[Insight-users] itkSYN itkBsplineSYN Image Registration- Optimizer Documentation??

[brian avants]

multi-resolution set up is usually the key to such issues.

but greater detail in your question / code is needed.

syn & variants are applicable for a broad range of images

as you can see in the literature & documentation.


brian




On Wed, Jan 22, 2014 at 11:57 AM, Emma Saunders
<emmasaunders123 at gmail.com>wrote:

> Thanks Nick and Brian for your input.
>
> I was attempting to change the learning rate as I am having poor results
> using both syn and itkBsplinesyn when registering two MR images of the
> entire thorax during respiration.  Are these methods specifically designed
> for certain types of images (e.g brain).  I initialize my framework with an
> Affine transform, using mattes MI, the metric value first falls off and
> then increases.  The result is thus only a small amount of movement, around
> 1/3 of what I would hope for.
>
> I can successfully register the images with a regular Bspline method using
> ITK, but was hoping to get a SYN method working in order to have a valid
> inverse.
>
> I suppose the question is, if I can get regular Bspline to give a descent
> result should the same hold for the SYN methods or am I missing something?
>
> Thanks for your help
>
>
>
>
> On Wed, Jan 22, 2014 at 3:30 PM, Emma Saunders <emmasaunders123 at gmail.com>wrote:
>
>> Hi all
>>
>> I am trying to use the Syn methods mentioned above.
>>
>> I would simply like to set my framework up to estimate the learning rate
>> at each iteration.  So far this has proved problematic as the optimization
>> process is contained within the methods itself and whenever I try to find a
>> solution need to look through ANTS header files and I still have no luck.
>>
>> Is there a simple way to set the learning rate given a scaleEstimator for
>> these methods (without using the ans registration helper ) or must one
>> simply choose one at the beginning and hope for the best??
>>
>> I am also finding it difficult to determine which optimizer these methods
>> use as in the ANTS headers there are two options a GradientDescent and a
>> conjugate gradient descent, any ideas??
>>
>> I initially used the ITK libraries as they provided a method that I could
>> begin using state of the art image processing algorithms without the steep
>> learning curve, is there any pending documentation or software guide for
>> the new ITKV4 framework.
>>
>> Thanks
>>
>> Emma
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20140122/182e407e/attachment.html>