[Show abstract][Hide abstract] ABSTRACT: Quantitative analysis of cardiac Magnetic Resonance (CMR) images requires accurate segmentation of myocardium. Although recent multi-atlas segmentation approaches have done a good job improving segmentation accuracy, they also increase the computational burden, which degrades their clinical utility. In this paper, we proposed a novel multi-atlas segmentation framework using an augmented atlas technique that is able to increase segmentation accuracy without increasing computational complexity. This is achieved by using roughly aligned neighborhood slices to improve patch-based label fusion accuracy. We evaluated the proposed approach on the MICCAI SATA Segmentation Challenge CAP dataset. Our results demonstrate that the proposed technique can achieve segmentation accuracy comparable to the state-of-the-art algorithms in much smaller amount of time.
Biomedical Imaging (ISBI), 2015 IEEE 12th International Symposium on, New York; 04/2015
[Show abstract][Hide abstract] ABSTRACT: 3D echocardiographic (3DE) imaging is a useful tool for assessing the complex geometry of the aortic valve apparatus. Segmentation of this structure in 3DE images is a challenging task that benefits from shape-guided deformable modeling methods, which enable inter-subject statistical shape comparison. Prior work demonstrates the efficacy of using continuous medial representation (cm-rep) as a shape descriptor for valve leaflets. However, its application to the entire aortic valve apparatus is limited since the structure has a branching medial geometry that cannot be explicitly parameterized in the original cm-rep framework. In this work, we show that the aortic valve apparatus can be accurately segmented using a new branching medial modeling paradigm. The segmentation method achieves a mean boundary displacement of 0.6 ± 0.1 mm (approximately one voxel) relative to manual segmentation on 11 3DE images of normal open aortic valves. This study demonstrates a promising approach for quantitative 3DE analysis of aortic valve morphology.
[Show abstract][Hide abstract] ABSTRACT: Comprehensive visual and quantitative analysis of in vivo human mitral valve morphology is central to the diagnosis and surgical treatment of mitral valve disease. Real-time 3D transesophageal echocardiography (3D TEE) is a practical, highly informative imaging modality for examining the mitral valve in a clinical setting. To facilitate visual and quantitative 3D TEE image analysis, we describe a fully automated method for segmenting the mitral leaflets in 3D TEE image data. The algorithm integrates complementary probabilistic segmentation and shape modeling techniques (multi-atlas joint label fusion and deformable modeling with continuous medial representation) to automatically generate 3D geometric models of the mitral leaflets from 3D TEE image data. These models are unique in that they establish a shape-based coordinate system on the valves of different subjects and represent the leaflets volumetrically, as structures with locally varying thickness. In this work, expert image analysis is the gold standard for evaluating automatic segmentation. Without any user interaction, we demonstrate that the automatic segmentation method accurately captures patient-specific leaflet geometry at both systole and diastole in 3D TEE data acquired from a mixed population of subjects with normal valve morphology and mitral valve disease.
Medical image analysis 10/2013; 18(1):118-129. DOI:10.1016/j.media.2013.10.001 · 3.65 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The basis of mitral annuloplasty ring design has progressed from qualitative surgical intuition to experimental and theoretical analysis of annular geometry with quantitative imaging techniques. In this work, we present an automated three-dimensional (3D) echocardiographic image analysis method that can be used to statistically assess variability in normal mitral annular geometry to support advancement in annuloplasty ring design.
Three-dimensional patient-specific models of the mitral annulus were automatically generated from 3D echocardiographic images acquired from subjects with normal mitral valve structure and function. Geometric annular measurements including annular circumference, annular height, septolateral diameter, intercommissural width, and the annular height to intercommissural width ratio were automatically calculated. A mean 3D annular contour was computed, and principal component analysis was used to evaluate variability in normal annular shape.
The following mean ± standard deviations were obtained from 3D echocardiographic image analysis: annular circumference, 107.0 ± 14.6 mm; annular height, 7.6 ± 2.8 mm; septolateral diameter, 28.5 ± 3.7 mm; intercommissural width, 33.0 ± 5.3 mm; and annular height to intercommissural width ratio, 22.7% ± 6.9%. Principal component analysis indicated that shape variability was primarily related to overall annular size, with more subtle variation in the skewness and height of the anterior annular peak, independent of annular diameter.
Patient-specific 3D echocardiographic-based modeling of the human mitral valve enables statistical analysis of physiologically normal mitral annular geometry. The tool can potentially lead to the development of a new generation of annuloplasty rings that restore the diseased mitral valve annulus back to a truly normal geometry.
The Annals of thoracic surgery 10/2013; 97(1). DOI:10.1016/j.athoracsur.2013.07.096 · 3.85 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We propose a new approach for statistical shape analysis of 3D anatomical objects based on features extracted from skeletons. Like prior work on medial representations, the approach involves deforming a template to target shapes in a way that preserves the branching structure of the skeleton and provides intersubject correspondence. However, unlike medial representations, which parameterize the skeleton surfaces explicitly, our representation is boundary-centric, and the skeleton is implicit. Similar to prior constrained modeling methods developed 2D objects or tube-like 3D objects, we impose symmetry constraints on tuples of boundary points in a way that guarantees the preservation of the skeleton's topology under deformation. Once discretized, the problem of deforming a template to a target shape is formulated as a quadratically constrained quadratic programming problem. The new technique is evaluated in terms of its ability to capture the shape of the corpus callosum tract extracted from diffusion-weighted MRI.
Information processing in medical imaging: proceedings of the ... conference 01/2013; 23:280-91. DOI:10.1007/978-3-642-38868-2_24
[Show abstract][Hide abstract] ABSTRACT: Here, we describe a novel method for volumetric segmentation of the amygdala from MRI images collected from 35 human subjects. This approach is adapted from open-source techniques employed previously with the hippocampus (Suh et al., 2011; Wang et al., 2011a,b). Using multi-atlas segmentation and machine learning-based correction, we were able to produce automated amygdala segments with high Dice (Mean = 0.918 for the left amygdala; 0.916 for the right amygdala) and Jaccard coefficients (Mean = 0.850 for the left; 0.846 for the right) compared to rigorously hand-traced volumes. This automated routine also produced amygdala segments with high intra-class correlations (consistency = 0.830, absolute agreement = 0.819 for the left; consistency = 0.786, absolute agreement = 0.783 for the right) and bivariate (r = 0.831 for the left; r = 0.797 for the right) compared to hand-drawn amygdala. Our results are discussed in relation to other cutting-edge segmentation techniques, as well as commonly available approaches to amygdala segmentation (e.g., Freesurfer). We believe this new technique has broad application to research with large sample sizes for which amygdala quantification might be needed.
Frontiers in Neuroscience 11/2012; 6:166. DOI:10.3389/fnins.2012.00166 · 3.66 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To evaluate the distribution of white matter (WM) disease in frontotemporal lobar degeneration (FTLD) and Alzheimer disease (AD) and to evaluate the relative usefulness of WM and gray matter (GM) for distinguishing these conditions in vivo.
Patients were classified as having FTLD (n = 50) or AD (n = 42) using autopsy-validated CSF values of total-tau:β-amyloid (t-tau:Aβ(1-42)) ratios. Patients underwent WM diffusion tensor imaging (DTI) and volumetric MRI of GM. We employed tract-specific analyses of WM fractional anisotropy (FA) and whole-brain GM density analyses. Individual patient classification was performed using receiver operator characteristic (ROC) curves with FA, GM, and a combination of the 2 modalities.
Regional FA and GM were significantly reduced in FTLD and AD relative to healthy seniors. Direct comparisons revealed significantly reduced FA in the corpus callosum in FTLD relative to AD. GM analyses revealed reductions in anterior temporal cortex for FTLD relative to AD, and in posterior cingulate and precuneus for AD relative to FTLD. ROC curves revealed that a multimodal combination of WM and GM provide optimal classification (area under the curve = 0.938), with 87% sensitivity and 83% specificity.
FTLD and AD have significant WM and GM defects. A combination of DTI and volumetric MRI modalities provides a quantitative method for distinguishing FTLD and AD in vivo.
[Show abstract][Hide abstract] ABSTRACT: In vivo human mitral valves (MV) were imaged using real-time 3D transesophageal echocardiography (rt-3DTEE), and volumetric images of the MV at mid-systole were analyzed by user-initialized segmentation and 3D deformable modeling with continuous medial representation, a compact representation of shape. The resulting MV models were loaded with physiologic pressures using finite element analysis (FEA). We present the regional leaflet stress distributions predicted in normal and diseased (regurgitant) MVs. Rt-3DTEE, semi-automated leaflet segmentation, 3D deformable modeling, and FEA modeling of the in vivo human MV is tenable and useful for evaluation of MV pathology.
Journal of Biomechanics 03/2012; 45(5):903-7. DOI:10.1016/j.jbiomech.2011.11.033 · 2.75 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Purpose: Patient-specific shape analysis of the mitral valve from
real-time 3D ultrasound (rt-3DUS) has broad application to the
assessment and surgical treatment of mitral valve disease. Our goal is
to demonstrate that continuous medial representation (cm-rep) is an
accurate valve shape representation that can be used for statistical
shape modeling over the cardiac cycle from rt-3DUS images. Methods:
Transesophageal rt-3DUS data acquired from 15 subjects with a range of
mitral valve pathology were analyzed. User-initialized segmentation with
level sets and symmetric diffeomorphic normalization delineated the
mitral leaflets at each time point in the rt-3DUS data series. A
deformable cm-rep was fitted to each segmented image of the mitral
leaflets in the time series, producing a 4D parametric representation of
valve shape in a single cardiac cycle. Model fitting accuracy was
evaluated by the Dice overlap, and shape interpolation and principal
component analysis (PCA) of 4D valve shape were performed. Results: Of
the 289 3D images analyzed, the average Dice overlap between each fitted
cm-rep and its target segmentation was 0.880+/-0.018 (max=0.912,
min=0.819). The results of PCA represented variability in valve
morphology and localized leaflet thickness across subjects. Conclusion:
Deformable medial modeling accurately captures valve geometry in rt-3DUS
images over the entire cardiac cycle and enables statistical shape
analysis of the mitral valve.
Proceedings of SPIE - The International Society for Optical Engineering 02/2012; 8320:8-. DOI:10.1117/12.910708 · 0.20 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Precise 3D modeling of the mitral valve has the potential to improve our understanding of valve morphology, particularly in the setting of mitral regurgitation (MR). Toward this goal, the authors have developed a user-initialized algorithm for reconstructing valve geometry from transesophageal 3D ultrasound (3D US) image data.
Semi-automated image analysis was performed on transesophageal 3D US images obtained from 14 subjects with MR ranging from trace to severe. Image analysis of the mitral valve at midsystole had two stages: user-initialized segmentation and 3D deformable modeling with continuous medial representation (cm-rep). Semi-automated segmentation began with user-identification of valve location in 2D projection images generated from 3D US data. The mitral leaflets were then automatically segmented in 3D using the level set method. Second, a bileaflet deformable medial model was fitted to the binary valve segmentation by Bayesian optimization. The resulting cm-rep provided a visual reconstruction of the mitral valve, from which localized measurements of valve morphology were automatically derived. The features extracted from the fitted cm-rep included annular area, annular circumference, annular height, intercommissural width, septolateral length, total tenting volume, and percent anterior tenting volume. These measurements were compared to those obtained by expert manual tracing. Regurgitant orifice area (ROA) measurements were compared to qualitative assessments of MR severity. The accuracy of valve shape representation with cm-rep was evaluated in terms of the Dice overlap between the fitted cm-rep and its target segmentation.
The morphological features and anatomic ROA derived from semi-automated image analysis were consistent with manual tracing of 3D US image data and with qualitative assessments of MR severity made on clinical radiology. The fitted cm-reps accurately captured valve shape and demonstrated patient-specific differences in valve morphology among subjects with varying degrees of MR severity. Minimal variation in the Dice overlap and morphological measurements was observed when different cm-rep templates were used to initialize model fitting.
This study demonstrates the use of deformable medial modeling for semi-automated 3D reconstruction of mitral valve geometry using transesophageal 3D US. The proposed algorithm provides a parametric geometrical representation of the mitral leaflets, which can be used to evaluate valve morphology in clinical ultrasound images.
Medical Physics 02/2012; 39(2):933-50. DOI:10.1118/1.3673773 · 2.64 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: A new approach for constructing deformable continuous medial models for anatomical structures is presented. Medial models describe geometrical objects by first specifying the skeleton of the object and then deriving the boundary surface corresponding to the skeleton. However, an arbitrary specification of a skeleton will not be "valid" unless a certain set of sufficient conditions is satisfied. The most challenging of these is the non-linear equality constraint that must hold along the boundaries of the manifolds forming the skeleton. The main contribution of this paper is to leverage the biharmonic partial differential equation as a mapping from a codimension-0 subset of Euclidean space to the space of skeletons that satisfy the equality constraint. The PDE supports robust numerical solution on freeform triangular meshes, providing additional flexibility for shape modeling. The approach is evaluated by generating continuous medial models for a large dataset of hippocampus shapes. Generalizations to modeling more complex shapes and to representing branching skeletons are demonstrated.
[Show abstract][Hide abstract] ABSTRACT: The m-rep, a representation of the interior of one or more objects, from which boundaries can be synthesized, is described
in detail. An m-rep consists of sheets of medial atoms; both sampled and parametrized representations of these sheets are
described. Means of forming objects made from a main sheet (figure) and attached protrusion or indentation subfigures are described, as are multiscale hierarchies of object complexes, objects,
figures, atoms, and voxels. The object-relative coordinate system provided by m-reps is presented. To allow the estimation
of probabilities on populations of m-reps, the m-rep can be understood as an element in a feature space that takes the mathematical
form of a symmetric space. Doing this provides the ability to estimate probabilities by a generalization of principal component
analysis to these curved spaces.