[Insight-developers] Feedback on possible major contribution on vascular extraction and analysis

Tue Jun 22 06:20:16 EDT 2010

Hi Karthik,

You mean Stephen Aylward? This seems interesting. Curiously I am currently
implementing his own algorithm in a generic manner, which seems to have
started to work J

http://www.kitware.com/publications/item/view/1121 (the paper)

Not perfectly yet, but I was using a fixed scale here

http://picasaweb.google.es/imacoli/Research#5485538955694925250

http://picasaweb.google.es/imacoli/Research#5485538969243056402

Please, let me know how we can collaborate in order not to duplicate
efforts. On a first look the purpose and scope seem to be the same. I have
to have a deeper look though.

Best regards

Iván

De: Karthik Krishnan [mailto:karthik.krishnan at kitware.com] 
Enviado el: viernes, 18 de junio de 2010 13:32
Para: Ivan Macia
CC: Insight Developers
Asunto: Re: [Insight-developers] Feedback on possible major contribution on
vascular extraction and analysis

Ivan:

This is wonderful. 

I'll let Stephen give you the run down; you should definitely take a look at
the ongoing effort TubeTk. The wiki page should get you started.  

    http://public.kitware.com/Wiki/TubeTK

The toolkit is based heavily on ITK and provides interfaces to Slicer and is
licensed under a BSD like (apache) license.

--
karthik

On Fri, Jun 18, 2010 at 3:53 PM, Ivan Macia <imacia at vicomtech.org> wrote:

Dear all,

Recently, I have been given the opportunity of taking again my Phd Thesis
work (I am currently full-time R&D engineer) that will focus on vascular
analysis. I am investigating the current and future techniques on vascular
extraction, analysis and modelling from medical images and their
applications. As a part of it, I am developing a module/library that hope
will help me in my research that will consist of the following:

- An open model for vascular structures via generic programming. This
includes a network, graph-like structure that can be traversed, an open
model for branches, centerlines and sections and corresponding operations
such as quantification of vessel attributes, etc.
- A generic algorithm for vessel tracking, with interchangeable components
such as metrics (vesselness or medialness functions), section estimators,
etc.
- Many more things in the future... :)

Philosophy:
- The module is (or tries to be) fully ITK compatible, is based on generic
programming techniques etc. in order to provide flexibility in the vascular
model definition and interchangeability of components. This allows to
replace a component such as the vesselness metric without rewriting the
whole algorithm, for example in order to benchmark different techniques.
- It makes use of previous exisiting components as much as possible, it
defines DataObjects and Filters (VesselDataObject,
ImageToVesselDataObjectFilter...) and reuses algorithms such as current
vesselness algorithms in ITK, ITK optimizers etc.
- It does not pretend to overlap other toolkits such as VMTK, which I think
is more oriented to geometric modelling, more than symbolic modelling.
Furthermore, it could be compatible with it.
- It is fully ITK oriented and focuses on the extraction and modelling of
vascular structures from medical images. 
- It may serve as a model that can later be used in VMTK, VTK for
visualization, Slicer, whatever.

The work is ongoing, the initial vascular model is defined and the vascular
tracking has just started to work with promising results. Further
functionalities and components will be added in the near future. There exist
about 50 classes right now, although some of them are not fully functional.

I don't want to be too pretentious and I would like my work to be used as
much as possible and this is why I would like to share it with the community
in the near future. I still need to develop several things and experiment
with them (and hopefully publish something interesting for the Thesis work),
but thought that may be useful to comment it in this forum, so others can
give their opinion and even add their own requirements or design
considerations at an early stage.

Regarding ITK I see the following possibilities:
- If the ITK committee aproves it it may incorporated (at least partly) in
the future as part of the toolkit itself following for example this process
(http://www.itk.org/Wiki/ITK_Oversight_Proposal_Process)
- It could be a separate toolkit such as Orfeo Toolbox.

Please, feel free to comment or discuss with me any aspect of this in public
or private.

Thanks in advance

Iván Macía 

_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html

Kitware offers ITK Training Courses, for more information visit:
http://kitware.com/products/protraining.html

Please keep messages on-topic and check the ITK FAQ at:
http://www.itk.org/Wiki/ITK_FAQ

Follow this link to subscribe/unsubscribe:
http://www.itk.org/mailman/listinfo/insight-developers

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20100622/47a9997e/attachment.htm>