SCIENTIFIC COMPUTING AND IMAGING INSTITUTE
at the University of Utah

Implantable cardioverter defibrillators (ICDs) are increasing used in abnormal configurations. We have developed a patient specific forward simulation model to predict efficacy of the defibrillation shock. Our goal was to develop a method of measuring the ICD surface potentials as the devices are tested during implantation surgery to use as verification of the simulation. A lead selection algorithm was used to develop a surface potential mapping system with 32 recording sites that do not interfere with implantation surgery. ICD discharge recordings were compared at similar locations to corresponding patient models. The reconstructed simulated surface potentials showed

To better understand the central nervous system, neurobiologists need to reconstruct the underlying neural circuitry from electron microscopy images. One of the necessary tasks is to segment the individual neurons. For this purpose, we propose a supervised learning approach to detect the cell membranes. The classifier was trained using AdaBoost, on local and context features. The features were selected to highlight the line characteristics of cell membranes. It is shown that using features from context positions allows for more information to be utilized in the classification. Together with the nonlinear discrimination ability of the AdaBoost classifier, this results in clearly noticeable improvements over previously used methods.

Keywords: crcns, neural networks

Understanding the neural circuitry of the retina requires us to map the connectivity of individual neurons in large neuronal tissue sections and analyze signal communication across processes from the electron microscopy images. One of the major bottlenecks in the critical path is the image mosaicing process where 2D slices are assembled from scanned microscopy image tiles. The problem of assembling the tiles is computationally non-trivial because of distortion of the specimen in the electron microscope due to heat and overlap between the scanned tiles. The complexity of the calculation arises from the massive size of the dataset and mathematical calculations required to calculate value of each pixel of the mosaic. We propose to use texture memory lookups to speedup the access to image tiles and data parallel computing enabled by the GPUs to accelerate this process. The proposed method results in noticeable improvements in speed of computation compared to other methods. Index Terms—Serial-section TEM, image mosaicing, GPGPU, CUDA, texture.

Keywords: crcns, electron microscopy, mosaics, retina, eye

The incorporation of massive amounts of data from different sources is a challenging task for the conception of any information visualization system. Especially the data heterogeneity often makes it necessary to include people from multiple domains with various fields of expertise. Hence, the inclusion of multiple users in a collaborative data analysis process introduces a whole new challenge for the design and conception of visualization applications. Using a multi-display environment to support co-located collaborative work seems to be a natural next step. However, adapting common visualization systems to multi-display environments poses several challenges.

We have come up with a number of design considerations for employing multiple-view visualizations in collaborative multi-display environments: adaptations of the visualization depending on display factors and user preferences, interaction techniques to facilitate information sharing and to guide the users' attention to relevant items in the environment, and the design of a flexible working environment, adjustable to varying group sizes and specific tasks.

Motivated by these considerations we propose a system relying on a spatial model of the environment as its main information source. We argue that the system design should be separated into basic multi-display environment functionality, such as multiple input handling and the management of the physical displays, and higher level functionality provided by the visualization system. An API offered by the multi-display framework thereby provides the necessary information about the environment and users to the visualization system.

We discuss in this paper efficient solvers for stochastic diffusion equations in random media. We employ generalized polynomial chaos (gPC) expansion to express the solution in a convergent series and obtain a set of deterministic equations for the expansion coefficients by Galerkin projection. Although the resulting system of diffusion equations are coupled, we show that one can construct fast numerical methods to solve them in a decoupled fashion. The methods are based on separation of the diagonal terms and off-diagonal terms in the matrix of the Galerkin system. We examine properties of this matrix and show that the proposed method is unconditionally stable for unsteady problems and convergent for steady problems with a convergent rate independent of discretization parameters. Numerical examples are provided, for both steady and unsteady random diffusions, to support the analysis.

Keywords: Generalized polynomial chaos, Stochastic Galerkin, Random diffusion, Uncertainty quantification

This paper presents a review of the current state-of-the-art of numerical methods for stochastic computations. The focus is on efficient high-order methods suitable for practical applications, with a particular emphasis on those based on generalized polynomial chaos (gPC) methodology. The framework of gPC is reviewed, along with its Galerkin and collocation approaches for solving stochastic equations. Properties of these methods are summarized by using results from literature. This paper also attempts to present the gPC based methods in a unified framework based on an extension of the classical spectral methods into multi-dimensional random spaces.

Keywords: Stochastic differential equations, generalized polynomial chaos, uncertainty quantification, spectral methods