Dear AFNI developers,
I’m running 3dTfitter in preparation for a PPI analysis.
The output time series of 3dTfitter has large spikes at the beginning and at the end of the run (spikes.jpg).
The time series of the ROI seem ok to me (roi.jpg).
Below, I include the commands I used, together with the output.
Could you help me in identifying what the source of the spikes might be?
Thank you!
All the best,
Irene
————————————————————————————
num_runs=1
TR=0.9 #(0.901)
stim_dur=1
min_frac=0.3
run_len=720
run_vol=800
3dDeconvolve -float -jobs 2 -x1D ppidir/{sub}_RegNoInt -input basedir/{sub}.results/pb05.${sub}.r??.scale+tlrc.HEAD -num_stimts 12
-polort A
-stim_file 1 basedir/{sub}.results/‘motion_demean.1D[0]’ -stim_base 1 -stim_label 1 motion_roll
-stim_file 2 basedir/{sub}.results/‘motion_demean.1D[1]’ -stim_base 2 -stim_label 2 motion_pitch
-stim_file 3 basedir/{sub}.results/‘motion_demean.1D[2]’ -stim_base 3 -stim_label 3 motion_yaw
-stim_file 4 basedir/{sub}.results/‘motion_demean.1D[3]’ -stim_base 4 -stim_label 4 motion_x
-stim_file 5 basedir/{sub}.results/‘motion_demean.1D[4]’ -stim_base 5 -stim_label 5 motion_y
-stim_file 6 basedir/{sub}.results/‘motion_demean.1D[5]’ -stim_base 6 -stim_label 6 motion_z
-stim_file 7 basedir/{sub}.results/‘motion_deriv.1D[0]’ -stim_base 7 -stim_label 7 motion_roll_deriv
-stim_file 8 basedir/{sub}.results/‘motion_deriv.1D[1]’ -stim_base 8 -stim_label 8 motion_pitch_deriv
-stim_file 9 basedir/{sub}.results/‘motion_deriv.1D[2]’ -stim_base 9 -stim_label 9 motion_yaw_deriv
-stim_file 10 basedir/{sub}.results/‘motion_deriv.1D[3]’ -stim_base 10 -stim_label 10 motion_x_deriv
-stim_file 11 basedir/{sub}.results/‘motion_deriv.1D[4]’ -stim_base 11 -stim_label 11 motion_y_deriv
-stim_file 12 basedir/{sub}.results/‘motion_deriv.1D[5]’ -stim_base 12 -stim_label 12 motion_z_deriv
-GOFORIT 5
-cbucket ppidir/{sub}_correction
3dSynthesize -cbucket ppidir/{sub}_correction+tlrc -matrix ppidir/{sub}_RegNoInt.xmat.1D -select allstim -prefix {ppidir}/{sub}_temp
3dcalc -float -a basedir/{sub}.results/pb05.${sub}.r01.scale+tlrc -b ppidir/{sub}_temp+tlrc -expr ‘a-b’ -prefix ppidir/{sub}_clean+tlrc
3dmaskave -mask basedir/{sub}.results/mask+tlrc -mrange 1 1 -quiet ppidir/{sub}_clean+tlrc> $ppidir/mask.1D
waver -dt ${TR} -GAM -inline 1@1 > $ppidir/gam_hrf.1D
3dTfitter -RHS mask.1D -FALTUNG ppidir/gam_hrf.1D {ppidir}/neuro_resp_mask.1D 012 0
————————————————————————————
++ 3dDeconvolve: AFNI version=AFNI_21.0.12 (Feb 25 2021) [64-bit]
++ Authored by: B. Douglas Ward, et al.
++ loading dataset /fenix/users/irepe/elefanten/elefanten_analysis/emo/singlesub/C1_sub01.results/pb05.C1_sub01.r01.scale+tlrc.HEAD
++ STAT automask has 241222 voxels (out of 311296 = 77.5%)
++ Skipping check for initial transients
++ Imaging duration=720.8 s; Automatic polort=5
++ Number of time points: 800 (no censoring)
- Number of parameters: 18 [18 baseline ; 0 signal]
++ total shared memory needed = 23,658,496 bytes (about 24 million)
++ mmap() memory allocated: 23,658,496 bytes (about 24 million)
++ Memory required for output bricks = 23,658,496 bytes (about 24 million)
++ Wrote matrix values to file /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/C1_sub01_RegNoInt.xmat.1D
++ ========= Things you can do with the matrix file =========
++ (a) Linear regression with ARMA(1,1) modeling of serial correlation:
3dREMLfit -matrix /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/C1_sub01_RegNoInt.xmat.1D -input /fenix/users/irepe/elefanten/elefanten_analysis/emo/singlesub/C1_sub01.results/pb05.C1_sub01.r01.scale+tlrc.HEAD
-Rbeta /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/C1_sub01_correction_REML
-Rbuck Decon_REML -Rvar Decon_REMLvar -verb
++ N.B.: 3dREMLfit command above written to file Decon.REML_cmd
++ (b) Visualization/analysis of the matrix via ExamineXmat.R
++ (c) Synthesis of sub-model datasets using 3dSynthesize
++ ==========================================================
++ ----- Signal+Baseline matrix condition [X] (800x18): 4.31517 ++ VERY GOOD ++
++ ----- Baseline-only matrix condition [X] (800x18): 4.31517 ++ VERY GOOD ++
++ ----- stim_base-only matrix condition [X] (800x12): 3.72941 ++ VERY GOOD ++
++ ----- polort-only matrix condition [X] (800x6): 1.00681 ++ VERY GOOD ++
++ +++++ Matrix inverse average error = 3.00302e-15 ++ VERY GOOD ++
++ Matrix setup time = 0.04 s
++ Voxels in dataset: 311296
++ Voxels per job: 155648
++ Job #1: processing voxels 155648 to 311295; elapsed time=9.367
++ Job #0: processing voxels 0 to 155647; elapsed time=9.377
++ voxel loop:0123456789.0123456789.0123456789.0123456789.0123456789.
++ Job #0 waiting for children to finish; elapsed time=35.312
++ Job #1 finished; elapsed time=35.402
++ Job #0 now finishing up; elapsed time=35.412
++ Wrote cbucket to /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/C1_sub01_correction+tlrc.BRIK
++ Wrote bucket dataset into ./Decon+tlrc.BRIK
++ Program finished; elapsed time=35.646
mv: cannot stat ‘3dDeconvolve.err’: No such file or directory
++ 3dSynthesize: AFNI version=AFNI_21.0.12 (Feb 25 2021) [64-bit]
++ Authored by: RW Cox
++ Output has 800 time points at TR=0.901
++ Calculating: …!
++ Output dataset /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/C1_sub01_temp+tlrc.BRIK
++ CPU time=0.00 s ; Elapsed=10.49 s
done 3dsynth sub${sub}
++ 3dcalc: AFNI version=AFNI_21.0.12 (Feb 25 2021) [64-bit]
++ Authored by: A cast of thousands
++ Output dataset /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/C1_sub01_clean+tlrc.BRIK
++ 3dmaskave: AFNI version=AFNI_21.0.12 (Feb 25 2021) [64-bit]
+++ 15 voxels survive the mask
done 3dmaskave_activation cluster sub${sub}
++ 3dTfitter: AFNI version=AFNI_21.0.12 (Feb 25 2021) [64-bit]
++ Authored by: RWCox
- default penalty scale factor=59.3309
++ Fit worked on all 1 voxels attempted - Writing FALTUNG dataset: /fenix/users/irepe/elefanten/elefanten_analysis/emo/PPI/singlesub/C1_sub01/neuro_resp_Clust_ATF_FFA_mask.1D
++ Total CPU time = 0.0 s
