# censored TTs disagrees with #TRs censored per stimulus class

Hey,

I have some task data (only one run) from a subject that I’ve analyzed using afni_proc. There were three stimulus classes provided. The output of the out.ss_review.redacted.txt has me somewhat perplexed. Below, it clearly states that there were 13 TRs censored per run, but then even farther below that it says that there were 11 2 10 (=23) TRs censored per stimulus.

At most, in the case of one run, I would expect sum(#TRs censored per stim) <= # TRS censored per run, since some TRs to be included in the baseline could also be censored and would not be reflected in #TRs censored per stim. There has to be something amiss somewhere. Any idea where?


subject ID                : MDD_1008
AFNI version              : AFNI_19.0.09
AFNI package              : linux_ubuntu_16_64
TRs removed (per run)     : 0
num stim classes provided : 3
final anatomy dset        : anat_final.MDD_1008+tlrc.HEAD
final stats dset          : stats.MDD_1008_REML+tlrc.HEAD
final voxel resolution    : 2.000000    2.000000        2.000000

motion limit              : 0.3
num TRs above mot limit   : 3
average motion (per TR)   : 0.0486667
average censored motion   : 0.0390909
max motion displacement   : 2.14551
max censored displacement : 1.86535
outlier limit             : 0.2
average outlier frac (TR) : 0.0249277
num TRs above out limit   : 10

num runs found            : 1
num TRs per run           : 254
num TRs per run (applied) : 241
num TRs per run (censored): 13
fraction censored per run : 0.0511811
TRs total (uncensored)    : 254
TRs total                 : 241
degrees of freedom used   : 20
degrees of freedom left   : 221

TRs censored              : 13
censor fraction           : 0.051181
num regs of interest      : 3
num TRs per stim (orig)   : 211 123 125
num TRs censored per stim : 11 2 10
fraction TRs censored     : 0.052 0.016 0.080
ave mot per sresp (orig)  : 0.044266 0.039016 0.048739
ave mot per sresp (cens)  : 0.034599 0.034229 0.032117

TSNR average              : 171.687
global correlation (GCOR) : 0.0397436
anat/EPI mask Dice coef   : 0.667881
maximum F-stat (masked)   : 9.38989
blur estimates (ACF)      : 0.849002 2.44407 10.7644
blur estimates (FWHM)     : 0 0 0

Hi Colm,

Those labels are short, so they are not fully descriptive. In the case of ‘num TRs censored per stim’, it refers to the regressor, not the stimulus presentation period. It is the number of non-zero BOLD response time points that are censored.

For less detail (on that term :), please see, “gen_ss_review_scripts.py -help_fields”.

  • rick