Dear Richard,
>I had a question about using the motion correction parameters as covariates
>in specifying models. If my understanding is correct, the parameters
>express relative spatial position and orientation in mm or radians from the
>reference scan.
>
>
I don't think this is entirely true. The parameters in the extra
regressors are the movement correction parameters for the *volume*, so
the same set of translations and rotations for the whole volume. The
actual displacements are different for voxels that are further away from
the centre of rotation, than those that are closer.
>If this is true, to correct for motion-related artifacts, wouldn't it be
>more appropriate to use the difference (delta) in each parameter from one
>image to the next, as this would seem to be a better estimate of motion
>(i.e. change in spatial position per unit time = velocity) at the the time
>of each image?
>
>
I don't think it is the speed of the movement that increases the effect
of artefacts, it really is the displacement. The artefact is due to the
spin excitation history of the voxel, i.e. energy transmitted to a brain
region before thatt region was at its current voxel location (maybe you
should ask an MR physicist instead of reading this...).
It is true though, that (in the case of rotation) the displacement
caused by a single parameter is not uniform across the brain. If you
shake your head, the displacement close to your neck is almost zero, and
the displacement is larger further away from your neck.
The BAMM software (http://www-bmu.psychiatry.cam.ac.uk/software) uses
the displacement in each voxel, rather than the movement parameters, to
remove the spin excitation history artifact. For a good description of
the method see
http://www-bmu.psychiatry.cam.ac.uk/sitewide/publications/journal/bullmore99met.pdf.
Note that because the displacement is different in each voxel, the
effects cannot be removed by putting regressors in the design matrix. In
BAMM, the artifacts are removed before the model is estimated.
>Likewise, next taking the scan-to-scan differences of those differences
>would estimate acceleration (i.e. change in velocity per unit time =
>acceleration).
>
>
I don't know how a fast movement (between two acquisitions) causes a
different artifact than a slow movement. Maybe it can be greater for
slow movements. I think solving this is very difficult...
With best wishes,
Alle Meije Wink
___________________________________________________________
To help you stay safe and secure online, we've developed the all new Yahoo! Security Centre. http://uk.security.yahoo.com
|