi_anat.py

Peter Molfese August 22, 2014 03:42PM

Re: poor alignment after align_epi_anat.py

jgowin August 22, 2014 03:52PM

Re: poor alignment after align_epi_anat.py

rick reynolds August 24, 2014 08:40PM

Re: poor alignment after align_epi_anat.py

jgowin August 25, 2014 05:31PM

Re: poor alignment after align_epi_anat.py

rick reynolds August 25, 2014 08:47PM

Re: poor alignment after align_epi_anat.py

jgowin August 27, 2014 09:51AM

Re: poor alignment after align_epi_anat.py

rick reynolds August 27, 2014 02:53PM