### Citations

2451 | A global geometric framework for nonlinear dimensionality reduction
- Tenenbaum, Silva, et al.
(Show Context)
Citation Context ...the atlas synthesis to the above centers. However, it has been shown that in high-dimensional space, Euclidean distance may not appropriately reflect the intrinsic similarity between different images =-=[36,37]-=-. There have been works addressing the problem in segmentation by manifold learning [36] instead of Euclidean space. But to the best of the author’s knowledge, the usage in groupwise registration has ... |

1336 |
Active shape models - their training and application
- Cootes, Taylor, et al.
- 1995
(Show Context)
Citation Context ...t work of the study is to separate the bladder from other tissues. To this end, a variety of segmentation methods can be employed. In this work, one most widely used method, Active Shape Models (ASM) =-=[32]-=-, is adopted as an example to segment bladder in the proposed framework. ASM, developed by Cootes et al. [32], is a statistical model for segmentation. The target object is expressed by a set of point... |

975 |
Principal warps: Thin-plate splines and the decomposition of deformations,”
- Bookstein
- 1989
(Show Context)
Citation Context ...ues normally either assume an initial rigid body or affine transformation, or are run after a rigid-body or affine algorithm has provided a starting estimate [14]. The thin-plate splines by Bookstein =-=[15]-=-, which comes from the availability of the point landmarks, is among the earliest transformations. Meyer et al. [16] use the interpolating thin-plate splines to accomplish the non-rigid deformation of... |

695 | Nonrigid registration using free-form deformations: application to breast MR images
- Rueckert, LI, et al.
- 1999
(Show Context)
Citation Context ... Meyer et al. [16] use the interpolating thin-plate splines to accomplish the non-rigid deformation of the points marked in the moving image. Instead of relying on thin-plate splines, Rueckert et al. =-=[17]-=- apply the B-spline to model those deformations. After that, a lot of non-linear registration methods are developed. Lately, Passera et al. [18] make comparison of parametric maps generated from trace... |

414 | Locality preserving projections.
- He, Niyogi
- 2003
(Show Context)
Citation Context ...n), a projection matrix is learned by manifold learning. There are several algorithms available for this task (e.g., LLE, LPP, Isomap [38]). In this application, Locality Preserving Projections (LPP) =-=[39]-=- is utilized because of it superior performance. The projection model can be expressed as Yi PXi, ð6Þ where P stands for the transformation matrix. After that, K-means clustering is applied to these... |

342 | Geometry images - Gu, Gortler, et al. |

340 | Imagematching as a diffusion process: an analogy with Maxwell’s demons
- Thirion
- 1998
(Show Context)
Citation Context ...ng the state-of-the-art both qualitatively and quantitatively [40]. They are the similarity transform based registration, the B-spline based registration [40], deformable registration based on Demons =-=[41]-=- and level set motion [42]. The pairwise registration is conducted by choosing one image as reference one and the otherthe B-spline based registration (e), deformable registration based on Demons (f) ... |

225 | Level-set methods: an overview and some recent results
- Fedkiw, Osher
- 2001
(Show Context)
Citation Context ...th qualitatively and quantitatively [40]. They are the similarity transform based registration, the B-spline based registration [40], deformable registration based on Demons [41] and level set motion =-=[42]-=-. The pairwise registration is conducted by choosing one image as reference one and the otherthe B-spline based registration (e), deformable registration based on Demons (f) and limprove a lot and the... |

216 | Unbiased diffeomorphic atlas construction for computational anatomy. Neuroimage 2004;23:S151–60
- Joshi, Davis, et al.
(Show Context)
Citation Context ...eously, the groupwiseent ways according to the specific problems, the selection of the reference image is mostly subjective, which bring in unavoidable bias to the following process of image analysis =-=[27]-=-. 1.2. The proposed method Existing works for registering organs such as brain, heart, and bone abound in a large amount of literatures [9]. But approaches for bladder registration are not so abundant... |

100 | Non-rigid image registration: Theory and practice - Crum, Hartkens, et al. |

87 | Data driven image models through continuous joint alignment.
- Learned-Miller
- 2006
(Show Context)
Citation Context ... work of Shattuck et al. [28], groupwise registration can be achieved by going around all possible subject combinations of pairwise registrations, which leads to heavy computation cost. Miller et al. =-=[29]-=- propose a transitionary space to the related objectives by minimizing the sum of all images. Zollei et al. [30] successfully apply Miller’s method to groupwise registration. They use the stochastic g... |

82 |
A.W.: Construction of a 3D probabilistic atlas of human cortical structures.
- Shattuck, Mirza, et al.
- 2008
(Show Context)
Citation Context ...al. [22], vast modified methods have been proposed to align all subjects simultaneously. Joshi et al. [27] propose to register all images to the constructed mean image. In the work of Shattuck et al. =-=[28]-=-, groupwise registration can be achieved by going around all possible subject combinations of pairwise registrations, which leads to heavy computation cost. Miller et al. [29] propose a transitionary ... |

66 |
Demonstraton of accuracy and clinical versatility of mutual information for automatic multimodality image fusion using affine and thin plate spline warped geometric deformations,’’ Med.
- Meyer
- 1997
(Show Context)
Citation Context ...lgorithm has provided a starting estimate [14]. The thin-plate splines by Bookstein [15], which comes from the availability of the point landmarks, is among the earliest transformations. Meyer et al. =-=[16]-=- use the interpolating thin-plate splines to accomplish the non-rigid deformation of the points marked in the moving image. Instead of relying on thin-plate splines, Rueckert et al. [17] apply the B-s... |

49 | Comparing Variations on the Active Appearance Model Algorithm
- Cootes, Kittipanya-ngam
- 2002
(Show Context)
Citation Context ...sed framework. ASM, developed by Cootes et al. [32], is a statistical model for segmentation. The target object is expressed by a set of points, which is constrained by Point Distribution Model (PDM) =-=[33]-=-. By incorporating the shape prior information from a set of training images, ASM iteratively transform to match an example of the object in a test image. However, though ASM is a popular segmentation... |

47 | Efficient population registration of 3D data
- Zollei, Learned-Miller, et al.
- 2005
(Show Context)
Citation Context ...nations of pairwise registrations, which leads to heavy computation cost. Miller et al. [29] propose a transitionary space to the related objectives by minimizing the sum of all images. Zollei et al. =-=[30]-=- successfully apply Miller’s method to groupwise registration. They use the stochastic gradient descent and affine transformation for an optimization problem. Balci et al. [31] extend Miller’s method ... |

45 | A unified informationtheoretic approach to groupwise non-rigid registration and model building.
- Twining, Cootes, et al.
- 2005
(Show Context)
Citation Context ...osed non-rigid registration will be unitized in the formwork of groupwise registration. In fact, the groupwise registration can be performed by iteratively utilizing traditional pairwise registration =-=[25,34]-=-. Several methods have been developed to achieve groupwise registration in this way. For example, Park et al. [26] propose to obtain the atlas with the closest image to the real template by performing... |

31 | Image registration methods: a survey. Image Vis. - Zitova, Flusser - 2003 |

25 | Least biased target selection in probabilistic atlas construction
- Park, PH, et al.
(Show Context)
Citation Context ...ration can be performed by iteratively utilizing traditional pairwise registration [25,34]. Several methods have been developed to achieve groupwise registration in this way. For example, Park et al. =-=[26]-=- propose to obtain the atlas with the closest image to the real template by performing all pairwise registrations. Nonetheless, this type of methods may be very time-consuming when the dataset is larg... |

25 | Free-form B-spline deformation model for groupwise registration
- SK, Golland, et al.
- 2007
(Show Context)
Citation Context ...l images. Zollei et al. [30] successfully apply Miller’s method to groupwise registration. They use the stochastic gradient descent and affine transformation for an optimization problem. Balci et al. =-=[31]-=- extend Miller’s method to non-rigid image registration. They represent the transformations of image by free-form deformations, which is based on the method ofspond to the reference one. There are man... |

20 | The Principal Axes Transformation – A Method for Image Restoration - Alpert, Bradshaw, et al. - 1990 |

19 |
Rigid body registration. In
- Ashburner, Friston
- 2003
(Show Context)
Citation Context ...odalities, such as the images between the above types. The rigid registration between the same modality generally match themselves by minimizing the mean square error (MSE) or mutual information (MI) =-=[13]-=-. As for different modalities, the aligning is more complex. Though there has been a large number of accurate and robust methods for rigid registration, in many cases it cannot satisfy the need for cl... |

13 |
Registration of MR prostate images with biomechanical modelling and nonlinear parameter estimation,”
- Alterovitz, Goldberg, et al.
- 2006
(Show Context)
Citation Context ...ith the reference one. In order to quantitatively compare the proposed method with other methods, two evaluation criterions, namely Mean Squared Error (MSE) [13] and Dice Similarity Coefficient (DSC) =-=[43]-=-, are employed. The MSE can reflect the intensity similarity between a pair of images, which is defined as MSE 1 N XN1 i 0 ðritiÞ2, ð8Þ where N stands for the total pixels in the image, ri is the... |

12 | Orientation of 3-d structures in medical images - Faber, Stokely - 1988 |

5 | Rigid medical image registration using PCA neural network,‖ - Shang, Lv, et al. - 2006 |

3 | The deformable feature map – a novel neurocomputing algorithm for adaptive plasticity in pattern analysis - Wismüller, Vietze, et al. |

3 |
Group-wise registration of large image dataset by hierarchical clustering and alignment
- Wang, Chen, et al.
(Show Context)
Citation Context ...ence image. On the other hand, with the development of imaging and storage technologies, more imaging data need to be analyzed, which demands more effective methods to register large number of images =-=[22]-=-. To this end, groupwise registration is proposed to register a set (more than double) of images simultaneously. By registering all of the images simultaneously, the groupwiseent ways according to the... |

2 |
Deformable Structure Registration of Bladder through Surface Map- ping
- Xiong, Viswanathan, et al.
(Show Context)
Citation Context ...x.doi.org/10.1016/j.neucom.2012.08.012 n Corresponding author. E-mail addresses: crabwq@opt.ac.cn (Q. Wang), zoucuiming@opt.ac.cn (C. Zou), yuany@opt.ac.cn (Y. Yuan), luhb@fmmu.edu.cn (H. Lu),therapy =-=[2,3]-=-. Registration is a basic task in image analysis. It means that two 1.1. Related workpoints between one fraction and another. But there are always changes of locations and shapes. In this case, an acc... |

2 |
State-of-the-art medical image registration methodologies: A survey. Multi modality state-of-theart medical image segmentation and registration methodologies
- Khalifa, Beache, et al.
- 2011
(Show Context)
Citation Context ...lists available at SciVerse ScienceDirect els Neurocom Neurocomputing 101 (2013) 181–189robust registration over the past years. The key point for rigidpingkun.yan@opt.ac.cn (P. Yan).the past decades =-=[9]-=-. First, according to the mechanism and the used methodology, existing methods can be generally categorized into rigid registration and non-rigid registration. Rigid registration has been developed fo... |

2 | Groupwise registration by hierarchical anatomical correspondence detection
- Wu, Wang, et al.
- 2010
(Show Context)
Citation Context ... to be developed. To this end, Joshi et al. [27] develop to construct the group mean image first and then registering all images to it, which can improve the computational efficiency. Also, Wu et al. =-=[35]-=- propose a hierarchical groupwise registration framework, which breaks down a large-scale groupwise registration problem into a set of small-scale ones. To do this, it first clusters all input images ... |

2 | Segmenting images by combining selected atlases on manifold
- Cao, Yuan, et al.
- 2011
(Show Context)
Citation Context ...the atlas synthesis to the above centers. However, it has been shown that in high-dimensional space, Euclidean distance may not appropriately reflect the intrinsic similarity between different images =-=[36,37]-=-. There have been works addressing the problem in segmentation by manifold learning [36] instead of Euclidean space. But to the best of the author’s knowledge, the usage in groupwise registration has ... |

1 | Registration for 3-d point cloud using angularinvariant feature, Neurocomputing 72 - Jiang, Cheng, et al. - 2009 |

1 | Three-dimensional surface registration: a neural network strategy, Neurocomputing 70 - Liu, Yan, et al. - 2006 |

1 |
A non-linear registration method for DCE-MRI and DCE-CT comparison in bladder tumors
- Passera, Mainardi, et al.
- 1998
(Show Context)
Citation Context ...stead of relying on thin-plate splines, Rueckert et al. [17] apply the B-spline to model those deformations. After that, a lot of non-linear registration methods are developed. Lately, Passera et al. =-=[18]-=- make comparison of parametric maps generated from tracer kinetic modeling between Dynamic Contrast Enhanced Computed X-ray Tomography (DCE-CT) and Dynamic Contrast-enhanced Magnetic Resonance Imaging... |

1 |
Algorithms for manifold learning, UCSD tech report
- Cayton
(Show Context)
Citation Context ...rder to reduce the m-dimensional vector Xi to n-dimensional Yi (m4n), a projection matrix is learned by manifold learning. There are several algorithms available for this task (e.g., LLE, LPP, Isomap =-=[38]-=-). In this application, Locality Preserving Projections (LPP) [39] is utilized because of it superior performance. The projection model can be expressed as Yi PXi, ð6Þ where P stands for the transfo... |

1 |
397–400.Key Laboratory of Transient Optics and Photonics, Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences
- Seghers, D’Agostino, et al.
(Show Context)
Citation Context ...osed non-rigid registration will be unitized in the formwork of groupwise registration. In fact, the groupwise registration can be performed by iteratively utilizing traditional pairwise registration =-=[25,34]-=-. Several methods have been developed to achieve groupwise registration in this way. For example, Park et al. [26] propose to obtain the atlas with the closest image to the real template by performing... |