Hello Marko,
thanks for the quick reply! I think I go with realign & unwarp then, but together with motion parameters.
What I had in mind:
1) "realign & unwarp" in combination with no motion regressors in the models = with the default settings one corrects for motion in the data but one does not control for interpolation errors due to translations and some of the rotations. But usually we want to do so by adding the movement regressors (?).
2) "realign & unwarp" means one has to reslice the data, which introduces some interpolation errors. Do the benefits outweigh these errors? With the fieldmap probably yes, but there might be circumstances where they are rather corruptive (leaving aside aspects like stimulus-correlated motion).
3) Some of the rotations are re-estimated several times during "realign & unwarp", but only the original rp files are saved and can be added to the model.
[Kind of off-topic and concerning realignemnt in general: Actually it would be interesting to have a closer look at motion parameters in relation to the high-pass filter. If the subject moves very slowly, so more or less a linear drift, what does remain after filtering? Rather nothing...]
Concerning the blipdir, yes with -1 the tip of the frontal lobe looks much better indeed. Was wondering about the Siemens terming >> however, so does A >> P mean something increases from A to P, or something starts at A, ... Anyway ;-)
Best,
Helmut
|