Hi Kate,
On 1 Aug 2007, at 19:38, Kate Fissell wrote:
> Hi,
>
> I am trying to use the bias field output by fast as a weight
> volume in flirt ( -inweight <volume> ). Could you explain
> how flirt uses the -inweight volume values as weights ? I see the
> FAQ says
> "To do this a weighting image must be made which has the value of 1.0
> everywhere except in the region of interest where a higher weight
> (e.g.
> 10.0) should be used."
> Does the weight volume have to be a volume with just 2 intensity
> values, 1
> and N (eg 10), or is that just an example ?
It can be any set of values. The weighting is simply a weighting
applied to the cost function - so those voxels with higher values
contribute more strongly to the registration cost function, those
with 0 contribute not at all. The weighting can be applied to the
input image or the reference image or both.
> In looking at the fast bias volume I think it is 1 at "uncorrected"
> voxels
> and then has bias values both above and below 1. I would think
> that to
> transform the bias volume into a weight volume, I might want to
> downweight voxels proportionate to their distance from 1, eg for
> bias volume B
> weight volume = 10 / (1 + abs(B-1))
> but I am not sure. That would make 10 the "no correection" value.
> Also, does it matter what weight I assign in the
> weight volume to voxels that are non-brain and set to zero in the
> flirt input volume ?
The above equation looks reasonable, particularly if it works!
However yes there's a problem for voxels outside the brain - in
general if you give them 0 weighting then they won't get used at all,
which means in effect that the brain/non-brain boundary _doesn't_ get
used in the registration, which is normally a bad thing. Hence you
might like to make sure that the bias field outside the brain gets
set to 1 (I think it might already anyway).
Cheers.
>
> thanks
> kate
------------------------------------------------------------------------
---
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
------------------------------------------------------------------------
---
|