Difference between revisions of "Projects/Structural/2007 Project Week Nonrigid Groupwise Registration"

From NAMIC Wiki
Jump to: navigation, search
 
(23 intermediate revisions by the same user not shown)
Line 1: Line 1:
 +
{|
 +
|[[Image:ProjectWeek-2007.png|thumb|left|320px|Return to [[2007_Programming/Project_Week_MIT|Project Week Main Page]] ]]
 +
<table width="420" border="1" align="center" >
 +
        <caption>
 +
        </caption>
 +
        <tr>
 +
          <td width="210" align="center"> [[Image:MeanSliceSerdar.jpg|thumb|center|Mean Image before registration]]  </td>
 +
          <td width="210" align="center"> [[Image:MeanSlice1Serdar.jpg|thumb|center|Mean Image after registration]]  </td>
 +
        </tr>
 +
        <tr>
 +
          <td width="210" align="center"> [[Image:STDSliceSerdar.jpg|thumb|center|STD Image before registration]] </td>
 +
          <td width="210" align="center"> [[Image:STDSlice1Serdar.jpg|thumb|center|STD Image after registration]] </td>
 +
        </tr>
 +
        <tr>
 +
          <td colspan="2"> Registration results on a synthetic dataset of 30 MR images. <td>
 +
        </tr>
 +
</table>
 +
|}
 +
 +
 +
 +
 +
__NOTOC__
 
===Key Investigators===
 
===Key Investigators===
* MIT: Serdar K Balci, Polina Golland
+
* MIT: Serdar K Balci, Polina Golland, Lilla Zollei, Sandy Wells
 
* Kitware: Luis Ibanez
 
* Kitware: Luis Ibanez
  
Line 20: Line 43:
 
<h1>Approach, Plan</h1>
 
<h1>Approach, Plan</h1>
  
We implemented a non-rigid groupwise registration method using sum of pixelwise entropies  
+
* Implement a non-rigid groupwise registration method  
as a joint alignment criterion.
+
** Joint alignment of images
We have chosen B-splines as our non-rigid deformation model and optimized ITK's B-Spline implementation by computing jacobian field locally.
+
** Minimize sum of univariate entropies
We extended ITK's pairwise registration framework to include groupwise registration methods.
+
** Efficient implementation using multi-threaded classes.
We provide an efficient implementation of the objective function using multi-threaded classes.
+
* B-splines as the non-rigid deformation model  
To compare our registration framework to a pairwise approach we implemented a template based method where we register every subject to the mean image using sum of squared differences.
+
** Optimize ITK's B-Spline implementation by computing jacobian field locally
 
+
*Extended ITK's pairwise registration framework to groupwise registration  
Our plan is to incorporate our implementation into ITK and to provide test cases for validation.
+
*To compare groupwise registration to pairwise approaches
 
+
** Implement a template based method where every subject is registered to the mean image using sum of squared differences.
  
 
</div>
 
</div>
Line 36: Line 59:
 
<h1>Progress</h1>
 
<h1>Progress</h1>
  
We implemented the project in ITK and submitted the code to ITK's Namic SANDBOX repositories.
+
Our project is already implemented in ITK, in this workshop we achieved to
 
+
*Submit the project to ITK's Namic SANDBOX repositories [http://www.na-mic.org/svn/NAMICSandBox/trunk/MultiImageRegistration/ link to code]
 
+
*Provide test cases using ITK's Brainweb data [http://public.kitware.com/pub/itk/Data/BrainWeb/BrainPart2.tgz  link to data]
 +
*Submit test results to the dashboard
 +
*Make modifications to the source code for memory efficiency
 +
*Build a package for Insight Journal submission (in progress)
 
</div>
 
</div>

Latest revision as of 14:11, 29 June 2007

Home < Projects < Structural < 2007 Project Week Nonrigid Groupwise Registration
Mean Image before registration
Mean Image after registration
STD Image before registration
STD Image after registration
Registration results on a synthetic dataset of 30 MR images.



Key Investigators

  • MIT: Serdar K Balci, Polina Golland, Lilla Zollei, Sandy Wells
  • Kitware: Luis Ibanez

Objective

We extended a previously demonstrated entropy based groupwise registration method to include a free-form deformation model based on B-splines.

The objective in groupwise registration is to bring subjects in a population into joint alignment in order to establish correspondences among anatomical structures.


Approach, Plan

  • Implement a non-rigid groupwise registration method
    • Joint alignment of images
    • Minimize sum of univariate entropies
    • Efficient implementation using multi-threaded classes.
  • B-splines as the non-rigid deformation model
    • Optimize ITK's B-Spline implementation by computing jacobian field locally
  • Extended ITK's pairwise registration framework to groupwise registration
  • To compare groupwise registration to pairwise approaches
    • Implement a template based method where every subject is registered to the mean image using sum of squared differences.

Progress

Our project is already implemented in ITK, in this workshop we achieved to

  • Submit the project to ITK's Namic SANDBOX repositories link to code
  • Provide test cases using ITK's Brainweb data link to data
  • Submit test results to the dashboard
  • Make modifications to the source code for memory efficiency
  • Build a package for Insight Journal submission (in progress)