Hi Gang,

I had a chance to run this with the 3dSynthesize option on the all_runs.${subject}+tlrc file with the interval specified to create a file with the regressors of no-interest for subtracting out and then extracting time series from a ROI. However, it seems to not be working properly. Here is the commands I ran:

1)3dSynthesize -prefix ${subject}_effectsNointerest -matrix Gamma_X_cbucket.xmat.1D -cbucket c_stats_${subject}.Gam+tlrc -select 0..19 28..33

Then I ran 3dcalc to subtract the above file from my all_runs.${subject}+tlrc file:
2)3dcalc -a all_runs.${subject}+tlrc -b ${subject}_effectsNointerest -expr 'a-b' -prefix Clean_Data_${subject}.tlrc

I then used an amygdala mask to extract the time series from the above Clean_Data_${subject} file:
3)3dmaskave -mask bilateral_amy.nii.gz -quiet CleanData_${subject}+tlrc'[0..238]' >${subject}.Seed_run_1.1D

The output of Seed_run_1.1D has two very large spikes (e.g., 99.83) with a few time points per spike. The remainder of signal in the .1D file is less then 1. Are these spikes due to large movement? I thought I had subtracted that out from the data in the 3dcalc step.

When I ran steps 1 & 2 from [afni.nimh.nih.gov] I do not have this problem.

Any ideas on ways to fix this issue?

I looked at the output of 3dSynthesize and saw that there were data points censored, could these large spikes be the time points that were censored? Should I not censor these time points (I believe there is another post regarding this)?

Thanks,

Danny