Difference between revisions of "Slicer3:Module:Level-Set Segmentation Framework-Documentation"
| (15 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
[[Documentation|Return to Slicer Documentation]] | [[Documentation|Return to Slicer Documentation]] | ||
| − | + | ||
===Level-Set Segmentation Framework=== | ===Level-Set Segmentation Framework=== | ||
| − | + | {| | |
| − | [[Image:capture1.png|thumb|280px| | + | |+ |
| − | [[Image:capture2.png|thumb|280px| | + | | __TOC__ |
| − | + | |[[Image:capture1.png|thumb|280px|Bubble seeding]] | |
| + | |[[Image:capture2.png|thumb|280px|Resulting segmentation]] | ||
| + | |} | ||
== General Information == | == General Information == | ||
| Line 20: | Line 22: | ||
The modules in the framework support different tasks in the segmentation realization: | The modules in the framework support different tasks in the segmentation realization: | ||
| − | -For a level-set segmentation to be performed a seed image must be provided. A module called | + | -For a level-set segmentation to be performed a seed image must be provided. A module called [[Slicer3:Module:Bubble_Maker-Documentation|Bubble Maker]] was developed, that takes fiducials and a vector a integers as input, and produces as output a label image composed of bubbles centered in the fiducials with radii such as the integers provided. |
| − | -For a level-set segmentation to be performed a feature image has to be extracted from the target image. That image is a smoothed, diffrentiated and remapped version of the original image. A tunable module called | + | -For a level-set segmentation to be performed a feature image has to be extracted from the target image. That image is a smoothed, diffrentiated and remapped version of the original image. A tunable module called [[Slicer3:Module:Target_Preprocessing-Documentation|Target Preprocessing]] was developed. |
| − | -In order for the preprocessing to take place in pseudo-interactiva time frames a region of interest must be provided by the user, preventing from unuseful calculations. A module called | + | -In order for the preprocessing to take place in pseudo-interactiva time frames a region of interest must be provided by the user, preventing from unuseful calculations. A module called [[Slicer3:Module:Region_Selector-Documentation|Region Selector]] was developed, that takes as input three rectangles, one in each coordinate plane to define a parallelepiped. |
| − | Finally a module called | + | Finally a module called [[Slicer3:Module:Level-Set_Label_Map_Evolver-Documentation|Level-set label map evolver]] was developed, which takes an initial label image and a feature image as input and performs a Geodesic Active Contours evolution on the label image according to the feature image and to a different terms in the level-set equation. The evolution takes place for a customizable number of iterations. |
The output is a label image that can be used to produce a model. | The output is a label image that can be used to produce a model. | ||
Latest revision as of 21:08, 11 August 2008
Home < Slicer3:Module:Level-Set Segmentation Framework-DocumentationReturn to Slicer Documentation
Level-Set Segmentation Framework
General Information
This framework corresponds with a new category of Slicer modules for Level-Set segmentation using Geodesic Active Countours (Caselles et al.)
This framework is useful when dealing with segmentations of regions that present edges with respect to the background.
Authors, Collaborators & Contact
- Author: Carlos S. Mendoza
- Contact: carlos.sanchez.mendoza@gmail.com
Module Description
The modules in the framework support different tasks in the segmentation realization:
-For a level-set segmentation to be performed a seed image must be provided. A module called Bubble Maker was developed, that takes fiducials and a vector a integers as input, and produces as output a label image composed of bubbles centered in the fiducials with radii such as the integers provided.
-For a level-set segmentation to be performed a feature image has to be extracted from the target image. That image is a smoothed, diffrentiated and remapped version of the original image. A tunable module called Target Preprocessing was developed.
-In order for the preprocessing to take place in pseudo-interactiva time frames a region of interest must be provided by the user, preventing from unuseful calculations. A module called Region Selector was developed, that takes as input three rectangles, one in each coordinate plane to define a parallelepiped.
Finally a module called Level-set label map evolver was developed, which takes an initial label image and a feature image as input and performs a Geodesic Active Contours evolution on the label image according to the feature image and to a different terms in the level-set equation. The evolution takes place for a customizable number of iterations.
The output is a label image that can be used to produce a model.
Usage
See modules documentation.
Known bugs
Follow this link to the Slicer3 bug tracker.
Usability issues
Follow this link to the Slicer3 bug tracker. Please select the usability issue category when browsing or contributing.
Source code & documentation
Customize following links for your module.
Links to documentation generated by doxygen.
Acknowledgment
This work was developed on financial support from the University of Sevilla, Spain. Most of the development took place in the Surgical Planning Laboratory, Harvard Medical School and Brigham and Women's Hospital, under the supervision of Mr. Steve Pieper Ph.D.
References
V. Caselles, R. Kimmel, and G. Sapiro. Geodesic active contours. International Journal on Computer Vision, 22(1):61–97, 1997.
