Catálogo de publicaciones - libros

Among the many features used for classification in computer-aided detection (CAD) systems targeting colonic polyps, those based on differences between the shapes of polyps and folds are most common. We introduce here an explicit parametric model for the or colon wall. The proposed model captures the overall shape of the haustra and we use it to derive the probability distribution of features relevant to polyp detection. The usefulness of the model is demonstrated through its application to a colon CAD algorithm.

Automated motion reconstruction of the left ventricle (LV) from 3D echocardiography provides insight into myocardium architecture and function. Low image quality and artifacts make 3D ultrasound image processing a challenging problem. We introduce a LV tracking method, which combines textural and structural information to overcome the image quality limitations. Our method automatically reconstructs the motion of the LV contour (endocardium and epicardium) from a sequence of 3D ultrasound images.

The branching pattern and geometry of coronary microvessels are of high interest to understand and model the blood flow distribution and the processes of contrast invasion, ischemic changes and repair in the heart in detail. Analysis is performed on high resolution, 3D volumes of the arterial microvasculature of entire goat hearts, which are acquired with an imaging cryomicrotome.

Multi-scale vessel detection is an important step required for a detailed quantitative analysis of the coronary microvasculature. Based on visual inspection, the derived lineness filter shows promising results on real data and digital phantoms, on the way towards accurate computerized reconstructions of entire coronary trees.

The novel lineness filter exploits the local first and second order multi-scale derivatives in order to give an intensity-independent response to line centers and to suppress unwanted responses to steep edges.

This paper incorporates multiscale vesselness filtering into the Livewire framework to simultaneously compute optimal medial axes and boundaries in vascular images. To this end, we extend the existing 2D graph search to 3D space to optimize not only for spatial variables (,), but also for radius values at each node. In addition, we minimize change for both scale and the smallest principle curvature and incorporate vessel boundary evidence in our optimization. When compared to two sets of DRIVE expert manual tracings, our proposed technique reduced the overall segmentation task time by 68.2%, had a similarity ratio of 0.772 (0.775 between manual), and was 98.2% reproducible.

This paper introduces a novel approach to quantify asymmetry in each point of a surface. The measure is based on analysing displacement vectors resulting from nonrigid image registration. A symmetric atlas, generated from control subjects is registered to a given subject image. A comparison of the resulting displacement vectors on the left and right side of the symmetry plane, gives a point-wise measure of asymmetry. The asymmetry measure was applied to the study of Crouzon syndrome using Micro CT scans of genetically modified mice. Crouzon syndrome is characterised by the premature fusion of cranial sutures, which gives rise to a highly asymmetric growth. Quantification and localisation of this asymmetry is of high value with respect to surgery planning and treatment evaluation. Using the proposed method, asymmetry was calculated in each point of the surface of Crouzon mice and wild-type mice (controls). Asymmetry appeared in similar regions for the two groups but the Crouzon mice were found significantly more asymmetric. The localisation ability of the method was in good agreement with ratings from a clinical expert. Validating the quantification ability is a less trivial task due to the lack of a gold standard. Nevertheless, a comparison with a different, but less accurate measure of asymmetry revealed good correlation.

We present an approach to detect anatomical structures by configurations of interest points, from a single example image. The representation of the configuration is based on Markov Random Fields, and the detection is performed in a single iteration by the algorithm. Instead of sequentially matching pairs of interest points, the method takes the entire set of points, their local descriptors and the spatial configuration into account to find an optimal mapping of modeled object to target image. The image information is captured by symmetry-based interest points and local descriptors derived from Gradient Vector Flow. Experimental results are reported for two data-sets showing the applicability to complex medical data.

The tracking and reconstruction of myocardial motion is critical to the diagnosis and treatment of heart disease. Currently, little has been done for the analysis of motion in long axis (LA) cardiac images. We propose a new fully automated motion reconstruction method for grid- tagged MRI that combines Gabor filters and deformable models. First, we use a Gabor filter bank to generate the corresponding phase map in the myocardium and estimate the location of grid tag intersections. Second, we use a non-rigid registration module driven by thin plate splines (TPS) to generate a transformation function between tag intersections in two consecutive images. Third, deformable spline models are initialized using Fourier domain analysis and tracked during the cardiac cycle using the TPS generated transformation function. The splines will then locally deform under the influence of gradient flow and image phase information. The final motion is decomposed into tangential and normal components corresponding to the local orientation of the heart wall. The new method has been tested on LA phantoms and heart data, and its performance has been quantitatively validated. The results show that our method can reconstruct the motion field in LA cardiac tagged MR images accurately and efficiently.

We present an algorithm to generate samples from probability distributions on the space of curves. We view a traditional curve evolution energy functional as a negative log probability distribution and sample from it using a Markov chain Monte Carlo (MCMC) algorithm. We define a proposal distribution by generating smooth perturbations to the normal of the curve and show how to compute the transition probabilities to ensure that the samples come from the posterior distribution. We demonstrate some advantages of sampling methods such as robustness to local minima, better characterization of multi-modal distributions, access to some measures of estimation error, and ability to easily incorporate constraints on the curve.

In this paper, we present a general framework for extracting 3D centerlines from volumetric datasets. Unlike the majority of previous approaches, we do not require a prior segmentation of the volume nor we do assume any particular tubular shape. Centerline extraction is performed using a morphology-guided level set model. Our approach consists of: i) learning the structural patterns of a tubular-like object, and ii) estimating the centerline of a tubular object as the path with minimal cost with respect to outward flux in gray level images. Such shortest path is found by solving the Eikonal equation. We compare the performance of our method with existing approaches in synthetic, CT, and multiphoton 3D images, obtaining substantial improvements, especially in the case of irregular tubular objects.

Despite many research efforts, accurate extraction of structures of interest still remains a difficult issue in many medical imaging applications. This is particularly the case for magnetic resonance (MR) images where image quality depends highly on the acquisition protocol. In this paper, we propose a variational region based algorithm that is able to deal with spatial perturbations of the image intensity directly. Image segmentation is obtained by using a -Convergence approximation for a multi-scale piecewise smooth model. This model overcomes the limitations of global region models while avoiding the high sensitivity of local approaches. The proposed model is implemented efficiently using recursive Gaussian convolutions. Numerical experiments on 2-dimensional human liver MR images show that our model compares favorably to existing methods.