Biomedical Engineering Reference
In-Depth Information
the blood pool and sometimes the outside tissues. Few automated segmentation
algorithms, with minimal manual intervention, are available on clinical consoles
to assist the segmentation task with a significant saving of time. Segmentation
of cardiac images is still a very active research area and level set segmenta-
tion methods have proved in the recent years to offer a very flexible three-
dimensional tool that can handle the volumetric and dynamic nature of the
data.
2.4.2
Open Source Software Tools for Level
Set Segmentation
2.4.2.1
Snake Automatic Partitioning (SNAP)
This software was developed by the Medical Image Display and Analysis Group
at the University of North Carolina and is available for download at www.
midag.cs.unc.edu. SNAP is a segmentation tool for volumetric image data using
3D level set methods with either a region-probability deformable model or a
gradient-based deformable model framework. Some interaction with parameter
settings of the segmentation method and prior-filtering is available. Interactive
visualization of the deformation process in provided.
2.4.2.2
Insight Segmentation and Registration Toolkit (ITK)
The National Library of Medicine Insight Segmentation and Registration Toolkit
(ITK) is an open-source software system to support the Visible Human Project.
The toolkit is available for free download at www.itk.org. Under active devel-
opment, ITK employs leading-edge segmentation and registration algorithms
in multiple dimensions. The Insight Toolkit was developed by six principal or-
ganizations, three commercial ( Kitware, GE Corporate R&D, and Insightful)
and three academic ( UNC Chapel Hill, University of Utah, and University of
Pennsylvania). Additional team members include Harvard Brigham & Women's
Hospital, University of Pittsburgh, and Columbia University. The funding for the
project is from the National Library of Medicine at the National Institutes of
Health. NLM in turn was supported by member institutions of NIH (see spon-
sors). Several level set segmentation methods are implemented in this toolkit
including: fast marching methods, shape detection segmentation, geodesic
