Hi Hema,

It is likely the motion censoring that is doing
you in, not the outliers.

How much motion censoring to do is a bit subject
group specific. Basically, try to censor as much
as possible, while keeping most of your patients
(or whatever group of subjects you have that makes
censoring an issue). Once you get a feel for a
subject group, it is good to be consistent.

For children and other high movers, it is common
to set a motion limit of 1.0, and sometimes higher
(thought higher disturbs me :). Maybe 0.5 is high
enough for your subjects.

You can get a feel for censoring effects by running
1d_tool.py -quick_censor_count on the motion files
(e.g. dfile_rall.1D, from each subject), e.g.

1d_tool.py -quick_censor_count 0.5 -infile dfile_rall.1D \
             -set_run_lengths 150 150 150

or even with a find command:

find subject_results -name dfile_rall.1D -exec   \
     1d_tool.py -quick_censor_count 0.5 -infile {} \
             -set_run_lengths 150 150 150 \;

- rick