Fast parametric elastic image registration

Abstract

Abstract-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.

Correlation between pretreatment FDG-PET biological target volume and anatomical location of failure after radiation therapy for head and neck cancers

“…The PET/CT data were transferred to U-M Plan, our in-house treatment planning system using a standard DICOM mechanism. The CT data from the PET/CT were registered to the treatment planning CT using a multiresolution B-spline transformation model [18,19] and a mutual information similarity metric [20,21]. The algorithm first registers the two datasets using a rotate-translate transformation model and uses this result as the starting point for a course to fine B-spline registration [19].…”

Organ-focused mutual information for nonrigid multimodal registration of liver CT and Gd–EOB–DTPA-enhanced MRI

“…In Sections 2.1 and 2.2, we will introduce the general framework and MI, describing briefly the equations presented in previous work (Kybic and Unser, 2003;Thévenaz and Unser, 2000). We will then explain our contribution in Section 2.3.…”

“…The intensity-based nonrigid registration algorithm used extends the previous B-spline method of Kybic and Unser (2003). The algorithm determines a set of B-spline coefficients that describe a nonrigid transformation that maximizes an image similarity measure.…”

“…B-spline functions have been widely used to represent deformations (Kybic and Unser, 2003;Ledesma-Carbayo et al, 2005;Oguro et al, 2009;Rueckert et al, 1999;Schnabel et al, 2001), motivated by their compact support, computational simplicity, good approximation properties, and implicit smoothness. We also use B-spline functions for representing continuous images derived from a set of samples (Kybic and Unser, 2003;Thévenaz and Unser, 2000). Moreover, B-spline basis functions are used as Parzen windows (Thévenaz and Unser, 2000) in the similarity criteria, as described later.…”

Automatic Classification of Left Ventricular Regional Wall Motion Abnormalities in Echocardiography Images Using Nonrigid Image Registration

“…In this step, we employ an iterative multiresolution optimization approach in which the resolution of the mesh of control points is increased, along with the image resolution, in a coarse to fine fashion to obtain a higher speed and robustness [41]. In addition, in each step, we apply an iterative gradient descent method [42] which steps in the direction of the gradient vector with a certain step size μ. The algorithm stops if the ║∇C║ ≤ ε for some small positive threshold, where ∇C is the gradient vector of the cost function.…”