Hi - I think your preprocessing steps are ok. Which output from dtifit are you feeding into
randomise, the FA data?
A t-test model in randomise would be sensible if you were comparing a set of zero-error FA
images with a set of non-zero (but equivalent) error FA images (eg generated from a set of
different but similar "badness" original added rotations. Is this what you are doing?
The simplest output from randomise to look at is the _vox image, ie uncorrected voxelwise p-
values for the t-test.
Cheers, Steve.
On Tue, 23 Jan 2007 11:02:51 +0100, Martin Kavec <[log in to unmask]> wrote:
>In the clinical DTI, I almost always see a subject's motion in the consecutive
>images acquired with different orientations. This may come from subject's
>rigid motion or eddy current and can be effectively compensated for by
>running motion-eddy correction prior to dti_fit and bedpost. I was curious
>however, how the motion affects FA, MD and tensor indices, if it is not
>corrected for and what is the critical level of motion inducing statistically
>significant changes in the above parameters.
>
>Here is what I briefly did:
>
>1. I took a single DTI subject data and corrected for the motion. here I
>assume that residual root-mean-square (RMS) is 0
>
>2. in the result of 1. I induced artificial motion by applying a random affine
>transform to each volume in the DTI series. Each random transform had RMS
>from small interval. I generated several series with the same and increasing
>amount of motion
>
>3. then I ran dti_fit on all the data (1. and 2.)
>
>4. merge results of dti_fit from 2 and 1 to the single 4D and ran randomise to
>get paired t-test
>
>
>Here I have a small problem as I am not completely sure, which result file of
>the randomise I should look at. This is because I get values above 1 in the
>tstat file and very small values (suggesting no significant difference) in
>the vox and max files even with high level of motion.
>
>Thanks in advance,
>
>Martin
>===========================================================
==============
|