Hi guys,
I have a question regarding volume registration.
We are interested to analyze fMRI data with a new tecnique that is a sort of nonparametric spectral estimation. We have noticed that the volume correction step (3dvolreg -maxite 35 -dfile mov.out -prefix final epi+orig) lead to different results in spectral estimation as compared to the raw data in which the max displacement is of 0.7 mm (so we think it is pretty small). Our fMRI data consists of 4 slices acquired with a TR of 300 ms for ten minutes and the voxel size is 3.75x3.75x5mm . These differences could be related to the interpolation schemes, or to the volume correction per se? In your opinion, does make sense to avoid motion correction if the displacement is "small"?
thank you,
claudio