Hi Antonio.
I'm afraid there is no hard rule for this as it depends on so many
factors. You are right in suggesting that fast motions are worse than slow
motions of the same magnitude (for reasons of "spin history" etc).
One (stupid) rule is to try to judge whether the activation looks like
motion artefact! A better qualitative way of assessing motion (apart from
the motion plots) is to run ICA (MELODIC) on the raw data to see how many
motion-related artefacts you get.
Finally, the new version of FSL (hopefully out as a beta release in two or
three weeks now) will contain slightly more informative motion plots - for
example see
http://www.fmrib.ox.ac.uk/~steve/ftp/fmri_4.2+.feat/mc/index.html
the final plots show motion summarised over all voxels, plotted "relative"
(each time point compared with the previous) and "absolute" (each time
point compared with the middle time point). If the relative curve shows
lots of motion, you're in for trouble.
Sorry not to be more quantitative! ttfn, Steve.
On Tue, 27 Aug 2002, Antonio Algaze wrote:
> Hi,
>
> Is there -from your vast experience- some rule of thumb to determine if a dataset is
> acceptable based on the MCFLIRT plots?
>
> For example, is one to consider:
> - absolute differences (e.g. Absolute Max - Absolute Min) throughout the entire run,
> and/or
> - the "abruptness" of the corrections (e.g. the time derivative of the plot) ?
>
> Any suggestions for sensible cutoff values related to these or other criteria, both for
> translations and rotations?
>
> Thanks for your kind help.
> Regards,
> Antonio
>
Stephen M. Smith
Head of Image Analysis, FMRIB
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|