Modules:HammerRegistration

From SlicerWiki
Revision as of 23:33, 28 April 2010 by Taox (talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Home < Modules:HammerRegistration

Return to Slicer 3.6 Documentation

Gallery of New Features

Module Name

Hammer Deformable Registration


Hammer Registration UI
Reference Image
Average of 40 registered images

General Information

Module Type & Category

Type: CLI

Category: Registration

Authors, Collaborators & Contact

  • Author: Guorong Wu, Xiaodong Tao, Jim Miller, Dinggang Shen
  • Contact: grwu at med.unc.edu, taox at research.ge.com

Module Description

HAMMER is an algorithm for elastic registration of medical images using geometric moment invariants as attributes and hierarchical attribute matching mechanism for finding deformation field. This module implements the algorithm described in 'HAMMER: Hierarchical Attribute Matching Mechanism for Elastic Registration', IEEE Trans. on Medical Imaging, 21(11):1421-1439, Nov 2002). Its inputs are skull stripped brain images with gray matter, white matter, and CSF segmentation.

Usage

Examples, Use Cases & Tutorials

  • This module, as an individual, takes as inputs two gray matter/white matter/ CSF segmented brain volumes and computes a deformation field based on attribute matching. The result deformation field can then be applied to the intensity volume corresponding to the moving image and generate a transformed intensity volume that matches the fixed image.
  • A command line example of the module is
 HammerRegistration -l 1,2,3 -i 50,50,50 
         FixedImageSegmentation.nrrd 
         MovingImageSegmentation.nrrd
         MovingImageIntensity.nrrd 
         ResampledMovingImageSegmentation.nrrd 
         ResampledMovingImageIntensity.nrrd
  • Tutorial at All-Hands Meeting 2010 can be downloaded here.

Quick Tour of Features and Use

  • Input/output panel:
Input/Output panels

In its current implementation, HammerRegistration takes three input volumes and generate two output volumes. Fixed Image is the tissue mask of the reference image. Moving Image is the tissue mask of the volume to be registered. Moving Intensity Image is the intensity volume corresponding to Moving Image. These volumes should not include extra-cranial tissue. The two output volumes -- Output Volume is the warped Moving Image and Output Intensity Image is the warped Moving Intensity Image. The output volumes are warped with the same deformation field.

  • Parameters panel:
Parameters panels

The two sets of parameters required are Tissue Labels, which are comma-separated CSF label (1 by default), gray matter label (2 by default), and white matter label (3 by default), and Iterations, which are number of iterations at three different resolutions. When using the default values of 50,20,20, it takes about 15 to 20 minutes for the module to register two volumes of size ~200x200x200.

  • Viewing panel:

Since Slicer3 takes into account the orientation of a volume, the re-oriented volume will not show any difference from the original volume.

Development

Dependencies

  • Depending on SkullStripper Module to remove extra-cranial tissue.
  • Depending on FuzzyTissueClassification Module to segment the brain tissue into gray matter, white matter, CSF.

Known bugs

None.

Usability issues

None.

Source code & documentation

Source Code: NITRC

XML Description: HammerRegistration.xml

Test: HammerRegistrationTest.cxx

Usage:

USAGE: 

   /Users/taox/dev/Slicer3-ext/HammerRegistration-build/lib/Slicer3/Plugins
                                        /HammerRegistration 
                                        [--returnparameterfile
                                        <std::string>]
                                        [--processinformationaddress
                                        <std::string>] [--xml] [--echo] [-i
                                        <std::vector<int>>] [-l
                                        <std::vector<int>>] [--]
                                        [--version] [-h] <std::string>
                                        <std::string> <std::string>
                                        <std::string> <std::string>


Where: 

   --returnparameterfile <std::string>
     Filename in which to write simple return parameters (int, float,
     int-vector, etc.) as opposed to bulk return parameters (image,
     geometry, transform, measurement, table).

   --processinformationaddress <std::string>
     Address of a structure to store process information (progress, abort,
     etc.). (default: 0)

   --xml
     Produce xml description of command line arguments (default: 0)

   --echo
     Echo the command line arguments (default: 0)

   -i <std::vector<int>>,  --iterations <std::vector<int>>
     Comma separated list of iterations, for low resolution, middle
     resolution, and high resolution. (default: 50,20,20)

   -l <std::vector<int>>,  --tissuelabel <std::vector<int>>
     Tissue label for csf, gm, and wm (in this order) in the input images.
     (default: 10,150,250)

   --,  --ignore_rest
     Ignores the rest of the labeled arguments following this flag.

   --version
     Displays version information and exits.

   -h,  --help
     Displays usage information and exits.

   <std::string>
     (required)  Fixed image to which to register

   <std::string>
     (required)  Moving image

   <std::string>
     (required)  Moving intensity image, (optional).

   <std::string>
     (required)  Resampled moving image to the fixed image coordinate
     frame.

   <std::string>
     (required)  Resampled moving intensity image to the fixed image
     coordinate frame (optional).


   Description: HAMMER is an algorithm for elastic registration of  
   medical images using geometric moment invariants as attributes and  
   hierarchical attribute matching mechanism for finding deformation  
   field. This module implements the algorithm described in 'HAMMER:  
   Hierarchical Attribute Matching Mechanism for Elastic Registration',  
   IEEE Trans. on Medical Imaging, 21(11):1421-1439, Nov 2002). Its  
   inputs are skull stripped brain images with gray matter, white   matter,
   and CSF segmentation.

   Author(s): Guorong Wu, Dinggang Shen, Xiaodong Tao, Jim
   Miller

   Acknowledgements: This work is part of the National Alliance for Medical
   Image Computing (NAMIC), funded by the National Institutes of Health
   through the NIH Roadmap for Medical Research, Grant U54 EB005149. This
   work is also partly supported by NIH Grant EB006733.

More Information

Acknowledgment

This work is part of the National Alliance for Medical Image Computing (NAMIC), funded by the National Institutes of Health through the NIH Roadmap for Medical Research, Grant U54 EB005149. This work is also partly supported by NIH Grant EB006733.

OASIS datasets were used to generate images on this page.

References

Dinggang Shen, Christos Davatzikos, HAMMER: Hierarchical Attribute Matching Mechanism for Elastic Registration.IEEE Trans Med Imaging. 2002 Nov;21(11):1421-39. PMID: 12575879.

Page on NA-MIC site can be found here.

UNC page on HAMMER.