Hi,
Yes, the absolute values after the smoothing stage in siena_flow2std
will be changed because the smoothing extends into the background.
It's the smoothing and not the dilation that has caused this.
You have two options to get back to the original scaling:
- apply the same smoothing to the standard space edge image that is
in the siena_flow2std script, and divide your A_to_B_flow_to_std (and/
or the 4D file after concatenating across subjects) by this to get
the normalisation right (make sure you output the smoothed mask image
as floating point).
or
- if you rerun using the updated siena_flow2std in FSL4.0 it should
now do that normalisation for you correctly so all should then be the
"right" scaling.
Cheers, Steve.
On 23 Aug 2007, at 18:18, Kempton, Matthew wrote:
> Dear SIENA users,
>
>
>
> We have a paper under review which uses a SIENA voxelwise analysis.
> We have been asked by a reviewer to report the peak boundary shift
> of a significant change we detected in part of the lateral ventricles.
>
>
>
> I had a look at the A_to_B_flow_to_std.img images for each subject
> at the co-ordinates in question. I hoped to get the displacement of
> brain boundary in mm from the voxel intensity at that point.
> However the mean displacement over all subjects seemed too small
> (about a 50th of a 1mm). I went back to the individual
> A_to_B_flow.img (flow image in native space) and the flow values
> were larger and more reasonable.
>
>
>
> I was wondering if values in the flow_to_std image are usually
> lower than the corresponding flow images. As the data we are
> working only shows very small displacement, to test this further we
> looked at 3 subjects who had been scanned 2 years apart where there
> was evidence of significant brain atrophy. For each subject we
> recorded voxel values at 10 'randomly' chosen points in the lateral
> ventricle brain/CSF boundary both in the A_to_B_flow_to_std.img and
> A_to_B_flow.img file. In the 3 subjects the voxel intensity in the
> A_to_B_flow.img seemed to be about right (1.5mm) while the voxel
> intensities in the A_to_B_flow_to_std.img file seemed about 10
> times too small (0.15mm).
>
>
>
> Has anybody else seen this and know why this happens? and is this
> deliberate or is it a 'side effect' of the dilation and smoothing
> steps… or have I just done something wrong?
>
>
>
> Best wishes,
>
>
>
> Matthew Kempton
>
> Institute of Psychiatry
>
> London
>
>
>
>
------------------------------------------------------------------------
---
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
------------------------------------------------------------------------
---
|