Modules:RegisterImages-Documentation-3.6

Return to Slicer 3.6 Documentation

Gallery of New Features

Register Images

Before Two LUTs are used to display the two volumes

During Note that both cores run at full capacity

After Fixed data set with red, moving with blue LUT.

After Moving versus Resampled Demonstration of the extent of deformation grid spacing = 60 pixels, 4 core machine, 9GB memory footprint, roughly 1.5 hours volume=512x512x472

General Information

Module Type & Category

Type: CLI Module calling ITK functions

Category: Registration

Authors, Collaborators & Contact

• Author1: Stephen Aylward, Kitware
• Author2: Casey Goodlett, Kitware
• Contributor1: Ron Kikinis (copying documentation)
• Contact: Casey Goodlett, Kitware

Module Description

This Module is an integrated framework providing access to ITK registration technologies. Algorithms can be run in single mode or pipelined. Depending on the size of the data sets, a significant amount of memory is needed. There is an option to trade off speed for memory. Most of the code is parallelized and will take advantage of multicore capabilities, if available.

New Features in 3.6

• Integration of transform nodes instead of reading/writing to files
• Integration of landmark module for specifying initialization
• Ability to specify a label mask on the fixed image for limiting regions of metric calculation
• Exposed metrics besides Mutual Information
• Simplified parameters
• Ability to include parameter presets distributed with Slicer

Major Features

The major features of the module include:

• Default parameters register many full-head and skull-stripped MRI: rigid, affine, and BSpline
• Offers a complete, pipeline-based registration solution
  • Load and apply existing transforms
  • Compute rigid, affine, and bspline transforms in sequence with a single command
• Intuitive parameters
  • Instead of setting obscure "scales" for parameters, you set global values for "Expected Offset", "Expected Rotation", ... to indicate how much mis-registration is anticipated in the data being registered
• MinimizeMemory option provides a way to compute bspline registrations using a dense set of control points and a large number of samples on "normal" computers (albeit computation time increases)
• SampleFromOverlap option allows images of vastly different sizes to be registered
  • Helps to avoid (but does not completely eliminate) the annoying ITK exception, "too many samples falls outside of the image"
• Based on an extensible and re-usable class structure.

Usage

• If you have a large data set or little memory, select the minimize memory option
• If you are planning to use the b-spline registration, be sure to read the discussion of parameter setting in the section below.

Use Cases, Examples

See Registration Use Case project

Quick Tour of Features and Use

See here for more information.

Main Registration Panel

Parameter Choices

• The default parameters should work for a wide range of cases
• In some cases (e.g. unusual acquisition conditions and/or highly inconsistent acquisition protocols) you will need to change the default parameters.
• In other situations, you may wish to tweak parameters to achieve your application-specific speed-vs-accuracy trade-off

See also this page on the NA-MIC wiki about parameter tuning for this module.

Overview

• Step 1: Loaded transform
  • You may load a pre-computed transform to initialize the registration.
  • If one is loaded, it is immediately applied (i.e., the moving image is resampled)
• Step 2: Initial registration
  • Options are:
    • None (sets the center of rotation to the center of the moving image)
    • Landmark (uses N-pairs of landmarks (passed as vectors) and a least-squared error metric to register the images using a rigid transform
    • Image Centers (shifts the images to align their centers)
    • Centers of Mass (shifts the images to align their centers of mass)
    • Second Moments (shifts and rotates the images to align the 1st and 2nd moments)
• Step 3: Registration
  • Options are:
    • None (applies the loaded transforms)
    • Initial
      • computes and applies the initial transform to the loaded registrations)
    • Rigid
      • computes a rigid transform and then applies it to the loaded registrations
    • Affine
      • computes an affine transform and then applies it to the loaded registrations
    • BSpline
      • computes a bspline transform and then applies it to the loaded registrations
    • PipelineRigid
      • computes a rigid transform (initialized using the results from the initial registration) and then applies it to the loaded registrations
    • PipelineAffine
      • computes a rigid transform (initialized using the results from the initial registration), uses those results to initialize and compute an affine transform, and then applies it to the loaded registrations
    • PipelineBSpline
      • computes a rigid transform (initialized using the results from the initial registration), uses those results to initialize and compute an affine transform, and then applies it to the loaded registrations, THEN computes and applies a BSpline transform

Detailed Information

IO Tab

Set the fixed and moving images using images in the scene Optionally set the ResampleImage to store the output image If not set, registration won't conduct the final resampling, saving computation time

IO panel Resample image is the resampled moving image.

Registration Parameters Tab

Load Transform provide the a transform for initializing the registration Save Transform final transform will be saved here Initialization see registration pipeline discussion Registration see registration pipeline discussion For rigid and affine registrations, one-plus-one evoluation optimization is first applied for N iterations, and then FRPR gradient-line-search optimization is applied. For more information, check the code: RegisterImagesModule/itkOptimizedImageToImageRegistrationmethod.h/txx For BSpline registration, a hierarchical registration scheme is used. An image pyramid having 3 levels is used to resample the images and the control grids. Heuristics are used to control the various resampling parameters. At each level, registration is conducted using FRPR gradient-line-search optimization. For more information, check the code: RegisterImagesModule/itkBSplineImageToImageRegistrationMethod.h/txx Metric Mutual Information metric. It is an multithreaded and optimized version of the Mattes MI method. For more information, check the code; Insight/Code/Review/itkOptMattesMutualInformationImageToImageMetric.h/txx Normalized Correlation metric. This is NOT multithreaded or optimized For more information, check the code; Insight/Code/Algorithms/itkNormalizedCorrelationImageToImageMetric.h/txx Mean Squares metric. It is an multithreaded and optimized version of the Mattes MI method. For more information, check the code; Insight/Code/Review/itkOptMeanSquaresImageToImageMetric.h/txx "Expected" values For rigid, affine, and bspline registration, parameter scales (refer to the Insight Software Guide) are represented as hyper-parameters in the RegisterImages module. "Expected Offset" controls the offset scales in rigid and affine registration the deformation vector scale in bspline registration "Expected Rotation" is roughly in terms of radians. It controls the rotation angles in rigid and affine registration "Expected Scale" is for scaling during affine registration "Expected Skew" is for skew for affine registration

Global Parameters This panel controls what will happen overall.

Advanced Registration Parameters Tab

Verbosity level Controls the level of detail in the reports in the log file Sample from fixed/moving overlap When the fixed image is much larger than the moving image, it is CRITICAL to set this flag and to pick a good initialization method. In that way, only the portion of the fixed image that is initially covered by the moving image will be used during registration. This prevents ITK from throwing an exception (error) stating that too many fixed-image samples miss (map outside of) the moving image. Fixed image mask A label image that specifies regions of the fixed image from which to draw samples for registration. This should have a label in regions you are interested in registration and have no label in regions you are not interested in or may contain significant differences between the image that should be excluded (pathology regions). Random number seed To ensure consistent performance, you can set a seed - repeated runs should produce identical results. Number of threads Number of multi-core/multi-processor threads to use during metric value computations. MimimizeMemory Turns off caching of intermediate values during bspline registration Provides a way to compute bspline registrations using a dense set of control points and a large number of samples on "normal" computers (albeit computation time increases) Rule of thumb, if the BSpline registration crashes - re-run with this option enabled. use windowed sinc for final interpolation If you have time to kill. Extremely slow and only marginally better than bspline resampling (the default).

Advanced Global Parameters Tab

Advanced Initial Registration Parameters

Fixed / Moving Landmarks Two fiducial lists of corresponding points in the fixed and moving images. Only selected fiducials are used and the fixed and moving landmarks must have the same number of landmarks in the same order. If supplied, then choose "Landmarks" as the initial registration method (see discussion on registration pipeline)

Initialization of rigid registration

Advanaced Rigid and Affine Parameters

MaxIterations Number of iterations for one-plus-one and for FRPR registration Sampling Ratio Portion of the image pixels to be used when computing the metric

Advanced Parameters for rigid registration Advanced Parameters for affine registration

Advanced BSpline Parameters

MaxIterations Number of iterations for one-plus-one and for FRPR registration Sampling Ratio Portion of the image pixels to be used when computing the metric Do the math...if you have 40 pixels between control points, then there will be 40^3 (64,000) pixels relevant to each control point. That excessive for directing one control point. Keep the sampling small. For 40 pixels between control points, a sampling density of 0.1 provide 6,400 pixels for metric computation at each control point - more than enough. When in doubt, turn on MinimizeMemory Control point spacing (pixels) Don't think about grid size - instead think about the level of detail that needs to be resolved (see discussion on sampling ratio). When in doubt, turn on MinimizeMemory

Advanced Parameters for B-Spline registration

Use Cases

Example 2: Affine Registration

• Task:
  • Affine registration of head MRI from two different subjects
• Data:
  • Using cases UNC-Healthy-Normal002 (fixed) and UNC-Healthy-Normal004 (moving)
  • Data provided by Dr. Bullitt at UNC.
  • Data is available from Kitware's MIDAS archive at http://hdl.handle.net/1926/542
  • Data can be automatically downloaded into ${RegisterImages_BINARY_DIR}/Testing/Data directory by enabling the CMake variable "BUILD_REGISTER_IMAGES_REAL_WORLD_TESTING"
    • Warning this also enables additional tests that can take 4+ hours to complete.
    • To see the code for automatically downloading from MIDAS (via svn), see Slicer3/Applications/CLI/RegisterImagesModule/Applications/CMakeLists.txt

Example 3: BSpline Registration

• Task:
  • BSpline registration of head MRI from two different subjects
• Data:
  • Using cases UNC-Healthy-Normal002 (fixed) and UNC-Healthy-Normal004 (moving)
  • Data provided by Dr. Bullitt at UNC.
  • Data is available from Kitware's MIDAS archive at http://hdl.handle.net/1926/542
  • Data can be automatically downloaded into ${RegisterImages_BINARY_DIR}/Testing/Data directory by enabling the CMake variable "BUILD_REGISTER_IMAGES_REAL_WORLD_TESTING"
    • Warning this also enables additional tests that can take 4+ hours to complete.
    • To see the code for automatically downloading from MIDAS (via svn), see Slicer3/Applications/CLI/RegisterImagesModule/Applications/CMakeLists.txt

Development

Dependencies

This module makes use of GenerateCLP, MRMLIO and ITK.

• GenerateCLP is used to defined the command-line options. GenerateCLP is distributed with Slicer (/Libs/GenerateCLP).
• MRMLIO is used to replace reading/writing images in ITK to instead passing pointers to shared memory. MRMLIO is distributed with Slicer (/Libs/MRMLIO).
• ITK is used to provide the algorithms. ITK must be compiled with the CMake variables USE_REVIEW, USE_TRANSFORM_IO, and USE_OPTIMIZED_REGISTRATION_METHODS enabled. These options are automatically enabled when ITK is build using Slicer's getBuildTest.tcl script.

Known bugs

• Module may fail (and not report a useful error to the console - check the log files) if insufficient memory is available. If so, please consider enabling the MinimizeMemory option in the Advanced Options tab.

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.

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. Information on the National Centers for Biomedical Computing can be obtained from National Centers for Biomedical Computing.

References

Publications related to this module go here. Links to pdfs would be useful.

• some email about initialization
• ITK Registration optimization
• Slicer2 vs Slicer3 registration validation
• Using BatchMake with Slicer RegisterImages module

Wishlist

• include moving filename into default names of results
• display fixed and moving results in compareview with proper LUTs (different colors for fixed and result, apply w/l from moving to results volume
• use in combination with VOI
• more detailed explanation of how to set parameters
• increase number of example solutions