SCI Publications

This paper presents Markov surfaces, a probabilistic algorithm for user-assisted segmentation of elongated structures in 3D images. The 3D segmentation problem is formulated as a path-finding problem, where path probabilities are described by Markov chains. Users define points, curves, or regions on 2D image slices, and the algorithm connects these user-defined features in a way that respects the underlying elongated structure in data. Transition probabilities in the Markov model are derived from intensity matches and interslice correspondences, which are generated from a slice-to-slice registration algorithm. Bezier interpolations between paths are applied to generate smooth surfaces. Subgrid accuracy is achieved by linear interpolations of image intensities and the interslice correspondences. Experimental results on synthetic and real data demonstrate that Markov surfaces can segment regions that are defined by texture, nearby context, and motion. A parallel implementation on a streaming parallel computer architecture, a graphics processor, makes the method interactive for 3D data.

We present a set of building blocks that provide scalable data movement capability to computational scientists and visualization researchers for writing their own parallel analysis. The set includes scalable tools for domain decomposition, process assignment, parallel I/O, global reduction, and local neighborhood communicationtasks that are common across many analysis applications. The global reduction is performed with a new algorithm, described in this paper, that efficiently merges blocks of analysis results into a smaller number of larger blocks. The merging is configurable in the number of blocks that are reduced in each round, the number of rounds, and the total number of resulting blocks. We highlight the use of our library in two analysis applications: parallel streamline generation and parallel Morse-Smale topological analysis. The first case uses an existing local neighborhood communication algorithm, whereas the latter uses the new merge algorithm.

Gradient domain processing remains a particularly computationally expensive technique even for relatively small images. When images become massive in size, giga or terapixel, these problems become particularly troublesome and the best serial techniques take on the order of hours or days to compute a solution. In this paper, we provide a simple framework for the parallel gradient domain processing. Specifically, we provide a parallel out-of-core method for the seamless stitching of gigapixel panoramas in a parallel MPI environment. Unlike existing techniques, the framework provides both a straightforward implementation, maintains strict control over the required/allocated resources, and makes no assumptions on the speed of convergence to an acceptable image. Furthermore, the approach shows good weak/strong scaling from several to hundreds of cores and runs on a variety of hardware.

Gradient domain processing is a computationally expensive image processing technique. Its use for processing massive images, giga or terapixels in size, can take several hours with serial techniques. To address this challenge, parallel algorithms are being developed to make this class of techniques applicable to the largest images available with running times that are more acceptable to the users. To this end we target the most ubiquitous form of computing power available today, which is small or medium scale clusters of commodity hardware. Such clusters are continuously increasing in scale, not only in the number of nodes, but also in the amount of parallelism available within each node in the form of multicore CPUs and GPUs. In this paper we present a hybrid parallel implementation of gradient domain processing for seamless stitching of gigapixel panoramas that utilizes MPI, threading and a CUDA based GPU component. We demonstrate the performance and scalability of our implementation by presenting results from two GPU clusters processing two large data sets.

Cardiac experimental electrophysiology is in need of a well-defined Minimum Information Standard for recording, annotating, and reporting experimental data. As a step toward establishing this, we present a draft standard, called Minimum Information about a Cardiac Electrophysiology Experiment (MICEE). The ultimate goal is to develop a useful tool for cardiac electrophysiologists which facilitates and improves dissemination of the minimum information necessary for reproduction of cardiac electrophysiology research, allowing for easier comparison and utilisation of findings by others. It is hoped that this will enhance the integration of individual results into experimental, computational, and conceptual models. In its present form, this draft is intended for assessment and development by the research community. We invite the reader to join this effort, and, if deemed productive, implement the Minimum Information about a Cardiac Electrophysiology Experiment standard in their own work.

Keywords: Minimum Information Standard; Cardiac electrophysiology; Data sharing; Reproducibility; Integration; Computational modelling

In this paper, we present two combinatorial methods to process 3-D steady vector fields, which both use graph algorithms to extract features from the underlying vector field. Combinatorial approaches are known to be less sensitive to noise than extracting individual trajectories. Both of the methods are a straightforward extension of an existing 2-D technique to 3-D fields. We observed that the first technique can generate overly coarse results and therefore we present a second method that works using the same concepts but produces more detailed results. We evaluate our method on a CFD-simulation of a gas furnace chamber. Finally, we discuss several possibilities for categorizing the invariant sets with respect to the flow.

Multiperspective images (MPIs) show more than what is visible from a single viewpoint and are a promising approach for alleviating the problem of occlusions. We present a comprehensive user study that investigates the effectiveness of MPIs for 3-D scene exploration. A total of 47 subjects performed searching, counting, and spatial orientation tasks using both conventional and multiperspective images. We use a flexible MPI framework that allows trading off disocclusion power for image simplicity. The framework also allows rendering MPI images at interactive rates, which enables investigating interactive navigation and dynamic 3-D scenes. The results of our experiments show that MPIs can greatly outperform conventional images. For searching, subjects performed on average 28% faster using an MPI. For counting, accuracy was on average 91% using MPIs as compared to 42% for conventional images.

Keywords: Interactive 3-D scene exploration, Navigation, Occlusions, User study, Visual interfaces

A population growth model that represents the growth trajectories of individual subjects is critical to study and understand neurodevelopment. This paper presents a framework for jointly estimating and modeling individual and population growth trajectories, and determining significant regional differences in growth pattern characteristics applied to longitudinal neuroimaging data. We use non-linear mixed effect modeling where temporal change is modeled by the Gompertz function. The Gompertz function uses intuitive parameters related to delay, rate of change, and expected asymptotic value; all descriptive measures which can answer clinical questions related to growth. Our proposed framework combines nonlinear modeling of individual trajectories, population analysis, and testing for regional differences. We apply this framework to the study of early maturation in white matter regions as measured with diffusion tensor imaging (DTI). Regional differences between anatomical regions of interest that are known to mature differently are analyzed and quantified. Experiments with image data from a large ongoing clinical study show that our framework provides descriptive, quantitative information on growth trajectories that can be directly interpreted by clinicians. To our knowledge, this is the first longitudinal analysis of growth functions to explain the trajectory of early brain maturation as it is represented in DTI.

Keywords: namic

In this paper, a method for interactive direct volume rendering is proposed for computing depth of field effects, which previously were shown to aid observers in depth and size perception of synthetically generated images. The presented technique extends those benefits to volume rendering visualizations of 3D scalar fields from CT/MRI scanners or numerical simulations. It is based on incremental filtering and as such does not depend on any precomputation, thus allowing interactive explorations of volumetric data sets via on-the-fly editing of the shading model parameters or (multi-dimensional) transfer functions.

In this paper, a new AdaBoost learning framework, called WNS-AdaBoost, is proposed for training discriminative models. The proposed approach significantly speeds up the learning process of adaptive boosting (AdaBoost) by reducing the number of data points. For this purpose, we introduce the weighted novelty selection (WNS) sampling strategy and combine it with AdaBoost to obtain an efficient and fast learning algorithm. WNS selects a representative subset of data thereby reducing the number of data points onto which AdaBoost is applied. In addition, WNS associates a weight with each selected data point such that the weighted subset approximates the distribution of all the training data. This ensures that AdaBoost can trained efficiently and with minimal loss of accuracy. The performance of WNS-AdaBoost is first demonstrated in a classification task. Then, WNS is employed in a probabilistic boosting-tree (PBT) structure for image segmentation. Results in these two applications show that the training time using WNS-AdaBoost is greatly reduced at the cost of only a few percent in accuracy.

Evaluating, comparing, and interpreting related pieces of information are tasks that are commonly performed during visual data analysis and in many kinds of information-intensive work. Synchronized visual highlighting of related elements is a well-known technique used to assist this task. An alternative approach, which is more invasive but also more expressive is visual linking in which line connections are rendered between related elements. In this work, we present context-preserving visual links as a new method for generating visual links. The method specifically aims to fulfill the following two goals: first, visual links should minimize the occlusion of important information; second, links should visually stand out from surrounding information by minimizing visual interference. We employ an image-based analysis of visual saliency to determine the important regions in the original representation. A consequence of the image-based approach is that our technique is application-independent and can be employed in a large number of visual data analysis scenarios in which the underlying content cannot or should not be altered. We conducted a controlled experiment that indicates that users can find linked elements in complex visualizations more quickly and with greater subjective satisfaction than in complex visualizations in which plain highlighting is used. Context-preserving visual links were perceived as visually more attractive than traditional visual links that do not account for the context information.

This article presents a simple framework for progressive processing of high-resolution images with minimal resources. We demonstrate this framework's effectiveness by implementing an adaptive, multi-resolution solver for gradient-based image processing that, for the first time, is capable of handling gigapixel imagery in real time. With our system, artists can use commodity hardware to interactively edit massive imagery and apply complex operators, such as seamless cloning, panorama stitching, and tone mapping.

We introduce a progressive Poisson solver that processes images in a purely coarse-to-fine manner, providing near instantaneous global approximations for interactive display (see Figure 1). We also allow for data-driven adaptive refinements to locally emulate the effects of a global solution. These techniques, combined with a fast, cache-friendly data access mechanism, allow the user to interactively explore and edit massive imagery, with the illusion of having a full solution at hand. In particular, we demonstrate the interactive modification of gigapixel panoramas that previously required extensive offline processing. Even with massive satellite images surpassing a hundred gigapixels in size, we enable repeated interactive editing in a dynamically changing environment. Images at these scales are significantly beyond the purview of previous methods yet are processed interactively using our techniques. Finally our system provides a robust and scalable out-of-core solver that consistently offers high-quality solutions while maintaining strict control over system resources.