Hi - I've not looked at this in detail or quantitatively, but my feeling
is that you can get reasonable segmentations if you have 1x1x1mm data
(though there is still lots of GM partial voluming). I don't think that
there are any special options that will help here, and it will probably
not help to remove other parts of the brain. You might try reducing the -b
and --b values (see FAST usage on the command line) to reduce the
smoothing of the segmentation.
You might benefit from looking at any publications that have done
flattening of the cerebellar surface (e.g. Monica Hurdal
http://www.math.fsu.edu/~mhurdal) as they will have looked into this I
guess.
Cheers, Steve.
On Tue, 13 Apr 2004, Graham Wideman wrote:
> Folks:
>
> Does anyone feel they have some success segmenting the cerebellum with
> FAST? If so, we'd appreciate comments on:
>
> -- What resolution and protocol of MRI did you find was needed?
>
> -- Any particular parameters that were especially needed for cerebellum?
>
> -- How did you judge you were getting a "good" segmentation?
>
> -- Can FASTs performance on the cerebellum be improved if we first remove
> all other matter and present just the cerebellum to FAST?
>
> -- Any other lessons learned?
>
> Thanks,
>
> Graham
>
Stephen M. Smith DPhil
Associate Director, FMRIB and Analysis Research Coordinator
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
|