afni.nimh.nih.gov

Hi Janina,

Here is an older post and thread on the subject:
"Re: Censoring Resting State Data"

1) The enorm value is an L2 norm (sqrt(sum sq)), while
FD is an L1 norm (sum abs). While that difference
is small, L1 does not make sense to me. A motion
difference of "6 0 0 0 0 0" is a far bigger motion than
"1 1 1 1 1 1", yet they would be the "same" with FD
(same, subject to slight scaling of rotation terms).

2) The enorm measure pre-dates FD, and it seems
more appropriate.

- rick