Hi Robin,

Thanks. We do not actually consider bumping rude. It is hard to keep track of all the threads, so it is helpful when we are reminded of something lingering.

1) Is that part of an align_epi_anat.py command? Can you post it (or you can mail me the full output.proc.* file).

2) Was it really removed from 6b? I don't recall it being in there. And that example is only about 2 weeks old, as a more complete version of example 6.

We usually apply despike in either resting state, or when we expect the subjects to be high-motion ones. For a straightforward task case, we usually do not apply it (though it is fine to do so).

3) Example 6b has -regress_3dD_stop because it also has -regress_reml_exec, and the expectation is to use the 3dREMLfit stats, rather than those from 3dDeconvolve. So the only point of -regress_3dD_stop is to save disk space.

Similar to despike, anaticor is more commonly applied in the case of resting state analysis, though it can be used for task as well. But the (voxelwise) anaticor regression cannot be done with 3dDeconvolve, for that 3dREMLfit is required.

4) Similarly, we tend to apply the 'deriv' parameters in the case of resting state analysis, not typically for task. Though again, it is okay to do so.


In a resting state analysis, we are more careful with motion and censoring in general. You will notice that all of these options (despike, anaticor, motion deriv regression) are applied in example 11.

afni_proc.py -show_example 'example 11'

- rick