Documentation/4.6/Extensions/PBNRR

From Slicer Wiki
Revision as of 06:49, 7 November 2016 by UpdateBot (talk | contribs) (Nightly -> 4.6)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Home < Documentation < 4.6 < Extensions < PBNRR


For the latest Slicer documentation, visit the read-the-docs.



Introduction and Acknowledgements

Extension: Physics-Based Non-Rigid Registration (PBNRR)
Acknowledgments: This work is funded mainly by the ARRA funds for the ITK-v4 implementation with grant number:NLMA2D2201000586P. In addition, this work is supported in part by NSFgrants:CCF-1139864, CCF-1136538, and CSI-1136536 and by the John Simon Guggenheim Foundation and the Richard T.Cheng Endowment.
Author: Fotis Drakopoulos
Contributors: Fotis Drakopoulos (CRTC), Yixun Liu (CRTC), Andriy Kot (CRTC), Andrey Fedorov (SPL B&W Harvard), Olivier Clatz (Asclepios INRIA), Nikos Chrisochoides (CRTC)
Contact: Nikos Chrisochoides, <email>npchris@gmail.com</email>
Website: https://crtc.cs.odu.edu/
License: BSD

Center for Real-time Computing  

Module Description

The module Non-Rigid Registers a moving to a fixed MRI. It uses a linear homogeneous bio-mechanical model to compute a dense deformation field that defines a transformation for every point in the fixed image to the moving image. The method includes three components (Feature Point Selection, Block Matching and Finite Element Solver) combine together to provide a user-friendly interface.

Use Cases

Case1 Input; top left: moving image, top right: mesh, bottom left: fixed image, bottom right: mask image.
Case1 Output; top left: warped image, top right: warped mesh, bottom left: direct deformation field, bottom right: inverse deformation field.
Case2 Input; top left: moving image, top right: mesh, bottom left: fixed image, bottom right: mask image.
Case2 Output; top left: warped image, top right: warped mesh, bottom left: direct deformation field, bottom right: inverse deformation field.

Tutorials

Panels and their use

  • Registration Parameters:
    • Block Radius: Radius (in image voxels) of the selected image blocks in each dimension (default: 1,1,1).
    • Window Radius: Radius (in image voxels) of the Block Matching window in each dimension (default: 5,5,5).
    • Selection Fraction: Fraction of the selected blocks from the total number of image blocks. Value should be between [0.01,1] (default: 0.05).
    • Rejection Fraction: Fraction of the rejected blocks from the number of the selected image blocks. Value should be between [0.01,1) (default: 0.25).
    • Outlier Rejection Steps: Number of outlier rejection steps. Value should be between [1,20] (default: 5).
    • Interpolation Steps: Number of interpolation steps. Value should be between [1,20] (default: 5).
    • Young Modulus: Young Modulus of the linear bio-mechanical model (default: 0.0021 N/mm2).
    • Poisson Ratio: Poisson Ratio of the linear bio-mechanical model. Value should be between [0.10,0.49] (default: 0.45).
  • Input/Output:
    • Input Moving Image: The input moving image.
    • Input Fixed Image: The input moving image.
    • Input Mask Image: The input moving image.
    • Input Mesh: The input moving image.
    • Output Volume: Moving image to the fixed image coordinate frame (optional).
    • Output Direct Deformation Field: Transform calculated that aligns the fixed and moving image. Maps positions in the moving coordinate frame to the fixed coordinate frame (optional).
    • Output Inverse Deformation Field: Transform calculated that aligns the fixed and moving image. Maps positions in the fixed coordinate frame to the moving coordinate frame (optional).
    • Output Warped Mesh: The warped tetrahedral mesh in vtk file format (optional).
PBNRR UI

Similar Modules

N/A

References

  • Olivier Clatz, Hervé Delingette, Ion-Florin Talos, Alexandra J. Golby, Ron Kikinis, Ferenc Jolesz, Nicholas Ayache, and Simon Warfield,“Robust Non-Rigid Registration to Capture Brain Shift from Intra-Operative MRI“, IEEE Transactions on Medical Imaging, 24(11):1417-1427, Nov. 2005.
  • Liu Y, Kot A, Drakopoulos F, Yao C, Fedorov A, Enquobahrie A, Clatz O and Chrisochoides NP (2014), ”An ITK implementation of a physics-based non-rigid registration method for brain deformation in image-guided neurosurgery”, Frontiers in Neuroinformatics 8:33. doi: 10.3389/fninf.2014.00033

Information for Developers