Fast parametric elastic image registration

By Jan Kybic, Michael A. Unser. IEEE Transactions on Image Processing, 2003

538 citations recorded by the source corpus.

We present an algorithm for fast elastic multidimensional intensity-based image registration with a parametric model of the deformation. It is fully automatic in its default mode of operation. In the case of hard real-world problems, it is capable of accepting expert hints in the form of soft landmark constraints. Much fewer landmarks are needed and the results are far superior compared to pure landmark registration. Particular attention has been paid to the factors influencing the speed of this algorithm. The B-spline deformation model is shown to be computationally more efficient than other alternatives. The algorithm has been successfully used for several two-dimensional (2-D) and three-dimensional (3-D) registration tasks in the medical domain, involving MRI, SPECT, CT, and ultrasound image modalities. We also present experiments in a controlled environment, permitting an exact evaluation of the registration accuracy. Test deformations are generated automatically using a random hierarchical fractional wavelet-based generator.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/a76201c881421fc6af36c9082f6e4b076bbb5714

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

elastix: A Toolbox for Intensity-Based Medical Image Registration (2010, unknown): “[41], normalized correlation (NC) [42], [43], mutual informa-”
Deformable Medical Image Registration: A Survey (2013, methodology): “Unser [136] compared the LM algorithm with GD, GD with a”
Elastic registration of biological images using vector-spline regularization (2005, methodology): “A deformation model based on B-splines is very versatile and can generate a large variety of nonlinear elastic deformations, while remaining easy to handle [12], [13], [14], [16], [20].”
DRAMMS: Deformable Registration via Attribute Matching and Mutual-Saliency Weighting (2009, background): “Examples of the latter category include (Glocker et al., 2008; Vercauteren et al., 2007, 2009; Ou and Davatzikos, 2009; D’Agostino et al., 2003; Kybic and Unser, 2003; Christensen et al., 1994; Collins et al., 1994; Thirion, 1998; Rueckert et al., 1999; Maes et al., 1997; Wells et al., 1996; Friston”
Diffeomorphic Registration Using B-Splines (2006, methodology): “A popular approach is to use spline-based transformations [8,17].”
A Review of Geometric Transformations for Nonrigid Body Registration (2008, background): “mation is defined as follows [122]:”
Medical image registration: a review (2014, background): “…for free-form deformation is probably the cubic B-spline (Rueckert et al. 1999; Studholme et al. 2000; Rohlﬁng and Maurer 2001; Rohlﬁng et al. 2003; Kybic and Unser 2003; Mattes et al. 2003; Kabus et al. 2004; Xie and Farin 2004; Balci et al. 2007; Bhagalia et al. 2009; Bai and Brady 2011;…”
Evaluation of Optimization Methods for Nonrigid Medical Image Registration Using Mutual Information and B-Splines (2007, methodology): “A possible subsampling strategy for the deterministic methods (GDD, GDL, QN, and NCG) is to select a single subset of voxels in the ﬁxed image and use these samples throughout the registration process [4], [43].”
Unsupervised End-to-end Learning for Deformable Medical Image Registration (2017, background): “Commonly adopted cost functions are the mean squared difference [32], mutual information [33], normalized mutual information [34] and normalized correlation [35].”
Spatio-temporal nonrigid registration for ultrasound cardiac motion estimation (2005, methodology): “As shown in [11], [38], and [40], B-splines constitute a good choice for the spatial basis functions, .”