## Nonrigid registration using free-form deformations: Application to breast MR images (1999)

### Cached

### Download Links

Venue: | IEEE Transactions on Medical Imaging |

Citations: | 375 - 23 self |

### BibTeX

@ARTICLE{Rueckert99nonrigidregistration,

author = {D. Rueckert and L. I. Sonoda and C. Hayes and D. L. G. Hill and M. O. Leach and D. J. Hawkes},

title = {Nonrigid registration using free-form deformations: Application to breast MR images},

journal = {IEEE Transactions on Medical Imaging},

year = {1999},

volume = {18},

pages = {712--721}

}

### Years of Citing Articles

### OpenURL

### Abstract

Abstract — In this paper we present a new approach for the nonrigid registration of contrast-enhanced breast MRI. A hierarchical transformation model of the motion of the breast has been developed. The global motion of the breast is modeled by an affine transformation while the local breast motion is described by a free-form deformation (FFD) based on B-splines. Normalized mutual information is used as a voxel-based similarity measure which is insensitive to intensity changes as a result of the contrast enhancement. Registration is achieved by minimizing a cost function, which represents a combination of the cost associated with the smoothness of the transformation and the cost associated with the image similarity. The algorithm has been applied to the fully automated registration of three-dimensional (3-D) breast MRI in volunteers and patients. In particular, we have compared the results of the proposed nonrigid registration algorithm to those obtained using rigid and affine registration techniques. The results clearly indicate that the nonrigid registration algorithm is much better able to recover the motion and deformation of the breast than rigid or affine registration algorithms. I.

### Citations

1270 |
Spline Models for Observational Data
- Wahba
- 1990
(Show Context)
Citation Context ... constrain the spline-based FFD transformation to be smooth, one can introduce a penalty term which regularizes the transformation. The general form of such a penalty term has been described by Wahba =-=[28]-=-. In 3-D, the penalty term takes the following form: (4) where denotes the volume of the image domain. This quantity is the 3-D counterpart of the 2-D bending energy of a thin-plate of metal and defin... |

752 | Alignment by maximization of mutual information
- WELLS
- 1997
(Show Context)
Citation Context ...quared differences or correlation, as a similarity measure. An alternative voxel-based similarity measure is mutual information (MI), which has been independently proposed by Collignon [10] and Viola =-=[11]-=-, and which has been shown to align images from different modalities accurately and robustly [29]. Mutual information is based on the concept of information theory and expresses the amount of informat... |

678 |
Principal Warps: Thin-Plate Splines and the Decomposition of Deformations
- Bookstein
- 1989
(Show Context)
Citation Context ...FD can be written as the 3-D tensor product of the familiar 1-D cubic B-splines where , , , , , and where represents the th basis function of the B-spline [22], [23] In contrast to thin-plate splines =-=[25]-=- or elastic-body splines [26], B-splines are locally controlled, which makes them computationally efficient even for a large number of control points. In particular, the basis functions of cubic B-spl... |

485 | Multimodality image registration by maximization of mutual information
- Maes, Collignon, et al.
- 1997
(Show Context)
Citation Context ...GID REGISTRATION USING FREE-FORM DEFORMATION 713 [8]). In particular, voxel-based similarity measures based on joint entropy [9], mutual information [10]–[13], and normalized mutual information [14], =-=[15]-=- have been shown to align images acquired with different imaging modalities, robustly. However, most of these approaches are limited either to rigid or affine transformations. In contrast, many nonrig... |

250 |
An overlap invariant entropy measure of 3D medical image alignment
- STUDHOLME, HILL, et al.
- 1999
(Show Context)
Citation Context ... NONRIGID REGISTRATION USING FREE-FORM DEFORMATION 713 [8]). In particular, voxel-based similarity measures based on joint entropy [9], mutual information [10]–[13], and normalized mutual information =-=[14]-=-, [15] have been shown to align images acquired with different imaging modalities, robustly. However, most of these approaches are limited either to rigid or affine transformations. In contrast, many ... |

215 |
Image matching as a diffusion process: an analogy with Maxwell’s demons
- Thirion
- 1998
(Show Context)
Citation Context ...owever, most of these approaches are limited either to rigid or affine transformations. In contrast, many nonrigid registration algorithms based on elastic deformations, such as animal [16] or demons =-=[17]-=-, rely on the assumption that the intensity of tissues between images remains constant. This is also true for nonrigid registration algorithms based on fluid deformations [18], [19]. A notable excepti... |

185 |
Hierarchical B-Spline Refinement
- FORSEY, BARTELS
- 1988
(Show Context)
Citation Context ...his case, the position of control point coincides with that of control point and the values of the new control points can be calculated directly from those of , using a B-spline subdivision algorithm =-=[27]-=-. In general, the local deformation of the breast tissues should be characterized by a smooth transformation. To constrain the spline-based FFD transformation to be smooth, one can introduce a penalty... |

130 | Matching 3-D anatomical surfaces with non-rigid deformations using octree-splines
- Szeliski, Lavallee
- 1996
(Show Context)
Citation Context ...e the 12 degrees of freedom of the transformation. In a similar fashion, the global motion model can be extended to higher order global transformations, such as trilinear or quadratic transformations =-=[21]-=-. B. Local Motion Model The affine transformation captures only the global motion of the breast. An additional transformation is required, which models the local deformation of the breast. The nature ... |

126 | Comparison and evaluation of retrospective intermodality brain image registration techniques
- West, Fitzpatrick, et al.
- 1997
(Show Context)
Citation Context ... measure is mutual information (MI), which has been independently proposed by Collignon [10] and Viola [11], and which has been shown to align images from different modalities accurately and robustly =-=[29]-=-. Mutual information is based on the concept of information theory and expresses the amount of information that one image contains about a second image where denote the marginal entropies of and denot... |

120 | Fast Fluid Registration of Medical Images
- Bro-Nielsen, Gramkow
- 1996
(Show Context)
Citation Context ...mal [16] or demons [17], rely on the assumption that the intensity of tissues between images remains constant. This is also true for nonrigid registration algorithms based on fluid deformations [18], =-=[19]-=-. A notable exception is the registration algorithm proposed by Meyer et. al [20], which is based on a thin-plate spline deformation and uses mutual information as a voxel-based similarity measure. Ho... |

108 | Scattered data interpolation with multilevel b-splines
- Lee, Wolberg, et al.
- 1997
(Show Context)
Citation Context ...nificantly across patients and with age. Therefore, it is difficult to describe the local deformation via parameterized transformations. Instead, we have chosen an FFD model, based on B-splines [22], =-=[23]-=-, which is a powerful tool for modeling 3-D deformable objects and has been previously applied to the tracking and motion analysis in cardiac images [24]. The basic idea of FFD’s is to deform an objec... |

64 |
A survey of medical image registration. Medical image analysis
- MAINTZ, VIERGEVER
- 1998
(Show Context)
Citation Context ...ny voxelbased similarity measures have shown promising results for multimodality image registration (for a detailed overview seesRUECKERT et al.: NONRIGID REGISTRATION USING FREE-FORM DEFORMATION 713 =-=[8]-=-). In particular, voxel-based similarity measures based on joint entropy [9], mutual information [10]–[13], and normalized mutual information [14], [15] have been shown to align images acquired with d... |

63 |
Automated multimodality image registration using information theory
- Collignon, Maes, et al.
- 1995
(Show Context)
Citation Context ...(for a detailed overview seesRUECKERT et al.: NONRIGID REGISTRATION USING FREE-FORM DEFORMATION 713 [8]). In particular, voxel-based similarity measures based on joint entropy [9], mutual information =-=[10]-=-–[13], and normalized mutual information [14], [15] have been shown to align images acquired with different imaging modalities, robustly. However, most of these approaches are limited either to rigid ... |

52 |
ªDemonstration of Accuracy and Clinical Versatility of Mutual Information for Automatic Multimodality Image Fusion Using Affine and Thin Plate Spline Warped Geometric Deformations,º Medical Image Analysis
- Meyer, Boes, et al.
- 1997
(Show Context)
Citation Context ...ween images remains constant. This is also true for nonrigid registration algorithms based on fluid deformations [18], [19]. A notable exception is the registration algorithm proposed by Meyer et. al =-=[20]-=-, which is based on a thin-plate spline deformation and uses mutual information as a voxel-based similarity measure. However, due to prohibitive computational complexity of the thin-plate spline warps... |

44 | Ayache N. Tracking and motion analysis of the left ventricle with deformable superquadratics. Med Image Anal 1996;1:129–149. Barron JL, Fleet D, Beauchemin S. Performance of optical flow techniques
- Bardinet, LD
(Show Context)
Citation Context ... chosen an FFD model, based on B-splines [22], [23], which is a powerful tool for modeling 3-D deformable objects and has been previously applied to the tracking and motion analysis in cardiac images =-=[24]-=-. The basic idea of FFD’s is to deform an object by manipulating an underlying mesh of control points. The resulting deformation controls the shape of the 3-D object and produces a smooth and continuo... |

43 |
A physics-based coordinate transformation for 3-D image matching
- Davis, Khotanzad, et al.
- 1997
(Show Context)
Citation Context ...tensor product of the familiar 1-D cubic B-splines where , , , , , and where represents the th basis function of the B-spline [22], [23] In contrast to thin-plate splines [25] or elastic-body splines =-=[26]-=-, B-splines are locally controlled, which makes them computationally efficient even for a large number of control points. In particular, the basis functions of cubic B-splines (2) (3)s714 IEEE TRANSAC... |

35 |
Multiresolution Voxel Similarity Measures for MR-PET Registration
- Studholme, Hill, et al.
- 1995
(Show Context)
Citation Context ...lity image registration (for a detailed overview seesRUECKERT et al.: NONRIGID REGISTRATION USING FREE-FORM DEFORMATION 713 [8]). In particular, voxel-based similarity measures based on joint entropy =-=[9]-=-, mutual information [10]–[13], and normalized mutual information [14], [15] have been shown to align images acquired with different imaging modalities, robustly. However, most of these approaches are... |

28 | Image metamorphosis with scattered feature constraints
- Lee, Wolberg, et al.
- 1995
(Show Context)
Citation Context ...ry significantly across patients and with age. Therefore, it is difficult to describe the local deformation via parameterized transformations. Instead, we have chosen an FFD model, based on B-splines =-=[22]-=-, [23], which is a powerful tool for modeling 3-D deformable objects and has been previously applied to the tracking and motion analysis in cardiac images [24]. The basic idea of FFD’s is to deform an... |

23 |
Automatic 3D segmentation of neuro-anatomical structures from MRI
- Collins, Evans, et al.
- 1995
(Show Context)
Citation Context ...es, robustly. However, most of these approaches are limited either to rigid or affine transformations. In contrast, many nonrigid registration algorithms based on elastic deformations, such as animal =-=[16]-=- or demons [17], rely on the assumption that the intensity of tissues between images remains constant. This is also true for nonrigid registration algorithms based on fluid deformations [18], [19]. A ... |

12 |
A three-component deformation model for image-guided surgery
- Edwards, Hill, et al.
- 1998
(Show Context)
Citation Context ...ges. The proposed combination of affine transformations and spline-based FFD’s provides a high degree of flexibility to model the motion of the breast. In contrast to physics-based deformation models =-=[30]-=-, the algorithm makes no assumptions about the elastic properties of the breast tissue. Even though physics-based deformation models might seem an attractive alternative, for example to model addition... |

10 | Automatic analysis of medical images using a deformable textbook
- Christensen, Miller, et al.
- 1995
(Show Context)
Citation Context ...as animal [16] or demons [17], rely on the assumption that the intensity of tissues between images remains constant. This is also true for nonrigid registration algorithms based on fluid deformations =-=[18]-=-, [19]. A notable exception is the registration algorithm proposed by Meyer et. al [20], which is based on a thin-plate spline deformation and uses mutual information as a voxel-based similarity measu... |

9 |
Analysis of dynamic MR breast images using a model of contrast enhancement
- Hayton, Brady, et al.
- 1997
(Show Context)
Citation Context ...n the pre- and postcontrast enhanced images remain constant. A similar approach has been suggested by Fischer et al. [6]. To overcome the problems caused by nonuniform intensity change, Hayton et al. =-=[7]-=- developed a pharmacokinetic model, which is combined with an optical-flow registration algorithm. This algorithm has been applied to the registration of twodimensional (2-D) breast MRI, but relies on... |

9 |
Automated 3D registration of MR and PET brain images by multiresolution optimization of voxel-similarity measures. Med Phys
- Studholme, DL, et al.
- 1997
(Show Context)
Citation Context ...a detailed overview seesRUECKERT et al.: NONRIGID REGISTRATION USING FREE-FORM DEFORMATION 713 [8]). In particular, voxel-based similarity measures based on joint entropy [9], mutual information [10]–=-=[13]-=-, and normalized mutual information [14], [15] have been shown to align images acquired with different imaging modalities, robustly. However, most of these approaches are limited either to rigid or af... |

8 |
Comparison and evaluation of rigid and nonrigid registration of breast MR images
- Denton, Sonoda, et al.
- 1999
(Show Context)
Citation Context ...n contrast-enhanced MRI. However, further work is needed to assess and evaluate the impact of the algorithm for the detection and diagnosis of breast cancer, which is the topic of a forthcoming paper =-=[31]-=-. Future work will involve the application of the proposed registration algorithm to data from the MRC-supported U.K. study of MRI as a method of screening women at genetic risk of breast cancer. ACKN... |

5 |
Automatic motion correction for breast MR imaging. Radiology 198
- Zuo, Jiang, et al.
- 1996
(Show Context)
Citation Context ...l respiratory and cardiac motion, complicates the estimation of the rate of uptake of contrast agent by the breast tissue. To facilitate the analysis of pre- and postcontrast enhanced MRI, Zuo et al. =-=[4]-=- proposed a registration algorithm which minimizes the ratio of variance between images. However, their algorithm is based on the assumption that the breast is only undergoing rigid motion. Kumar et a... |

4 |
Application of 3D registration for detecting lesions in magnetic resonance breast scans
- Kumar, Asmuth, et al.
- 1996
(Show Context)
Citation Context ...oposed a registration algorithm which minimizes the ratio of variance between images. However, their algorithm is based on the assumption that the breast is only undergoing rigid motion. Kumar et al. =-=[5]-=- proposed a nonrigid registration technique which uses an optical-flow type algorithm, but is based on the assumption that the intensities in the pre- and postcontrast enhanced images remain constant.... |

2 |
MR imaging of the breast
- Weinreb, Newstead
- 1995
(Show Context)
Citation Context ...pausal women with a genetic predisposition to develop breast cancer. This has led to the investigation of alternative imaging modalities, such as MRI, for the detection and diagnosis of breast cancer =-=[2]-=-. Even though MRI mammography has disadvantages, such as a low spatial resolution of around 1 mm and the need for contrast agents, it has a number of advantages, including the tomographic, and therefo... |

1 |
Local elastic motion correction
- Fischer, Otte, et al.
- 1998
(Show Context)
Citation Context ... optical-flow type algorithm, but is based on the assumption that the intensities in the pre- and postcontrast enhanced images remain constant. A similar approach has been suggested by Fischer et al. =-=[6]-=-. To overcome the problems caused by nonuniform intensity change, Hayton et al. [7] developed a pharmacokinetic model, which is combined with an optical-flow registration algorithm. This algorithm has... |