S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Trans. Med. Imag., vol. 13, pp. 122--132, 1994.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....deformation of the ventricle. In some cases, the tracking can be helped by adding hard constraints when there is a priori knowledge of the deformation at some reliable anchor points. This is the case in tagged images, where some tissues are marked and can be tracked after a simple preprocessing [2, 24, 44, 30, 27]. We hope to apply our work to this recent technique of imaging in a the near future. Figure 22: Twist component estimation on a synthetic example. Top: from original data. From left to right and from top to bottom: from models obtained using strategies 1, 2, 3 and 4. Figure 23: Trajectories of ....

S. Kumar and D. Goldgof. Automatic Tracking of SPAMM Grid and the Estimation of Deformation Parameters from Cardiac MR Images. In IEEE Transactions on Medical Imaging, pages 122132, March 1994.

....deformation of the ventricle. In some cases, the tracking can be helped by adding hard constraints when there is a priori knowledge of the deformation at some reliable anchor points. This is the case in tagged images, where some tissues are marked and can be tracked after a simple preprocessing [2, 24, 44, 30, 27]. We hope to apply our work to this recent technique of imaging in a the near future. Acknowledgements We would like to thank Gr#goire Malandain who contributed to the segmentation of the images presented in this article; and J#r#me Declerck, Serge Benayoun and Alexis Gourdon for constructive ....

S. Kumar and D. Goldgof. Automatic Tracking of SPAMM Grid and the Estimation of Deformation Parameters from Cardiac MR Images. In IEEE Transactions on Medical Imaging, pages 122132, March 1994.

....polynomial is not known. Young and Axel [3] proposed fitting a finite element model to similar displacement measurements from grid tagged images. The finite element mesh in [3] contains only 16 elements resulting in a fairly coarse resolution estimate of the displacement field. Kumar and Goldgof [4] proposed fitting a thin plate spline to intersection correspondences in grid tagged images. Denney and Prince [5] proposed an estimation theoretic approach to the interpolation problem. This approach uses a multidimensional stochastic vector field model for the displacement field and the Fisher ....

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Transactions on Medical Imaging, 13(1):122--132, 1994.

....contours. Guttman, et al. 4] used a template matching approach along with a least squares error criterion. Contours are used in this approach to restrict the domain over which the template match is performed. Young, et al. 6] Amini, et al. 9] Radeva, et al. 14] and Kumar and Goldgof [15] used snakes to identify tag lines based on image intensity and spatial continuity constraints. Contours are used in these approaches to break the continuity constraints at the epicardial and endocardial boundaries. In [5] Guttman, et al. proposed a tag identification algorithm that does not ....

....are a topic of future research. The ML MAP algorithm contains a tag center estimation algorithm that uses a Gaussian tag profile to generate image forces for a snake algorithm. This approach is similar in spirit to snake based tag estimation algorithms proposed by other researchers [5,6,15,9]. The ML MAP algorithm, however, differs from these other snake based algorithms in the following ways. First, the maximum likelihood (ML) estimation framework provides a set of spatially varying weights for comparing the image data with the Gaussian tag template based on the physics of the ....

S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Transactions on Medical Imaging, vol. 13, no. 1, pp. 122--132, 1994.

....11] detection of tag stripes is performed by graph search and tag profile fitting. In order to determine the tag locations, tag profiles are simulated as a function of time using physics of MRI. The profile fitting approach has been improved by utilizing a template matching procedure in [2, 1] In [2, 12, 21], authors have adopted a 2D snake technique for different image slices to recover within slice tag motion. Their approach aims to minimize an external energy which is the sum of intensities for each slice, together with an internal energy which provides smoothness. As pointed out in [12] ....

....In [2, 12, 21] authors have adopted a 2D snake technique for different image slices to recover within slice tag motion. Their approach aims to minimize an external energy which is the sum of intensities for each slice, together with an internal energy which provides smoothness. As pointed out in [12], detection of tag data with varying contrast needs spatially varying parameters for snakes, making automated localization non trivial. There exists another class of approaches to analysis of LV motion based on optical flow analysis of tagged MRI [4, 17, 7] Our goal in this work is to develop a ....

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Transactions on Medical Imaging, 13(1):122--132, 1994.

....such as strain can be computed. Young and Axel [6] proposed fitting a finite element model to similar displacement measurements obtained from grid tagged images. Their finite element mesh contains only 16 elements, which results in a coarse estimate of the displacement field. Kumar and Goldgof [7] proposed fitting a thin plate spline to intersection correspondences from similar images. Denney and Prince [8] proposed an estimation theoretic approach to this interpolation problem. Their approach uses a multi dimensional stochastic vector field model for the displacement field and the Fischer ....

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Transactions on Medical Imaging, 13(1):122--132, 1994.

....applications like model based image compression [TH94] or medical diagnosis [Aya93] In medical imaging processing such sequences are important since they provide both anatomical and physiological information. A particularly important application concerns the analysis of the cardiac deformations [AD92, SD92, HG93b, CHA94, KG94]. Three dimensional (3D) medical imagery allows to recover the complete dynamic of the left ventricle wall. Furthermore, the time resolution permits a precise study of the heart beats, not limited to diastole and systole. So it is now conceivable to compute accurately the left ventricle wall ....

S. Kumar and D. Goldgof. Automatic tracking of spamm grid and the estimation of deformation parameters from cardiac mr images. IEEE Transactions on Medical Imaging, 13(1):122--132, March 1994.

....ventricle (LV) Guttman, et al. 6] used a template matching approach along with a leastsquares error criterion. Contours are used in this approach to restrict the domain over which the template match is performed. Young, et al. 5] Amini, et al. 7] Radeva, et al. 8] and Kumar and Goldgof [9] used snakes to identify tag lines based on image intensity and spatial continuity constraints. Contours are used in these approaches to break the continuity constraints at the epicardial and endocardial boundaries. Contour identification, however, is a time consuming process and requires a ....

....occupying the same physical position. For these reasons, we optimize Equation (6) for each tag center subject to spatial continuity constraints and a constraint on tag separation [11] The resulting algorithm is similar to the snake based tag estimation algorithms proposed by other researchers [11, 5, 9, 7]. The algorithm in this paper differs from these other snake based algorithms in that 1) we use a snake force based on the tag center likelihood function derived in the previous section and 2) the myocardium is segmented in the ROI by using a maximum a posteriori (MAP) hypothesis test. We denote ....

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Transactions on Medical Imaging, 13(1):122--132, 1994.

....as illustrated in Fig. 4. Tracking tag lines and measuring this component of motion therefore requires two goals to be met: identifying points of minimal brightness and determining m. Methods for accomplishing these goals can be divided into template matching [19, 20, 21] and active geometry [7, 22, 23] techniques, both of which we describe here. Template Matching In template matching, which is illustrated in Fig. 5, a point on a tag line is found by comparing the image brightness along a strip of pixels to an expected tag pattern template. The location at which the template and the measured ....

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Trans. Med. Imag., 13:122--132, 1993.

....deformation of the ventricle. In some cases, the tracking can be helped by adding hard constraints when there is a priori knowledge of the deformation at some reliable anchor points. This is the case in tagged images, where some tissues are marked and can be tracked after a simple preprocessing [2, 24, 44, 30, 27]. We hope to apply our work to this recent technique of imaging in a the near future. Analyzing heart deformation submitted to MEDIA 37 Figure 22: Twist component estimation on a synthetic example. Top: from original data. From left to right and from top to bottom: from models obtained using ....

S. Kumar and D. Goldgof. Automatic Tracking of SPAMM Grid and the Estimation of Deformation Parameters from Cardiac MR Images. In IEEE Transactions on Medical Imaging, pages 122132, March 1994.

....graph search and tag profile fitting. In order to determine the tag locations, tag profiles are simulated as a function of time using physics of MRI. The profile fitting approach has been improved by utilizing a template matching procedure in [1, 2, 3] or by additional local analysis [21, 13] In [1, 34, 17], authors have adopted a 2D snake analysis system for different image slices in order to recover within slice tag motion. Their approach aims to minimize an external energy which is the sum of intensities for each slice, together with an internal energy which provides smoothness. As pointed out in ....

....authors have adopted a 2D snake analysis system for different image slices in order to recover within slice tag motion. Their approach aims to minimize an external energy which is the sum of intensities for each slice, together with an internal energy which provides smoothness. As pointed out in [17], detection of tag data with varying contrast needs spatially varying parameters for snakes, making automated localization non trivial. Other approaches to capture heart motion is based on optical flow analysis of tagged MRI [28, 12] To provide satisfactory results knowledge is required about ....

S. Kumar and D. Goldgof, "Automatic Tracking of SPAMM Grid and the Estimation of Deformation Parameters from Cardiac MR Images", IEEE Trans. Med. Imaging, Vol. 13, 1994.

....a stack of 2D images for each time frame. Guttman, et al. 71, 72] proposed a semi automated algorithm for detecting and tracking endocardial and epicardial contours and tag lines from planar tagged MR images based on differential edge detection and template matching techniques. Kumar and Goldgof [73] proposed an automated method for tracking tag lines in grid tagged images based on the use of snakes [74, 75] which track lines by minimizing an energy function containing penalty terms for continuity, curvature and image intensity. Once the features have been identified and tracked, some type ....

....in later images, as shown in Figure 4.1. Points along these tag lines 123 (a) b) Figure 4.1: a) Tagged LV shortly after end diastole. b) Tagged LV at end systole. can be identified with a semi automated tracking algorithm such as that proposed by Guttman et al. 72] and Kumar and Goldgof [73]. We show in Section 4.2.2 that by comparing the location of a point on a deformed tag line to the reference location of the tag plane, we can measure the displacement of that point in the direction normal to the reference tag plane. To measure displacements in other directions, tag planes are ....

[Article contains additional citation context not shown here]

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Transactions on Medical Imaging, 13(1):122--132, 1994.

....using implanted beads in the myocardium, which has provided clinical information about regional myocardial strain [5] However the use of these invasive markers is limited. Alternatively, tissue tagging with MRI makes it possible to embed in the image a large number of material landmarks [6] 7] [8], 9] 10] 11] These non invasive landmarks move along with the tissue. During the second stage the correspondence between landmarks is established over time. The correspondence can be achieved manually [10] or using automatic procedures that exploit the spatial organization of the landmarks ....

....[9] 10] 11] These non invasive landmarks move along with the tissue. During the second stage the correspondence between landmarks is established over time. The correspondence can be achieved manually [10] or using automatic procedures that exploit the spatial organization of the landmarks [8], or the continuity of motion. Finally, the last stage consists of recovering the motion of the object from the sparse set of landmarks trajectories. Unfortunately, without further information, the problem is underconstrained. The landmarks trajectories provide only a sparse sampling of the ....

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac images. IEEE Trans. on Medical Imaging, Vol. 13, No.1:122--132, March 1994.

....these tag planes show the tags as dark lines which are nearly straight in images taken shortly after end diastole and are curved in later images. Points along these tag lines can be identified with a semi automated tracking algorithm such as that proposed by Guttman et al. 3] Kumar and Goldgof [4], or Young et al. 5] Because of the possibility of motion par This research was supported by Whitaker Foundation Biomedical Engineering Research Grant 91 0108, NIH grant R01 HL45090, and National Science FoundationPresidential Faculty Fellow Award MIP9350336. T. S. Denney Jr is with the ....

....One difficulty with this approach is that the correct order of the interpolating polynomial is not known. Young and Axel [9] proposed fitting a finite element model to displacement measurements from grid tagged images using heuristically chosen interpolation functions. Kumar and Goldgof [4] proposed a method for estimating 2 D LV motion from grid tagged images using a thin plate spline interpolating function. In this paper, we develop a framework for computing a dense displacement field from sparse measurements by formulating the problem as a stochastic estimation problem. A priori ....

[Article contains additional citation context not shown here]

S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images", IEEE Transactions on Medical Imaging, vol. 13, no. 1, pp. 122--132, 1994.

....scheme has the advantage of being easy to implement. After a regular mesh is defined (usually quadrilateral) a simple postprocessing can be performed to reshape the mesh to match feature points. This scheme is useful for the image where objects can be tagged by the invasive or noninvasive markers [4] [10] The obvious drawback is that not all images can be attached with artificial markers. In addition, the initial mesh can not be stretched to a degree that original mesh structure is destroyed, and thus no longer suitable for the nonrigid motion tracking. Figure 3 (c) shows an adaptive ....

S. Kumar and D. Goldgof, "Automatic Tracking of SPAMM Grid and the Estimation of Deformation Parameters from Cardiac MR Images", IEEE Engineering on Medical Imaging, 13(1), pp. 122132, 1994.

No context found.

S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Trans. Med. Imag., vol. 13, pp. 122--132, 1994.

No context found.

S. Kumar and D. Goldgof. Automatic Tracking of SPAMM Grid and the Estimation of Deformation Parameters from Cardiac MR Images. In IEEE Transactions on Medical Imaging, pages 122132, March 1994.

No context found.

Kumar and S. D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Trans. Med. Imag., vol. 13, pp. 122--132, Mar. 1994.

No context found.

S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Transactions on Medical Imaging, vol. 13, pp. 122--132, Mar 1994.

No context found.

S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Trans. Med. Imaging, 13:122--132, 1994.

No context found.

S. Kumar and D. Goldgof, "Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images," IEEE Trans. Med. Imag., vol. 13, pp. 122--132, 1994. 172 Image Segmentation Using Deformable Models

No context found.

S. Kumar and D. Goldgof. Automatic tracking of SPAMM grid and the estimation of deformation parameters from cardiac MR images. IEEE Transactions on Medical Imaging, 13(1):122--132, 1994.

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