Actually, what I mean can be mentioned as a transformation model (e.g.
an elastic body biomechanical morph model) which is primarily fed by
constraints (=skeleton-projection morphs + 2D white-gray boundary
transformation field as is done in freesurfer). Surely these
constraints don't cover the whole 3D white matter field, but, by means
of the elastic biomechanical model (or any other appropriate model
type?) the transformation field can be diffused and extrapolated in
the rest of the image.
I have just found an article, "Improve tractography alignment using
combined volumetric and surface registration" which is close to my
idea, meaning that the 2D surface displacement field is diffused in
the whole 3D space.
But, in this article no application of valuable DW data is implemented
for inter-subject alignment of images (only structural data are used).
Meaning that a tract's starting and ending point could be aligned
perfectly with this purely structural tarnsformation, but this
approach doesn't address remnant intersubject variations in tract
shapes and paths which I believe can be overcome by implementing the
TBSS-derieved constraints (=skeleton projection vectors) in the model.
I would like to know your expert comment on such a combined
structural-diffusion alignment methodology.
Sincerely yours,
Sourena
On 9/9/11, Stephen Smith <[log in to unmask]> wrote:
> Hi - I'm not quite sure what you're asking. The
> FNIRT-warp+skeleton-projection is only valid for the skeleton, and doesn't
> make sense to try to interpolate that out into the rest of the image…..?
> Cheers.
>
>
>
> On 8 Sep 2011, at 19:02, Sourena Soheili wrote:
>
>> Thanks professor Smith regarding your response,
>>
>> "I would like to combine TBSS transformation vectors and white-gray matter
>> boundary normalization warps among subjects to produce a transformation
>> field covering the entire white-matter diffusion data (interpolating
>> transformation in regions not belonging to the skeleton or boundary).
>> The initial FNIRT warp covers the whole image and not just the skeleton -
>> so you could use this (see warp files in the FA directory and the TBSS
>> scripts for more info) - however the final skeleton projection step
>> applies only to the skeleton, not all of the image."
>> But the transformation data which I would like to know if is achievable as
>> raw data is the skeleton projection vectors, not the FNIRT warp. Is it
>> possible?
>>
>> Thanks,
>> Sourena
>>
>> On Tue, Aug 23, 2011 at 1:29 PM, Stephen Smith <[log in to unmask]>
>> wrote:
>> Hi
>>
>> On 20 Aug 2011, at 08:55, Sourena Soheilinezhad wrote:
>>
>>> Hi,
>>>
>>> Is it possible to export TBSS calculated warps with the current FSL
>>> version?
>>>
>>> I would like to combine TBSS transformation vectors and white-gray matter
>>> boundary normalization warps among subjects to produce a transformation
>>> field covering the entire white-matter diffusion data (interpolating
>>> transformation in regions not belonging to the skeleton or boundary).
>>
>> The initial FNIRT warp covers the whole image and not just the skeleton -
>> so you could use this (see warp files in the FA directory and the TBSS
>> scripts for more info) - however the final skeleton projection step
>> applies only to the skeleton, not all of the image.
>>
>>> Also please give me a comment if carrying out direct analysis on the
>>> diffusion vectors (e.g. Watson-Williams spherical statistic test on V1)
>>> could yield further benefit than FA, since this test calculates
>>> significant directional difference among vector groups.
>>
>> Possibly - though this needs a lot of thought wrt what the most sensible
>> transformation (particularly rotation obviously) should be applied before
>> comparing this across subjects. One might argue that if the registration
>> is done well and rotations applied appropriately, there should be nothing
>> left to vary across subjects!
>>
>> Cheers.
>>
>>
>>
>>
>>>
>>> Thanks in advance,
>>> Sourena
>>>
>>
>>
>> ---------------------------------------------------------------------------
>> 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
>> ---------------------------------------------------------------------------
>>
>>
>>
>>
>
>
> ---------------------------------------------------------------------------
> 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
> ---------------------------------------------------------------------------
>
>
>
>
|