Thanks so much for the response! have been away from the office and just had time to read it. I will be examining the data that you suggested.

I thought of another potential issue with our data:

Here is our generic 3dDeconvolve:

3dDeconvolve -input SUBID_pleas_h_2runs+tlrc \
-concat '1D: 0 360' \
-mask pleas_h_full_mask+tlrc \
-censor /volumes/ARIANA_HD/Users/cmurphy/Desktop/fmri_meta_data/param_analysis/censor_2runs.1D \
-num_stimts 4 -local_times \
-stim_times 1 '/volumes/ARIANA_HD/Users/cmurphy/Desktop/fmri_meta_data/param_analysis/PUMP_AB.1D/BLOCK(8,1)' \
-stim_label 1 CAFFEINE \
-stim_times 2 '/volumes/ARIANA_HD/Users/cmurphy/Desktop/fmri_meta_data/param_analysis/PUMP_BC.1D' 'BLOCK(8,1)' \
-stim_label 2 SACCHARIN \
-stim_times 3 '/volumes/ARIANA_HD/Users/cmurphy/Desktop/fmri_meta_data/param_analysis/PUMP_CA.1D' 'BLOCK(8,1) ' \
-stim_label 3 SUCROSE \
-stim_times 4 '/volumes/ARIANA_HD/Users/cmurphy/Desktop/fmri_meta_data/param_analysis/WATER.1D' 'BLOCK(8,1)' \
-stim_label 4 WATER \
-tshift \
-num_glt 6 \
-gltsym 'SYM: CAFFEINE -WATER' -glt_label 1 "caffeine-water_H" \
-gltsym 'SYM: SACCHARIN -WATER' -glt_label 2 "saccharine-water_H" \
-gltsym 'SYM: SUCROSE -WATER' -glt_label 3 "sucrose-water_H" \
-gltsym 'SYM: SUCROSE -CAFFEINE' -glt_label 4 "sucrose-caffeine_H" \
-gltsym 'SYM: SUCROSE -SACCHARIN' -glt_label 5 "sucrose-saccharin_H" \
-gltsym 'SYM: SACCHARIN -CAFFEINE' -glt_label 6 "saccharin-caffeine_H" \
-full_first -fout -tout -rout \
-bucket SUBID_2runs_H_P_bucket \
-xjpeg SUBID_2runs_H_P_bucket_xmat.jpg \
-x1D SUBID_2runs_H_P_bucket_xmat.x1D


Here is the potential issue:

At the end of our stim_times command line we have our BLOCK duration set to 8 seconds…. Our actual BLOCK time in our protocol is 10 seconds. We put 8 as the duration because we wanted to censor out the first 2 seconds of each block. This was written into the command a while ago and has just been accepted…

I have a feeling that was a big mistake.

My guess is that setting the duration to 8 does not censor out the first 2 seconds of the 10s block? In order to censor out the first 2 seconds of each block we would need a censor .1D file that had 1's and 0's in the appropriate times (in TR's) to censor out the first 2 second of each block. (or we could use CENSORTR …(specific TR's))

The first 2 seconds is when our subjects are swallowing the stimuli… this causes some head motion, hence, the desire to censor.

Am i correct in assuming that putting 8 as the duration will not censor the first two seconds?

If so, we should put 10 as the duration and censor out the first 2 seconds (1 TR) with a censor file or CENSORTR… but how will this affect the model?