Difference between revisions of "Slicer3:Module:Level-Set Label Map Evolver-Documentation"

From Slicer Wiki
Jump to: navigation, search
 
(20 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
[[Documentation|Return to Slicer Documentation]]
 
[[Documentation|Return to Slicer Documentation]]
__NOTOC__
 
 
===Module Name===
 
===Module Name===
 
Level-Set Label Map Evolver
 
Level-Set Label Map Evolver
 +
{|
 +
|-
 +
|__TOC__
 +
|
 +
|
 +
|
 +
|[[image:Evolver1.jpg|600 px]]
 +
|}
  
 
== General Information ==
 
== General Information ==
Line 9: Line 16:
 
Type: CLI (Interactive to-be)
 
Type: CLI (Interactive to-be)
  
Category: Segmentation
+
Category: Level-Set Segmentation
 +
 
 
===Authors, Collaborators & Contact===
 
===Authors, Collaborators & Contact===
 
* Carlos S. Mendoza: Universidad de Sevilla
 
* Carlos S. Mendoza: Universidad de Sevilla
Line 15: Line 23:
  
 
===Module Description===
 
===Module Description===
This module is an under development version of an interactive level-set evolution module based on [http://www.itksnap.org ITK-SNAP] segmentation approach.  
+
This module is capable of evolving a label image according to a level-set customizable equation for a given number of iterations according to the preprocessed version of a given input image, thus segmenting that image. Ultimately it will be migrated to an interactive approach in which the level-set evolution can be controlled as it is being displayed, running step by step.
 +
 
 +
The level-set equation incorporates terms for propagation, curvature, and advection.
 +
 
 +
== Usage ==
 +
 
 +
Once the [[Slicer3:Module:Region_Selector-Documentation|Region Selector]], the [[Slicer3:Module:Target_Preprocessing-Documentation|Target Preprocessing]] and the [[Slicer3:Module:Bubble_Maker-Documentation|Bubble Maker]] modules have been run we can feed the bubble image and preprocessed volume to this module.
 +
 
 +
{|
 +
|-
 +
|[[image:evolver3.jpg|300 px]]
 +
| The parameters have the following meaning:
 +
 
 +
*Initial Label Image: The output of the [[Slicer3:Module:Bubble_Maker-Documentation|Bubble Maker]] module.
 +
 
 +
*Feature Image: The output of the [[Slicer3:Module:Target_Preprocessing-Documentation|Target Preprocessing]] module.
 +
 
 +
*Propagation Weight: Relates to the amount of inflating force applied inside the surface (can be negative).
 +
 
 +
*Curvature Weight: Influences roundness of the resulting interface. Greater weight, greater roundness. Prevents leaking through narrow openings.
 +
 
 +
*Advection Weight: It weights the influence of the advection term, which behaves like a magnet attracting the interface towards edges.
 +
 
 +
*Number of Iterations: Indicates duration of the evolution.
 +
 
 +
*Output Label Image: Resulting segmentation.  
 +
 
 +
 
  
It will be capable of evolving a label image according to a level-set customizable equation for a given number of iterations according to the preprocessed version of a given input image, thus segmenting that image. Ultimately it will be migrated to an interactive approach in which the level-set evolution can be controlled as it is being displayed, running step by step.
 
  
The level-set equation will incorporate terms for propagation, curvature, advection and shape priors.
+
|}
 +
 
 +
The usage of this module implies a certain amount of trial and error, tuning parameters according to the incorrect segmentations obtained in the first runs. If you get leaking adjust curvature weight and propagation force or run for less iterations, according to what is happening in order regions(under-growth). A couple trials will provide the user with a feeling of the influence of the different parameters.
 +
 
 +
For further details on parameter interpretation and mathematical formalities address to Caselles et al. paper.
  
For the initial label image to be created, another CLI module is under development. The "Bubble Maker" module will take as input a set of fiducials and radii and create a label image made of bubbles. Those bubbles should fit inside our structure of interest in order for the resulting label image to work as an initial stage in the level-set segmentation.
 
== Usage ==
 
TODO.
 
===Examples, Use Cases & Tutorials===
 
TODO.
 
===Quick Tour of Features and Use===
 
TODO.
 
== Development ==
 
TODO.
 
===Dependencies===
 
TODO.
 
 
===Known bugs===
 
===Known bugs===
 
Follow this [http://na-mic.org/Mantis/main_page.php link] to the Slicer3 bug tracker.  
 
Follow this [http://na-mic.org/Mantis/main_page.php link] to the Slicer3 bug tracker.  
Line 37: Line 64:
 
Follow this [http://na-mic.org/Mantis/main_page.php link] to the Slicer3 bug tracker. Please select the '''usability issue category''' when browsing or contributing.
 
Follow this [http://na-mic.org/Mantis/main_page.php link] to the Slicer3 bug tracker. Please select the '''usability issue category''' when browsing or contributing.
 
===Source code & documentation===
 
===Source code & documentation===
TODO.
+
Customize following [http://www.na-mic.org/ViewVC/index.cgi/ links] for your module.
== More Information ==
+
 
TODO.
+
[http://www.na-mic.org/Slicer/Documentation/Slicer3/html/ Links] to documentation generated by doxygen.
 +
 
 
===Acknowledgment===
 
===Acknowledgment===
TODO.
+
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===
 
===References===
TODO.
+
V. Caselles, R. Kimmel, and G. Sapiro. Geodesic active contours. International Journal
 
+
on Computer Vision, 22(1):61–97, 1997.
 
 
This module is under development.
 

Latest revision as of 20:00, 11 August 2008

Home < Slicer3:Module:Level-Set Label Map Evolver-Documentation

Return to Slicer Documentation

Module Name

Level-Set Label Map Evolver

Evolver1.jpg

General Information

Module Type & Category

Type: CLI (Interactive to-be)

Category: Level-Set Segmentation

Authors, Collaborators & Contact

  • Carlos S. Mendoza: Universidad de Sevilla
  • Contact: carlos.sanchez.mendoza@gmail.com

Module Description

This module is capable of evolving a label image according to a level-set customizable equation for a given number of iterations according to the preprocessed version of a given input image, thus segmenting that image. Ultimately it will be migrated to an interactive approach in which the level-set evolution can be controlled as it is being displayed, running step by step.

The level-set equation incorporates terms for propagation, curvature, and advection.

Usage

Once the Region Selector, the Target Preprocessing and the Bubble Maker modules have been run we can feed the bubble image and preprocessed volume to this module.

Evolver3.jpg The parameters have the following meaning:
  • Propagation Weight: Relates to the amount of inflating force applied inside the surface (can be negative).
  • Curvature Weight: Influences roundness of the resulting interface. Greater weight, greater roundness. Prevents leaking through narrow openings.
  • Advection Weight: It weights the influence of the advection term, which behaves like a magnet attracting the interface towards edges.
  • Number of Iterations: Indicates duration of the evolution.
  • Output Label Image: Resulting segmentation.



The usage of this module implies a certain amount of trial and error, tuning parameters according to the incorrect segmentations obtained in the first runs. If you get leaking adjust curvature weight and propagation force or run for less iterations, according to what is happening in order regions(under-growth). A couple trials will provide the user with a feeling of the influence of the different parameters.

For further details on parameter interpretation and mathematical formalities address to Caselles et al. paper.

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.