Dear all,
I am having a problem using CENSORTR in 3dDeconvolve when I also have a stim_base file in the model (this case motion files). The output I get when I try and run it is below (Script runs fine when I take the –CENSORTR option out).
I did not want these points in analysis since visual inspection of motion file reveled movement above our threshold, and when I first ran the analysis I did not take motion as a factor in the model.
My question is whether these points really do need to be censored out when you put the motion variables in your regressor model. I am not sure.
Thank you very much in advance (The script is below the output).
L.
OUTPUT
--------------
Liat data:US_Only_Exp_GamRegress.sh
++ No '-bucket' option given ==> using '-bucket Decon'
++ 3dDeconvolve: AFNI version=AFNI_2007_05_29_1644 (Jun 15 2007) [32-bit]
++ Authored by: B. Douglas Ward, et al.
++ reading dataset 4856_r1_psc+orig 4856_r2_psc+orig 4856_r3_psc+orig 4856_r4_psc+orig 4856_r5_psc+orig
++ Auto-catenated datasets start at: 0 102 204 306 408
*+ WARNING: Input polort=1; Imaging duration=204.0 s; Recommended minimum polort=2
++ Wrote matrix values to file Decon.xmat.1D
++ Signal+Baseline matrix condition [X] (490x10): 1.02265 ++ VERY GOOD ++
++ Baseline-only matrix condition [X] (490x10): 1.02265 ++ VERY GOOD ++
++ -polort-only matrix condition [X] (490x10): 1.02265 ++ VERY GOOD ++
++ Matrix inverse average error = 1.54875e-16 ++ VERY GOOD ++
++ Matrix setup time = 0.21 s
++ Calculations starting; elapsed time=2.291
++ voxel loop:0123456789.0123456789.0123456789.0123456789.0123456789.
++ Calculations finished; elapsed time=9.664
*+ WARNING: changed output dataset name from 'Decon' to 'Decon_AA3'
++ Wrote bucket dataset into ./Decon_AA3+orig.BRIK
++ #Flops=4.9477e+09 Average Dot Product=19.6
tcsh: -nobout: Command not found.
tcsh: mv: No match.
tcsh: mv: No match.
SCRIPT
========
#!/bin/tcsh
foreach subj (4856) #(4315 4322 4358 4378 4384 4467 4501 4689 4693 4703 4704 4710 4716 4720 4814 4843 4852 4870)
cd /Volumes/d1/Liat/data/${subj}/functional/afni
fixMacfile.sh ${subj}.txt #fix formatting from Mac to unix
# this step convolves the stim file with the gamma function
foreach n (0 1 2 3)
waver -GAM -numout 510 -peak 1 -TR 2.0 -input "${subj}.txt[${n}]" >w${n}.1D
end
1dcat w0.1D w1.1D w2.1D w3.1D >${subj}_gam_hrf.1D
rm w*.1D
#3dTcat -prefix ${subj}_all_runs_pscA\
#${subj}_r1_psc+orig \
#${subj}_r2_psc+orig \
#${subj}_r3_psc+orig \
#${subj}_r4_psc+orig \
#${subj}_r5_psc+orig
3dDeconvolve -input ${subj}_r1_psc+orig ${subj}_r2_psc+orig ${subj}_r3_psc+orig ${subj}_r4_psc+orig ${subj}_r5_psc+orig \
-CENSORTR 3:38..57 \
-nobout \
-bucket ${subj}_statsMotion \
-cbucket ${subj}_coeff_fittsMotion \
-full_first \
-xsave \
-polort 2 \
-fitts ${subj}_fittsMotion\
-legendre \
-num_stimts 10 \
-stim_file 1 ${subj}_gam_hrf.1D'[0]' -stim_label 1 pi \
-stim_file 2 ${subj}_gam_hrf.1D'[1]' -stim_label 2 pt \
-stim_file 3 ${subj}_gam_hrf.1D'[2]' -stim_label 3 ni \
-stim_file 4 ${subj}_gam_hrf.1D'[3]' -stim_label 4 nt \
-stim_file 5 ${subj}_all_motion.1D'[0]' -stim_base 5 \
-stim_file 6 ${subj}_all_motion.1D'[1]' -stim_base 6 \
-stim_file 7 ${subj}_all_motion.1D'[2]' -stim_base 7 \
-stim_file 8 ${subj}_all_motion.1D'[3]' -stim_base 8 \
-stim_file 9 ${subj}_all_motion.1D'[4]' -stim_base 9 \
-stim_file 10 ${subj}_all_motion.1D'[5]' -stim_base 10 \
-tout \
-num_glt 4 \
-gltsym 'SYM: pt -pi' -glt_label 1 pt-pi \
-gltsym 'SYM: nt -ni' -glt_label 2 nt-ni \
-gltsym 'SYM: nt -pt' -glt_label 3 nt-pt \
-gltsym 'SYM: ni -pi' -glt_label 4 ni-pi
end