Carlos Alberola

Nonrigid registration of 3D tensor medical data

New medical imaging modalities offering multi-valued data, such as phase contrast MRA and diffusion tensor MRI, require general representations for the development of automated algorithms. In this paper we propose a unified framework for the registration of medical volumetric multi-valued data using local matching. The paper extends the usual concept of similarity between two pieces of data to be matched, commonly used with scalar (intensity) data, to the general tensor case. Our approach to registration is based on a multiresolution scheme, where the deformation field estimated in a coarser level is propagated to provide an initial deformation in the next finer one. In each level, local matching of areas with a high degree of local structure and subsequent interpolation are performed. Consequently, we provide an algorithm to assess the amount of structure in generic multi-valued data by means of gradient and correlation computations. The interpolation step is carried out by means of the Kriging estimator, which provides a novel framework for the interpolation of sparse vector fields in medical applications. The feasibility of the approach is illustrated by results on synthetic and clinical data.

diSNei: A Collaborative Environment for Medical Images Analysis and Visualization

In this paper we describe our environment diSNei, a graphical tool for collaborative image analysis and visualization of models created out of slices of volume data; this application allows a number of users to simultaneous and coordinatedly analyze medical images, create graphical models, navigate through them and superimpose raw data onto the models. The application is intended to help physicians interpret data in the case that ambiguous situations may appear, by means of collaboration with other colleagues. It is therefore an integrated environment for expertise interchange among physicians and we believe that it is a powerful tool for academic purposes as well. Other outstanding application features are its being multiplatform, and, particularly, the fact that it can run on NT computers, and the support for stereo rendering so as to obtain a deep sensation of inmersion into the models.

A Bayesian Approach to in vivo Kidney Ultrasound Contour Detection Using Markov Random Fields

Automatic detection of structures in medical images is of great importance for the implementation of tools that can obtain accurate measurements for an eventual diagnosis. In this paper, a new method for the creation of such tools is presented. We focus on in vivo kidney ultrasound, a target in which classical methods fail due to the inherent difficulty of such an imaging modality and organ. The proposed method operates on every slice by detecting kidney contours under a probabilistic Bayesian framework. We make use of Markov Random Fields ideas to model the problem and find the solution. A computer easy-touse interface to the model is also presented.Automatic detection of structures in medical images is of great importance for the implementation of tools that can obtain accurate measurements for an eventual diagnosis. In this paper, a new method for the creation of such tools is presented. We focus on in vivo kidney ultrasound, a target in which classical methods fail due to the inherent difficulty of such an imaging modality and organ. The proposed method operates on every slice by detecting kidney contours under a probabilistic Bayesian framework. We make use of Markov Random Fields ideas to model the problem and find the solution. A computer easy-to-use interface to the model is also presented.

A general approach to nonrigid registration: decoupled optimization

To a great extent, the success or advanced image-guided medical procedures hinges on non-rigid volume registration. For example, non-rigid registration must be applied in interventional approaches where intra-operative information is used to update high-quality preoperative data; in follow-up studies in order to assess time-evolution of development; aging, pathology or treatment; and in many other applications including inter-subject variability and population-based atlas construction. In this paper we examine several computational schemes that warp one volumetric dataset onto another. We also explore the inevitable trade-off between the computational load and the incorporation of sophisticated similarity measures necessary for multimodal volumes. Estimated deformation fields are based on the variational formulation of partial derivative equations (PDEs), which includes a similarity and a regularization term. We compare numerical solutions to this problem using the Euler-Lagrange equations (EL), the finite element discretization (FE), and decoupled optimization over the possible deformations (DO).

Multiresolution compression schemes for medical image and volume data

In this paper we propose a multiresolution compression scheme applied to image and volume data. It is intended for rapid browsing in graphical files, in which some regions of interest may exist; therefore the compression algorithms focus on incorporating quick full-reconstruction procedures, both in images and volumes. This scheme is based on vector quantization of the coefficients of a wavelet decomposition. The differences in the methods lie on how vectors in the multiresolution decompositions are selected for vector quantization. We have also developed a client-server application using this scheme, which allows images with increasing levels of detail to be represented in the client system as they are received from the server and decoded. Numerical results show the performances of all the selections made.

Markov Random Fields in the Context of Stereo Vision

The term stereo vision refers to the ability of an observer (either a human or a machine) to recover the three-dimensional information of a scene by means of (at least) two images taken from different viewpoints. Under the scope of this problem—and provided that cameras are calibrated—two subproblems are typically considered, namely, the correspondence problem, and the reconstruction problem (Trucco & Verri, 1998). The former refers to the search for points in the two images that are projections of the same physical point in space. Since the images are taken from different viewpoints, every point in the scene will project onto different image points, i.e, onto points with different coordinates in every image coordinate system. It is precisely this disparity in the location of image points that gives the information needed to reconstruct the point position in space. The second problem, i.e., the reconstruction problem, deals with calculating the disparity between a set of corresponding points in the two images to create a disparity map, and to convert this into a three-dimensional map. In this context, we will show howMarkov Random Fields (MRFs) can be effectively used. It is well known that MRFs constitute a powerful tool to incorporate spatial local interactions in a global context (Geman & Geman, 1984). So, in this chapter, we will consider local interactions that define proper MRFs to develop a model that can be applied in the process of recovery of the 3D structure of the real world using stereo pairs of images. To this end, we will briefly describe the whole stereo reconstruction process (Fig. 1), including the process of selection of features, some important aspects regarding the calibration of the camera system and related geometric transformations of the images and, finally, probabilistic analyses usable in the definition of MRFs to solve the correspondence problem. In the model to describe, both a priori and a posteriori probabilities will be separately considered and derived making use of reasonable selections of the potentials (Winkler, 1995) that define the MRFs on the basis of specific analytic models. In the next section, a general overview of a stereo system will be shown. In Sec. 3, a brief overview of some well known stereo correspondence algorithms is given. Sec. 4describes the main stages of a stereo correspondence system in which MRFS can be applied. Sec. 5describes the camera model that will be considered in this chapter together with some important related issues like: camera calibration, the epipolar constraint and image rectification. Sec. 6describes the concept of Markov random fields, and related procedures, like simulated annealing. Sec. 7 introduces MRFs for the edge detection problem. Sec. 8 describes, in detail, how MRFs can 3

Multimodal Interfaces for Laparoscopic Training

Multimodal interfaces are providing promising simulation solutions for training different practitioners as surgeons. These environments present visual and haptic interaction to the trainee, as in a real intervention. They offer numerous advantages over the traditional learning process, like the possibility of monitoring the skills and delivering constructive feedback. This chapter presents a multimodal interface for laparoscopic training describing the functionality and the main technical components of the SINERGIA laparoscopic simulator. The reader will learn about how the visual and haptic interaction of the surgeon is emulated, and how different multimodal training scenarios can be designed and built. Simulator’s architecture and some of its core technologies such as the collision detection and handling algorithm are described in more detail. On the other hand, this chapter introduces two recent research contributions in this field: a better understanding of how surgeons perceive the consistency of tissues, and a conceptual framework for the analysis, design and validation of multimodal simulation technologies in surgical training. In the near future, physicians will be trained and accredited using these multimodal solutions, leading to a safer and more efficient surgery.

Anisotropic Regularization of Posterior Probability Maps Using Vector Space Projections. Application to MRI Segmentation

In this paper we address the problem of regularized data classification. To this extent we propose to regularize spatially the class-posterior probability maps, to be used by a MAP classification rule, by applying a non-iterative anisotropic filter to each of the class-posterior maps. Since the filter cannot guarantee that the smoothed maps preserve their probabilities meaning (i.e., probabilities must be in the range [0,1] and the class-probabilities must sum up to one), we project the smoothed maps onto a probability subspace. Promising results are presented for synthetic and real MRI datasets.