In replying to Bob Cox's comment, I am just wondering if it's possible to use 3drotate with the -1Dfile option to read in motion parameters from the -1Dfile output from 3dvolreg and apply them to same dataset.

To be more precise. What I mean is that:

Step 1: use 3dvolreg on DATA+orig to output motion parameters (dfile named motion.1D)

Step 2: use 3drotate with -1Dfile option to read the motion parameters, motion.1D (say, the output named rotated_DATA+orig)

Step 3: use 3dvolreg on rotated_DATA+orig (dfile named rotated_motion.1D)

I checked the rotated_motion.1D file. Seems to me that all the motion has been corrected. I tried this method with a number of datasets which some of them have very big motion.

My questions are:
I am told that if there is too much motion, something more than 1.5-2mm for translation and more than 2 degrees for rotation, then motion correction may not work for this dataset and we may have to discard some of the timepoints of this dataset in order to obtain reasonable results.

First, If the above method (step 1 to step 3) that is acceptable, I guess the test subject can move as much as he/she can in the MRI scanner provided that his brain is always within the Field of View of the scanner?

Second, the above method (step 1 to step 3) is a no-no, why?

Thank you very much!

Ricky