A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3-D neuroradiological data sets," IEEE Trans. Med. Imag., vol. 18, no. 10, pp. 828--839, Oct. 1999.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....(see figure 1) In practice, correspondence has often been established using manually defined landmarks ; this is both time consuming and subjective. The problems are exacerbated when the approach is applied to 3D images. Several previous attempts have been made to automate model building [1,11,12,14,15,22,25]. The problem of establishing dense correspondence over a set of training bound aries can be posed as that of defining the parameterisation and alignment of each shape in the training set, leading to a dense correspondence between equivalently parameterised boundary points. Arbitrary ....

....over a set of training bound aries can be posed as that of defining the parameterisation and alignment of each shape in the training set, leading to a dense correspondence between equivalently parameterised boundary points. Arbitrary parameterisations of the training boundaries have been proposed [1,14] , but these do not guarantee correct correspondences. Shape features (e.g. regions of high curvature) have been used to establish point correspondences, with boundary length interpo lation between these points [2,13,24,25] Although this approach corresponds with human intuition, it is still ....

[Article contains additional citation context not shown here]

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3- D neuroradiological data sets". IEEE Transactions On Medical Imaging, 18(10): p. 828-839. 1999.

....true for the 3D case, since more slices have to be analyzed and the number of landmark points that is necessary to describe the shape increases dramatically. Several authors have proposed automatic procedures for finding correspond ing landmark points based on segmented images. Kelemen et al. [2] find corre sponding points by parameterisation of the training shape boundaries. Wang et This work is funded by Philips Research Laboratories, Hamburg, Germany. al. 3] identify the landmarks using curvature information. Luo and O Donnell [4] have manufactured a dense set of landmarks using ....

A. Kelemen, G. Szkely, and G. Gerig, "Elastic model-based segmentation of 3-d neuroradiological data sets," IEEE Transactions on Medical Imaging, vol. 18, no. 10, pp. 828-839, 1999.

....ester, M13 9PT, UK. e maih rhodri.h.davies stud.man.ac.uk) J.C. Waterton is with AstraZeneca, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK. are exacerbated when the approach is applied to 3D im ages. Several previous attempts have been made to automate model building [3] 4] 5] [6], 7] 8] 9] The problem of establishing dense correspondence over a set of training boundaries can be posed as that of defining a parameterization for each of the training set, leading to a dense correspondence between equivalently parameterized boundary points. Arbitrary parameterizations of ....

....correspondence over a set of training boundaries can be posed as that of defining a parameterization for each of the training set, leading to a dense correspondence between equivalently parameterized boundary points. Arbitrary parameterizations of the training bound aries have been proposed [3] [6] , but these fail to address the issue of optimality. Shape features (e.g. regions of high curvature) have been used to establish point corre spondences, with boundary length interpolation between these points It0] tt] 12] 9] Although this approach corresponds with human intuition, it is ....

[Article contains additional citation context not shown here]

A. Kelemen G. Szekely, and G. Gerig, "Elastic model-based seg- mentation of 3-D neuroradiologicM data sets", IEEE Transactions On Medical Imaging, vol. 18, no. 10, pp. 828-839, 1999.

....when the internal object is relatively small and its shape is highly variable, as is the case for the hippocampus, this approach may not generate an accurate result, due to its sensitivity to the imperfections involved with the registration and warping steps. Alternatively, a deformable model [Kelemen 1999] has been proposed for the segmentation of the hippocampus. This method uses shape and boundary profile information to guide the segmentation procedure of hippocampus. Semi automatic methods may provide a more realistic approach for hippocampal segmentation because they combine the automatic ....

A. Kelemen, G. Szekely, and G. Gerig. "Elastic model-based segmentation of 3-D neuroradiological data sets". IEEE Trans. on MedicalImaging, 18(10): 828-839, October 1999.

....properties of the underlying growth field were examined via principal component analysis, and the principal eigenvector that reflected a large percentage of the variation of the growth field was used to predict the direction of growth. Related are a number of statistical shape models [27] [30] that use principal component analysis in the context of segmentation and template matching. These models are not concerned with prediction per se, but rather with modeling normal anatomical variability for the scope of shape segmentation and quantification. In this paper, we describe a framework ....

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3-D neuroradiological data sets," IEEE Trans. Med. Imag., vol. 18, pp. 828--839, Oct. 1999.

....user effort when they are used to edit target structures that are surrounded by structures with a similar intensity range. Many algorithms for segmentation of medical images use a priori knowledge in the form of probabilistic or geometric models to guide the segmentation process [3] 8] [12]. Typically, these models are derived through an automated supervised learning process [13] that produces a general model from training samples that an expert defines. In the past decade, researchers introduced a variety of deformable models [14] 17] which minimize energy functions to find the ....

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3-D neuroradiological data sets," IEEE Trans. Med. Imag., vol. 18, pp. 828--839, Oct. 1999.

No context found.

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3d neuroradiological data sets," IEEE Trans. Med. Imaging, vol. 18, pp. 828--839, October 1999.

No context found.

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3d neuroradiological data sets," IEEE Transactions on Medical Imaging 18, pp. 828--839, October 1999.

....spherical harmonics (SPHARM) and a medial manifold description (skeletonization) using a sampled medial representation (M rep) Both representations can describe the same objects but are fundamentally different in the way they represent the object shapes. The SPHARM description (see Brechb uhler [1, 2]) is a global, fine scale description that represents shapes of spherical topology. The basis functions of the parameterized surface are spherical harmonics. SPHARM is a smooth, accurate surface shape representation given a sufficiently small approximation error (see Fig. 1a) Based on a uniform ....

....sphere and optimized for a homogeneous distribution of parameters. The parametrization of single objects has been later combined with the concept of statistical shape models proposed by Cootes et al. 7, 8] but using the coefficients of the parametrization rather than a point distribution model [9, 2]. 2.1.1. Spherical harmonics descriptors Spherical harmonic basis functions Y l , l # l of degree l and order m are defined on # [ 0; # ] 0; 2#) by the following definitions (see Fig. 2 left for a visualization of the basis function) Y l (#, #) 2l 1 4# (l m) l ....

[Article contains additional citation context not shown here]

G. Kelemen, A.and Sz ekely and G. Gerig, "Elastic model-based segmentation of 3d neuroradiological data sets," IEEE Trans. Med. Imaging, vol. 18, pp. 828--839, October 1999.

....as point wise changes in size, can be easily interpreted locally [5, 6] A difficulty, however, is the inability of deformation fields to capture subtle changes that might be related to the various scales at which the object s geometric features are manifested. The approach taken by Kelemen [7] evaluates a pop1 ulation of 3D hippocampal shapes based on a boundary description by spherical harmonic basis functions (SPHARM) which was proposed by Brechb uhler [8, 9] The SPHARM shape description delivers an implicit correspondence between shapes on the boundary, which is used in the ....

....shapes. By holding the topology of the model fixed, an implicit correspondence between shapes is given and statistical shape analysis can directlt be applied. This paper applies a technique originally developed for model based segmentation, the SPHARM shape representation of object surfaces [8, 7] , to analyze brain structures. In particular, we address the clinical research problem of studying similarity of brain structures in idential (monozygotic, MZ) and non identical (dizygotic, DZ) twin pairs. The paper is organized as follows. First we discuss the SPHARM description and its use for ....

[Article contains additional citation context not shown here]

G. Kelemen, A.and Sz ekely and G. Gerig, "Elastic model-based segmentation of 3d neuroradiological data sets," IEEE Trans. Med. Imaging, vol. 18, pp. 828--839, October 1999.

....graph called medial graph with edges representing either inter or intra figural links. SPHARM The SPHARM description is a parametric surface description that can only represent objects of spherical topology [4] The basis functions of the parameterized surface are spherical harmonics. Kelemen [14] demonstrated that SPHARM can be used to express shape deformations. SPHARM is a smooth, accurate fine scale shape representation, given a sufficiently small approximation error. Based on a uniform icosahedron subdivision of the spherical parameterization, we obtain a Point Distribution Model ....

....the shape variability in the training population, thus making the computations of our scheme more stable. We assume that the shape space is an appropriate representation of the object s biological variability. PCA is applied to SPHARM objects #c i of the training population as described by Kelemen [14]. The PCA results in an average coefficient vector # c and the eigenmodes of deformation (# 1 , #v 1 ) # n 1 , #v n 1 ) The first k eigenmodes (# 1 , #v 1 ) # k , #v k ) are chosen to cover at least 95 of the population s variability. A discrete description of the shape ....

A. Kelemen, G. Sz ekely, and G. Gerig, "Elastic model-based segmentation of 3d neuroradiological data sets," IEEE Trans. Med. Imaging, vol. 18, pp. 828--839, October 1999.

....only statistical significant but also neuroanatomically relevant and intuitive. 2. Shape Modeling Surface based shape representation: We applied a technique for surface parametrization that uses expansion into a series of spherical harmonics (SPHARM) Brechbuehler, 1995, Szekely et al. 1996, Kelemen, 1999]. The development parallels the seminal work of Cootes Taylor [Cootes et al. 1995] on active shape models but is based on a parametric object description (inspired by Staib and Duncan [Staib, 1996] rather than a point distribution model. SPHARM is a global parametrization method, i.e. small ....

....studies. 3.1 Statistical analysis of amygdala hippocampal asymmetry in schizophrenia. We studied the asymmetry of the hippocampal complex for a group of 15 controls and 15 schizophrenics (collaboration with M. E. Shenton, Harvard) Asymmetry was assessed by segmentation using deformable models (Kelemen, 1999), by flipping one object across the midsagittal plane, by aligning the reference and the mirrored object using the coordinates of the first ellipsoid, and by calculating the MSD between the two surfaces (Fig.1) Fig. 1. Analysis of amygdalahippocampal left right asymmetry. The left hippocampal ....

[Article contains additional citation context not shown here]

Kelemen, A., Szekely, G. and Gerig, G. (1999) Elastic Model-Based Segmentation of 3D Neuroradiological Data Sets, IEEE Transactions On Medical Imaging, 18, 828-839.

No context found.

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3-D neuroradiological data sets," IEEE Trans. Med. Imag., vol. 18, no. 10, pp. 828--839, Oct. 1999.

No context found.

A. Kelemen, G. Szkely, and G. Gerig, "Elastic model-based segmentation of 3-D neuroradiological data sets," IEEE Trans. Med. Imag., vol. 18, pp. 828--839, Oct. 1999.

No context found.

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3d neuroradiological data sets," IEEE Transactions on Medical Imaging 18, pp. 828--839, October 1999.

No context found.

A. Kelemen, G. Szekely, and G. Gerig, "Elastic model-based segmentation of 3-D neuroradiological data sets," IEEE Transactions on Medical Imaging, vol. 18, no. 10, pp. 828--839, 1999.

No context found.

Kelemen, A, G Szekely and G Gerig (1999). Elastic Model-Based Segmentation of 3D Neuroradiological Data Sets. IEEE Transactions On Medical Imaging 18: 828-839.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC