Hi,

I have 2 questions regarding motion correction.

1. The motion correction plot of one of the subjects shows many spikes. I want to improve the the quality of the motion correction and I was thinking to replace the timepoints where there was a spike motion with the mean of the adjacent timepoints images. (e.g spike at timepoint 15 ==> replace the timepoint with the mean of timepoint 14 and 16)
I know how to do it in SPM, but the only problem is that when I convert the .img files back into AFNI format the datum type for that timepoint (15)ranges from 0 to 32767 (glmmax) ,while the rest of the timepoints from 0 to ~2000 . So that timepoint is an outlier in my brik.

The question is that how can we implement this procedure using only AFNI commands?

2. The second question is about how to choose the base brick in motion correction. In my experiment I have 7 runs and I was not able to find a unique subbrick to realigned all timepoints to it. I wonder if is it correct to realign each run to itself in the first step and then in the second step to find a common subbrik and realign all the realigned runs to that subbrik.

Thank you.
Regards,
Teodora