Dense image registration through MRFs and efficient linear programming

By Ben Glocker, Nikos Komodakis, G. Tziritas, Nassir Navab, Nikos Paragios. Medical Image Anal., 2008

482 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computational Mechanics, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/88d040a394da779919d54d9a13a2cca792c4a6ff

How researchers see this paper

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

In Aerospace Engineering

In Computational Mechanics

In Computer Vision and Pattern Recognition

Citations in the corpus

Evaluation of Registration Methods on Thoracic CT: The EMPIRE10 Challenge (2011, unknown): “• E (DROP) [44], [45]: After initial pre-alignment, the dense intensity-based registration is performed using hierarchical FFDs and iterative discrete labeling of MRFs for the energy minimization.”
MSM: a new flexible framework for Multimodal Surface Matching☆ (2014, unknown): “Consequently, it has become common practice within discrete volumetric approaches (Glocker et al., 2008) to reduce the degrees of freedom by using a multi-resolution series of control point grids within a B-Spline deformation framework, similar to those used by continuous approaches (Rueckert et al.”
Deformable medical image registration: setting the state of the art with discrete methods. (2011, methodology): “We should note that further experimental results for a variety of other medical registration settings can be found in our earlier works on multi-modal registration (81), atlas-based segmentation via registration (82), deformable stitching for whole-body MRI (83), and others (31,59,61,84,85).”
Unsupervised Learning for Fast Probabilistic Diffeomorphic Registration (2018, methodology): “Traditional methods solve an optimization over the space of deformations, such as elastic-type models [4], B splines [25], dense vector fields [27], or discrete methods [8,12].”
Recursive Cascaded Networks for Unsupervised Medical Image Registration (2019, methodology): “Among traditional algorithms, an iterative approach is commonly suggested [2, 3, 4, 7, 10, 18, 27, 52], where they formulate each iteration as a progressive optimization problem.”
Survey of Non-Rigid Registration Tools in Medicine (2017, background): “& GRAD: The sum of gradient inner products [45]”
Learning clinically useful information from images: Past, present and future (2016, methodology): “…discrete optimisation methods that had been successfully applied to image segmentation tasks turned out to be very well suited for FFD-based non-rigid registration [Glocker et al. (2008)] and allow the registration of 3D medical images in time frames that are acceptable in clinical workﬂows.”
Slice-to-volume medical image registration: a survey (2017, background): “Nevertheless, recent works on image registration have focused on discrete formulations (Glocker and Sotiras, 2011; Glocker et al., 2008), both for projective and sliced 2D/3D registration (Zikic et al., 2010b,a; Ferrante and Paragios, 2013; Ferrante et al., 2015a,b).”
VoxelMorph: A Learning Framework for Deformable Medical Image Registration (2018, methodology): “These include elastictype models [8], [22], [23], statistical parametric mapping [24], free-form deformations with b-splines [25], discrete methods [17], [18] and Demons [19], [26].”
DRAMMS: Deformable Registration via Attribute Matching and Mutual-Saliency Weighting (2009, methodology): “Existing deformable registration methods can be generally classiﬁed into two main categories: voxel-wise methods (e.g., [1,2,3,4,5,6,7]) and landmark/feature-based methods (e.g., [8,9,10,11,12,13,14]).”
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 (2015, methodology): “The myocardial segmentation masks available from other subject(s) are generally propagated to unseen data in Atlas-based techniques [2] using non-rigid registration algorithms like diﬀeomorphic demons (dDemons) [20], FFD-MI [3], probabilistic label fusion or SVM [2].”
Multimodal Surface Matching with Higher-Order Smoothness Constraints☆ (2017, methodology): “The optimal rotation for each control-point Rp 2 Sp is found using discrete optimisation (Robinson et al., 2014; Glocker et al., 2008).”
Deformable Medical Image Registration: A Survey (2013, methodology): “FastPD was rst used for image registration to infer the displacements of a grid-based deformation model in [412, 413].”
An Unsupervised Learning Model for Deformable Medical Image Registration (2018, background): “There is extensive work in 3D medical image registration [2, 4, 6, 7, 13, 18, 42].”