View source for ITK Release 4/A2D2 Projects/Registration Parameter Tuning

= Motivation =

#ITK registration pipelines are highly parameterized.
#Registration performance highly dependent on the selected parameter values.
#Data specific magic numbers obtained empirically.

Look at your code and see how often the following magic numbers are used: {1…10}, {0,…10}, {5,10,15…}, {<math>10^n</math>}, {<math>2^n</math>}, prime numbers, 42. Most likely, these are not optimal values.

= Goal =

The goal of this project to develop a framework for automated parameter tuning of ITK non-rigid
registration pipelines.

Formulate this task as an optimization over pipeline parameter values:
#Use the tight coupling between registration and segmentation of image pairs.
#Optimized function based on comparison of the given image segmentation and that induced by the registration.<br>

[[File:optimizationFramework.jpg|800px]]

= Data =

The data used in the project can be found in [http://midas.kitware.com/community/view/33 this] MIDAS repository.

= Team =

* Ziv Yaniv, PI, (Georgetown University)
* Andinet Enquobahrie (Kitware)
* Michael Grauer ( Kitware )
* Ben Fuerst (Georgetown University/TUM)

= Ongoing =

# Analysis of the ITK optimization framework - [[Proposals:Refactoring_of_optimization_framework|refactor ITK hierarchy]] to distinguish between functions that provide derivative information and those that do not.
#XML reading with support for schema - [http://xerces.apache.org/xerces-c/ Xerces], check license compatibility.
#Look into the use of [[Proposals:Condor|Condor ]] for distributed computing.