Fast free-form deformation using graphics processing units

By Marc Modat, Gerard R. Ridgway, Zeike A. Taylor, Manja Lehmann, Josephine Barnes, D. Hawkes, Nick C Fox, Sébastien Ourselin. Comput. Methods Programs Biomed., 2010

1,203 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/21b5e050525ecbec2119e47409a296eb29edc115

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

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Rationale, design and methodology of the image analysis protocol for studies of patients with cerebral small vessel disease and mild stroke (2015, methodology): “For each dataset, nonlinear registration (Modat et al. 2010) is used to align the tissue masks at baseline in the structural space (T2W) with the parametric maps in the diffusion space.”
Medical image processing on the GPU - Past, present and future (2013, background): “…(Ha et al., 2010; Oh et al., 2011; Huang et al., 2011), diffeomorphic registration algorithms (Han et al., 2010; Huang et al., 2010), free-form deformation (Modat et al., 2010), motion tracking of video microscopy (Liu et al., 2012) and mass-conserving image registration (Castillo et al., 2012).”
Attenuation Correction Synthesis for Hybrid PET-MR Scanners: Application to Brain Studies (2014, methodology): “This inter-subject coordinate mapping was obtained using a symmetric global registration followed by a cubic B-spline parametrized nonrigid registration, using normalized mutual information as a measure of similarity [27].”
A survey of medical image registration on graphics hardware (2011, background): “[53] M.”
Unsupervised Learning of Probabilistic Diffeomorphic Registration for Images and Surfaces (2019, methodology): “In our work, we bridge these two paradigms to offer classical guarantees within a machine learning approach.”
Deep learning in medical image registration: a review (2019, methodology): “used DVF generated from traditional DIRs including PlastiMatch [111], NiftyReg [127] and VarReg [162] as ground truth [126].”
Survey of Non-Rigid Registration Tools in Medicine (2017, methodology): “The tool implements the Fast Free-Form Deformation algorithm [74] and offers a substantially decreased computational time compared to the classical FreeForm Deformation (FFD) algorithm [65].”
Geodesic Information Flows: Spatially-Variant Graphs and Their Application to Segmentation and Fusion (2015, methodology): “In our work, we use a symmetric variant of a non-rigid free-form registration algorithm [26].”
3D Whole Brain Segmentation using Spatially Localized Atlas Network Tiles (2019, background): “Many efforts have been proposed to accelerate the process using GPU acceleration [65], atlas selection [25], fast patch matching [66] etc.”
Deep learning in medical image registration: a survey (2019, methodology): “used DVF generated from traditional DIRs including PlastiMatch [111], NiftyReg [127] and VarReg [162] as ground truth [126].”
Deformable MR Prostate Segmentation via Deep Feature Learning and Sparse Patch Matching (2015, methodology): “Then, the free-form deformation (FFD) [37] is further adopted to the linearly aligned images for deformable registration.”
GPU computing in medical physics: a review. (2011, methodology): “68,69 B-spline image registration, a method popular for its ﬂexibility and robustness, computes a ﬁne displacement ﬁeld by interpolating a coarse grid of uniformly spaced control points using piecewise continuous B-splines basis functions.”
VoxelMorph: A Learning Framework for Deformable Medical Image Registration (2018, background): “Recent GPU implementations have reduced this runtime to just minutes, but require a GPU for each registration [4].”
Local Temperature Scaling for Probability Calibration (2020, methodology): “For preprocessing, all images are first affinely registered to the ICBM MNI152 nonlinear atlas [18] using NiftyReg [49, 50, 60] and intensity normalized via histogram equalization.”
Evaluation of Registration Methods on Thoracic CT: The EMPIRE10 Challenge (2011, neutral): “• M (Nifty Reggers) [56], [57]: A block-matching technique was used to perform an initial affine alignment.”
Multi-Atlas Segmentation of Biomedical Images: A Survey (2014, methodology): “This approach has already been used for the GPU-acceleration of the registration step (Jorge Cardoso et al., 2013; Duc et al., 2013; Han et al., 2009; Modat et al., 2010).”
TransMorph: Transformer for unsupervised medical image registration (2021, positive): “• NiftyReg 5 (Modat et al., 2010): We used the sum of squared differences (SSD) as the objective function and bending energy as the regularizer for all registration tasks.”
Deformable Medical Image Registration: A Survey (2013, methodology): “Other e orts include the encoding of similarity metrics such as Normalized Mutual Information [55,56].”