Catálogo de publicaciones - libros

To more accurately separate each pulmonary nodule from its background in a low dose computer tomography (LDCT) chest image, two new adaptive probability models of visual appearance of small 2D and large 3D pulmonary nodules are used to control evolution of deformable boundaries. The appearance prior is modeled with a translation and rotation invariant Markov-Gibbs random field of voxel intensities with pairwise interaction analytically identified from a set of training nodules. Appearance of the nodules and their background in a current multi-modal chest image is also represented with a marginal probability distribution of voxel intensities. The nodule appearance model is isolated from the mixed distribution using its close approximation with a linear combination of discrete Gaussians. Experiments with real LDCT chest images confirm high accuracy of the proposed approach.

In this paper, we propose a fast intensity-based registration algorithm for the analysis of contrast-enhanced breast MR images. Motion between pre-contrast and post-contrast images has been modeled by a combination of rigid transformation and free-form deformation. By modeling the conditional probability function to be Gaussian and considering the normalized mutual information (NMI) criterion, we create a pair of auxiliary images to speed up the registration process. The auxiliary images are registered to the actual images by optimizing the simple sum of squared difference (SSD) criterion. The overall registration is achieved by linearly combining the deformation observed in the auxiliary images. One well-known problem of non-rigid registration of contrast enhanced images is the contraction of enhanced lesion volume. We address this problem by rejecting the intensity outliers from registration. Results have shown that our method could achieve accurate registration of the data while successfully prevent the contraction of the contrast enhanced lesion volume.

Current approaches to quantitative analysis of DCE-MRI with non-linear models involve the convolution of an arterial input function (AIF) with the contrast agent concentration at a voxel or regional level. Full quantification provides meaningful biological parameters but is complicated by the issues related to convergence, (de-)convolution of the AIF, and goodness of fit. To overcome these problems, this paper presents a penalized spline smoothing approach to model the data in a semi-parametric way. With this method, the AIF is convolved with a set of B-splines to produce the design matrix, and modeling of the resulting deconvolved biological parameters is obtained in a way that is similar to the parametric models. Further kinetic parameters are obtained by fitting a non-linear model to the estimated response function and detailed validation of the method, both with simulated and data is provided.

This paper describes an automatic tissue segmentation algorithm for brain MRI of young children. Existing segmentation methods developed for the adult brain do not take into account the specific tissue properties present in the brain MRI of young children. We examine the suitability of state-of-the-art methods developed for the adult brain when applied to the segmentation of the young child brain MRI. We develop a method of creation of a population-specific atlas from young children using a single manual segmentation. The method is based on non-linear propagation of the segmentation into population and subsequent affine alignment into a reference space and averaging. Using this approach we significantly improve the performance of the popular EM segmentation algorithm on brain MRI of young children.

This paper presents a learning based method for automatic extraction of the major cortical sulci from MRI volumes or extracted surfaces. Instead of using a few pre-defined rules such as the mean curvature properties, to detect the major sulci, the algorithm learns a discriminative model by selecting and combining features from a large pool of candidates. We used the Probabilistic Boosting Tree algorithm [16] to learn the model, which implicitly discovers and combines rules based on manually annotated sulci traced by neuroanatomists. The algorithm almost has no parameters to tune and is fast because of the adoption of integral volume and 3D Haar filters. For a given approximately registered MRI volume, the algorithm computes the probability of how likely it is that each voxel lies on a major sulcus curve. Dynamic programming is then applied to extract the curve based on the probability map and a shape prior. Because the algorithm can be applied to MRI volumes directly, there is no need to perform preprocessing such as tissue segmentation or mapping to a canonical space. The learning aspect makes the approach flexible and it also works on extracted cortical surfaces.

This paper addresses the problem of creating probabilistic brain atlases from manually labeled training data. We propose a general mesh-based atlas representation, and compare different atlas models by evaluating their posterior probabilities and the posterior probabilities of their parameters. Using such a Baysian framework, we show that the widely used ”average” brain atlases constitute relatively poor priors, partly because they tend to overfit the training data, and partly because they do not allow to align corresponding anatomical features across datasets. We also demonstrate that much more powerful representations can be built using content-adaptive meshes that incorporate non-rigid deformation field models. We believe extracting optimal prior probability distributions from training data is crucial in light of the central role priors play in many automated brain MRI analysis techniques.

The process of constructing an atlas typically involves selecting one individual from a sample on which to base or root the atlas. If the individual selected is far from the population mean, then the resulting atlas is biased towards this individual. This, in turn, can bias any inferences made with the atlas. Unbiased atlas construction addresses this issue by either basing the atlas on the individual which is the median of the sample or by an iterative technique whereby the atlas converges to the unknown population mean. In this paper, we explore the question of whether a single atlas is appropriate for a given sample or whether there is sufficient image based evidence from which we can infer multiple atlases, each constructed from a subset of the data. We refer to this process as . Essentially, we determine whether the sample, and hence the population, is multi-modal and is best represented by an atlas per mode. We use the mean shift algorithm to identify the modes of the sample and multidimensional scaling to visualize the clustering process.

In order to more reliably recover cardiac information from noise-corrupted patient-specific measurements, it is essential to employ meaningful constraining models and adopt appropriate optimization criteria to couple the models with the measurements. While biomechanical models have been extensively used for myocardial motion recovery with encouraging results, the passive nature of such constraints limits their ability to fully count for the deformation caused by active forces of the myocytes. To overcome such limitations, we propose to adopt a as the prior constraint for heart motion analysis. The model is comprised of a cardiac electric wave propagation model, an electromechanical coupling model, and a biomechanical model, and thus more completely describes the macroscopic cardiac physiology. Embedded within a multiframe state-space framework, the uncertainties of the model and the patient-specific measurements are systematically dealt with to arrive at optimal estimates of the cardiac kinematics and possibly beyond. Experiments have been conducted on synthetic data and MR image sequences to illustrate its abilities and benefits.

We present a fully automated method to estimate the location and orientation of the left ventricle (LV) in four-dimensional (4D) cardiac magnetic resonance (CMR) images without any user input. The method is based on low-level image processing techniques incorporating anatomical knowledge and is able to provide rapid, robust feedback for automated scan planning or further processing. The method relies on a novel combination of temporal Fourier analysis of image cines with simple contour detection to achieve a fast localization of the heart. Quantitative validation was performed using 4D CMR datasets from 330 patients (54024 images) with a range of cardiac and vascular disease by comparing manual location with the automatic results. The method failed on one case, and showed average bias and precision of under 5mm in apical, mid-ventricular and basal slices in the remaining 329. The errors in automatic orientation were similar to the errors in scan planning as performed by experienced technicians.

We propose a pairwise active appearance model (PAAM) to characterize statistical regularities in shape, appearance, and motion presented by a target that undergoes a series of motion phases, such as the left ventricle in echocardiography. The PAAM depicts the transition in motion phase through a Markov chain and the transition in both shape and appearance through a conditional Gaussian distribution. We learn from a database the joint Gaussian distribution of the shapes and appearances belonging to two consecutive motion phases (i.e., a pair of motion phases), from which we analytically compute the conditional Gaussian distribution. We utilize the PAAM in tracking the left ventricle contour in echocardiography and obtain improved tracking results in terms of localization accuracy when compared with expert-specified contours.