Litcius/Paper detail

On Finite Difference Jacobian Computation in Deformable Image  Registration

Yihao Liu, Junyu Chen, Shuwen Wei, Aaron Carass, Jerry L. Prince

2024International Journal of Computer Vision11 citationsDOIOpen Access PDF

Abstract

Abstract Producing spatial transformations that are diffeomorphic is a key goal in deformable image registration. As a diffeomorphic transformation should have positive Jacobian determinant $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> everywhere, the number of pixels (2D) or voxels (3D) with $$\vert J\vert &lt;0$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> <mml:mo>&lt;</mml:mo> <mml:mn>0</mml:mn> </mml:mrow> </mml:math> has been used to test for diffeomorphism and also to measure the irregularity of the transformation. For digital transformations, $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> is commonly approximated using a central difference, but this strategy can yield positive $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> ’s for transformations that are clearly not diffeomorphic—even at the pixel or voxel resolution level. To show this, we first investigate the geometric meaning of different finite difference approximations of $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> . We show that to determine if a deformation is diffeomorphic for digital images, the use of any individual finite difference approximation of $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> is insufficient. We further demonstrate that for a 2D transformation, four unique finite difference approximations of $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> ’s must be positive to ensure that the entire domain is invertible and free of folding at the pixel level. For a 3D transformation, ten unique finite differences approximations of $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> ’s are required to be positive. Our proposed digital diffeomorphism criteria solves several errors inherent in the central difference approximation of $$\vert J\vert $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mo>|</mml:mo> <mml:mi>J</mml:mi> <mml:mo>|</mml:mo> </mml:mrow> </mml:math> and accurately detects non-diffeomorphic digital transformations. The source code of this work is available at https://github.com/yihao6/digital_diffeomorphism .

Topics & Concepts

AlgorithmArtificial intelligenceComputer scienceMedical Image Segmentation TechniquesRobotics and Sensor-Based LocalizationDNA and Biological Computing
On Finite Difference Jacobian Computation in Deformable Image  Registration | Litcius