in aea, are -multicost and -AddEdge incompatible?

AFNI Message Board
1

Hey,

I’m trying to align a lot of EPI data with poor contrast (there are no pre steady state volumes with sufficient T1 weighting to make using the LPC cost function by default viable for all subjects) with my T1 data.

To get a sense of what cost function provides the best alignment for a particular subject I’m running (from linux_openmp_64: Sep 4 2016 (Version AFNI_16.2.16)):


align_epi_anat.py -anat2epi			\
		  -anat ${anatFile}		\
		  -epi ${epiFile}		\
		  -epi_base 0			\
		  -volreg off			\
		  -tshift off			\
		  -cost lpc			\
		  -multi_cost lpa lpc+ZZ mi     \
		  -AddEdge

When I look in the subject’s AddEdge directory the only files in there are those referring to the default LPC cost metric, For example, for one subject I get the following:


105_A.anat_unif.zp_al+orig.BRIK.gz        105_A.anat_unif.zp_al_rs+orig.HEAD     _ae.ExamineList.log                                  vr_base_ns_ec+orig.BRIK.gz
105_A.anat_unif.zp_al+orig.HEAD           105_A.anat_unif.zp_ns_e3+orig.BRIK.gz  vr_base_ns_105_A.anat_unif.zp_al_rs_ec+orig.BRIK.gz  vr_base_ns_ec+orig.HEAD
105_A.anat_unif.zp_al_rs_e3+orig.BRIK.gz  105_A.anat_unif.zp_ns_e3+orig.HEAD     vr_base_ns_105_A.anat_unif.zp_al_rs_ec+orig.HEAD     vr_base_ns+orig.BRIK.gz
105_A.anat_unif.zp_al_rs_e3+orig.HEAD     105_A.anat_unif.zp_ns_ec+orig.BRIK.gz  vr_base_ns_105_A.anat_unif.zp_ns_ec+orig.BRIK.gz     vr_base_ns+orig.HEAD
105_A.anat_unif.zp_al_rs_ec+orig.BRIK.gz  105_A.anat_unif.zp_ns_ec+orig.HEAD     vr_base_ns_105_A.anat_unif.zp_ns_ec+orig.HEAD
105_A.anat_unif.zp_al_rs_ec+orig.HEAD     105_A.anat_unif.zp_ns+orig.BRIK.gz     vr_base_ns_e3+orig.BRIK.gz
105_A.anat_unif.zp_al_rs+orig.BRIK.gz     105_A.anat_unif.zp_ns+orig.HEAD        vr_base_ns_e3+orig.HEAD

Are these combinations of arguments incompatible? If so could aea be modified so that it is possible to use multicost with addedge?

I suspect the problem lies around line 3449 in align_epi_anat.py where the previous contents of AddEdge are deleted within the loop iterating over the all of the cost functions.

2

any takers?

3

I would be the “taker”, but I haven’t had time to look. You can use the @AddEdge script independently. Just resample all datasets to the same grid.

4

Hi there,

Bumping this thread. I’ve been trying to do the same thing (use both -multi_cost and -AddEdge options for -align_epi_anat.py), and I’ve been running into the same problems. It seems like the two arguments are incompatible, so I have to pick one over the other, and I haven’t been as successful with running @AddEdge by alone. Has there been any progress made on integrating these arguments so they can both play nice with afni_proc.py?

Best,
Debbie