Hi, I am analyzing a fairly straight forward block design experiment, but in addition to estimating the stimulus driven activity I need to estimate the activity from individual blocks of fixation. Here it would seemingly make more sense to avoid adding any HRF to the model during these blocks (but of course keep the HRF for the stimulus blocks) but is there a way to do this with 3dDeconvolve? Using the -stim_times_IM option I see that it is possible to obtain the beta weights for each block of the experiment, but what is the appropriate way to obtain an estimate of activity during individual blocks of rest?

If for example, I wanted to do a linear contrast of the second fixation block vs. the 4th fixation block (there are actual reasons for wanting to do this), is there a way to do this without indicating a basis function for these periods of fixation? I feel that:

3dDeconvolve -input expdata+tlrc.HEAD \
-stim_times_IM 1 Fixate.1D 'BLOCK(20,1)' \
-stim_times_IM 2 Stimulus.1D 'BLOCK(20,1)'

is problematic since it would be modeling the resting blocks explicitly.

Is it preferable to avoid this completely and instead extract the residuals of these time points from the nuisance regressors, via 3dTproject, separately?
Any advice would be greatly appreciated, thank you so much for your time.
-db