Hi,
You're right that it probably doesn't make much sense to do a group
analysis on this dataset. I would have thought that you don't
necessarily even want to use the skeletonised data, as for a first
pass analysis at least, I would think you would want to visualise
where any change is happening to get a feel for what's in the data.
Hence I think I would recommend something like:
- take just one subject's timepoints data and run tbss_1_preproc
- imcp the first timepoint to "target" in FAi and run tbss_2_reg
and tbss_3_postreg and tbss_4_prestats
- view stats/all_FA as a movie loop to check alignments
- do timeseries stats on all_FA (and possibly also
all_FA_skeletonised if you wish)
But - how to do the modelling; I would put linear (temporal) trend
and maybe qudratic depending on how many timepoints you have. Of
course, this may not be as sensitive to change as using DTI-
bootstrapping to get measures of uncertainty by randomising across
the different directions data - but on the latter you would need to
implement your own way of doing this (see papers by Jones, etc on
bootstrapping the different directions data etc).
Cheers, Steve.
On 5 Jun 2007, at 01:39, Stephen Rose wrote:
> Hi
>
> I was wondering whether anyone has any advice about the possibility
> of analysing serial stroke DTI
> data using TBSS and randomise. We are interested in measuring the
> changes in FA over time
> during recovery of stroke. As this is a very heterogenous group
> (e.g. stroke lesions occur in
> different anatomical locations, size etc), simple grouping of data
> would be of little use. Is it
> possible to analyse changes in FA over a number of time points for
> individual subjects using
> randomise? If so any advice regarding generation of design and
> contrast files would be much
> appreciated. Can individual subjects be assigned as a “different
> group” when setting up the design
> files etc?
------------------------------------------------------------------------
---
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
------------------------------------------------------------------------
---
|