Hi afni experts,

I have data over 8 runs (each run about 5 min.) and I am concerned that I might have some motion issues. I am thinking of applying motion censoring in 3ddeconvolve (in addition to regressing motion parameters). In order to determine the time points to censor, I am using "maxdisp1D" command in 3dvolreg (my commands are below), but for most subjects in quite a few of the runs (especially the later runs), almost all (and sometimes all) of the time points in a run are above my threshold (4mm). I believe the reason for this is slow drift over the course of my runs. I don't want to delete half of my data, is there another way to get around this? Should I try a different method for motion censoring?

3dvolreg -tshift 1 -prefix BOLD_R7_Volreg \

-maxdisp1D maxMotion.1D\

-linear -verbose -base '../6/BOLD_R1_Despike+orig[78]' \

-dfile BOLD_R7_motion.txt BOLD_R7_Despike+orig

cat maxMotion.1D | awk '{if ($1 > 4) print 1; else print 0}' > censor.1D


Thanks!

Rengin