Hi,
You only need one weighting volume, which contains values
of zero and one (for this case that is all you need). I would
suggest creating the weighting volume as a copy of the input
image - that is, the one with the lesions. You will need to
create this image by hand unless you have some reliable
way of detecting the lesions (which is normally not possible).
The easiest way to create the weighting volume is to make a
mask in FSLView. Do this by loading in your image with
lesions, "create mask" from the File menu and then draw your
lesions by selecting the pen tool and drawing them on. Once
you've done that, save it as a mask. Note that this is *not*
the weighting volume as it has the opposite values (zeros
outside the lesions and ones in the lesions). So you then
need to swap the values, which you can easily do with
fslmaths. The appropriate command is:
fslmaths mylesionmask -mul -1 -add 1 -thr 0.5 -bin mylesionweights
Then use the image "mylesionweights" in Flirt (as the
input image weighting). You do not need to do anything
with reference weighting and just use the usual reference
image template.
Hope this helps.
All the best,
Mark
On 23 Apr 2008, at 02:04, Fabienne Cazalis wrote:
> Dear FSL experts,
>
> I want to register brains with lesions to a template. I understand
> that not
> doing anything incurs in poor registration, and that using lesion-
> masking is
> not efficient.
>
> FSL website suggests using the "Weighting Volumes" option in FLIRT
> (instuctions pasted below). I understand that I can feed a model-
> brain in
> which each voxel is attributed a value going from zero (=non brain)
> to (to
> how much? 1? more?), and then I can also feed a home-made lesioned-
> brain
> with similarly valued-voxels, in which the lesion will be Zero-ed, and
> therefore successfully ignored by FLIRT.
>
> That's where my understanding stops, as I have no idea on how am I
> to create
> such images (the weighting image and the image that is weighted).
>
> Can you help?
> thank you very much!
> fabienne cazalis
>
>
> [info from fsl website]
>> *Weighting Volumes* - impose voxel-wise weighting to reference and/or
>> input images, to affects the cost function. The weighting images
>> must be
>> the same size as the image they are weighting (e.g. refweight and
>> reference
>> images) and the voxel values of the weighting image represent how
>> much
>> weighting that particular voxel is given in the cost function.
>> Therefore,
>> by setting weights to zero, some areas of the image can be
>> effectively
>> ignored, which is useful in masking out pathologies so that they do
>> not
>> affect the registration. In this way very accurate registrations
>> can be
>> made between pathological and ``normal'' images. This cannot be
>> achieved by
>> masking the images prior to registration, as that induces artificial
>> boundaries which bias the registration. Furthermore, some areas can
>> be
>> given extra weighting (such as the ventricles) so that the
>> registration is
>> most accurate near these structures, but still uses information
>> from the
>> rest of the image (e.g. the cortical surface) to improve the
>> robustness of
>> the registration.
>
|