Summer2009:The Vascular Modeling Toolkit in 3D Slicer

• 

  Project Week Main Page

• 

  Example segmentation of cerebral vessels and a tumor using VMTK in 3D Slicer

• 

  Example segmentation of vessels in an arm bearing a vascular access for hemodialysis

Key Investigators

• Daniel Haehn (Student of Medical Informatics, University of Heidelberg)

Objective

The Vascular Modeling Toolkit (VMTK) is a collection of libraries and tools for 3D reconstruction, geometric analysis, mesh generation and surface data analysis for image-based modeling of blood vessels. It should be very interesting to offer such techniques in 3D Slicer.

Approach, Plan

To provide VMTK tools in 3D Slicer the vtkVmtk library has to be available as an add-on. With the connection of VMTK and 3D Slicer processing pipelines between VMTK code and other algorithms can be established.

The plan is to clean-up existing VMTK in 3D Slicer code, provide additional VMTK functionality in 3D Slicer as well as to optimize the work-flow between the existing VMTK modules. This includes more level set evolution methods and variable calculation of feature images during level set segmentation.

Progress

The library of VMTK was made available in 3D Slicer using a GUI-less loadable module approach in order to provide a flexible way of distributing and including it. Therefore all VMTK functions are available to 3D Slicer developers.

Python scripted modules offering VMTK level set segmentation methods and VMTK vessel enhancement filters within 3D Slicer were created. These include the following algorithms:

• Colliding Fronts and Fast Marching Upwind Gradient initialization
• Geodesic Active Contour and Curves evolution
• FrangiVesselness, SatoVesselness and VesselEnhancingDiffusion filters

The created modules are available at the VMTK in 3D Slicer NITRC page: http://www.nitrc.org/projects/slicervmtklvlst/

Summer Project Week Update:

• Projects of the UIOWA concerning VMTK were helped and introduced to the existing Slicer3 modules. Also students from Harvard SEAS were interested in the ability to track a vessel tree using VMTK.
• Fast Marching Upwind Gradient initialization: target points were made optional and the algorithm can run with one or more source points (seeds).
• Level-set initialization can now be performed on an enhanced vesselness image. During evolution stage this initial level set can be inflated within the original volume to avoid the loss of information.
• For using the gui-less VMTK library module as a Slicer3 extension, necessary updates to the CMake files of VMTK were performed to remove deprecated INSTALL_FILES and INSTALL_TARGET calls. These changes make vtkVmtk Slicer3 compliant by using CMake 2.6, were sent to Luca Antiga and hopefully be in the official VMTK release soon :D.

References

• Antiga L, Piccinelli M, Botti L, Ene­Iordache B, Remuzzi A, Steinmann DA. (2008) An image­based modeling framework for patient­specific computational hemodynamics. Med Biol Eng Comput 46(11):1097­1112
• Antiga L, Steinman DA (2008) The Vascular Modeling Toolkit. http://www.vmtk.org/
• Piccinelli M, Veneziani A, Steinman DA, Remuzzi A, Antiga L (2009) A framework for geometric analysis of vascular structures: applications to cerebral aneurysms. IEEE Trans Med Imaging. In press.