Visualization Of Large Data Sets Research Articles

A New Training Method for Large Self Organizing Maps

Self Organizing Maps (SOMs) are widely used neural networks for classification or visualization of large datasets. Like many neural network simulations, implementations of the SOM algorithm need a scan of all the neural units in order to simulate the work of a parallel machine. This paper reports a new learning algorithm that speeds up the training of a SOM with a little loss of the performance on many quality tests. The very low computation time, means that this algorithm can be used as a fast visualization tool for large multidimensional datasets.

Automated Platform for High-Resolution Tissue Imaging Using Nanospray Desorption Electrospray Ionization Mass Spectrometry

An automated platform has been developed for acquisition and visualization of mass spectrometry imaging (MSI) data using nanospray desorption electrospray ionization (nano-DESI). The new system enables robust operation of the nano-DESI imaging source over many hours by precisely controlling the distance between the sample and the nano-DESI probe. This is achieved by mounting the sample holder onto an automated XYZ stage, defining the tilt of the sample plane, and recalculating the vertical position of the stage at each point. This approach is useful for imaging of relatively flat samples such as thin tissue sections. Custom software called MSI QuickView was developed for visualization of large data sets generated in imaging experiments. MSI QuickView enables fast visualization of the imaging data during data acquisition and detailed processing after the entire image is acquired. The performance of the system is demonstrated by imaging rat brain tissue sections. Low background noise enables simultaneous detection of lipids and metabolites in the tissue section. High-resolution mass analysis combined with tandem mass spectometry (MS/MS) experiments enabled identification of the observed species. In addition, the high dynamic range (>2000) of the technique allowed us to generate ion images of low-abundance isobaric lipids. A high-spatial resolution image was acquired over a small region of the tissue section revealing the distribution of an abundant brain metabolite, creatine, on the boundary between the white and gray matter. The observed distribution is consistent with the literature data obtained using magnetic resonance spectroscopy.

The Molecule Cloud - compact visualization of large collections of molecules

BackgroundAnalysis and visualization of large collections of molecules is one of the most frequent challenges cheminformatics experts in pharmaceutical industry are facing. Various sophisticated methods are available to perform this task, including clustering, dimensionality reduction or scaffold frequency analysis. In any case, however, viewing and analyzing large tables with molecular structures is necessary. We present a new visualization technique, providing basic information about the composition of molecular data sets at a single glance.SummaryA method is presented here allowing visual representation of the most common structural features of chemical databases in a form of a cloud diagram. The frequency of molecules containing particular substructure is indicated by the size of respective structural image. The method is useful to quickly perceive the most prominent structural features present in the data set. This approach was inspired by popular word cloud diagrams that are used to visualize textual information in a compact form. Therefore we call this approach “Molecule Cloud”. The method also supports visualization of additional information, for example biological activity of molecules containing this scaffold or the protein target class typical for particular scaffolds, by color coding. Detailed description of the algorithm is provided, allowing easy implementation of the method by any cheminformatics toolkit. The layout algorithm is available as open source Java code.ConclusionsVisualization of large molecular data sets using the Molecule Cloud approach allows scientists to get information about the composition of molecular databases and their most frequent structural features easily. The method may be used in the areas where analysis of large molecular collections is needed, for example processing of high throughput screening results, virtual screening or compound purchasing. Several example visualizations of large data sets, including PubChem, ChEMBL and ZINC databases using the Molecule Cloud diagrams are provided.

Self-Organizing Maps for In Silico Screening and Data Visualization.

Self-organizing maps, which are unsupervised artificial neural networks, have become a very useful tool in a wide area of disciplines, including medicinal chemistry. Here, we will focus on two applications of self-organizing maps: the use of self-organizing maps for in silico screening and for clustering and visualisation of large datasets. Additionally, the importance of parameter selection is discussed and some modifications to the original algorithm are summarised.

Fast Visualization Programming of Ocean Remote Sensing Information Based on IDL

The most commonly used method for ocean information visualization is plotting two-dimensional distribution images by using the IDLgrContour object in IDL,But when the data volume exceeds 10M, it process looks like rolling window curtains.Draw on the experience of contour display type,here we put forward the application technique of efficient visualization of large data sets,include basic framework,colorbar configuration and realizing process.The example shows it can meet the fast display requirement of common ocean remote sensing information with excellent expansibility and flexibility.

Efficient Data Projection for Visual Analysis of Large Data Sets Using Neural Networks

The most classical visualization methods, including multidimensional scaling and its particular case – Sammon's mapping, encounter difficulties when analyzing large data sets. One of possible ways to solve the problem is the application of artificial neural networks. This paper presents the visualization of large data sets using the feed-forward neural network – SAMANN. This back propagation-like learning rule has been developed to allow a feed-forward artificial neural network to learn Sammon's mapping in an unsupervised way. In its initial form, SAMANN training is computation expensive. In this paper, we discover conditions optimizing the computational expenditure in visualization even of large data sets. It is shown possibility to reduce the original dimensionality of data to a lower one using small number of iterations. The visualization results of real-world data sets are presented.

Multi-Class Data Exploration Using Space Transformed Visualization Plots

Visualization of large datasets is computationally expensive. For this reason, enveloping methods have been used to visualize such datasets. Using enveloping methods, we visualize summary statistics of the data in the space transformed visualization (STV) plots, such as the traditional parallel coordinate plot (TPCP), instead of the actual data records. Existing enveloping methods, however, are limited only to the TPCP and they can also be misleading. This is because the parallel coordinates are parameter transformations and the summary statistics computed for the original data records are not preserved throughout the transformation to the parallel coordinates space. We propose enveloping methods that avoid this drawback and that can be applied not only to the TPCP but also to a family of STV plots such as the smooth parallel coordinate plot (SPCP) and the Andrews plot. We apply the proposed methods to the min–max, the quartiles, and the concentration interval envelopes (CIEs). These enveloping methods allow us to visually describe the geometry of given classes without the need of visualizing each single data record. These methods are effective for visualizing large datasets, as illustrated for real datasets, because they mitigate the cluttering effect in visualizing large-sized classes in the STV plots. Supplemental materials, including R-code, are available online to enable readers to reproduce the graphs in this article and/or apply the proposed methods to their own data.

The Future of Biomedical Content in Applied Optics

Applied Optics has had a long and distinguished history of content related to optics and photonics in the life sciences. Many seminal papers in tissue optics, biomedical spectroscopy, biomedical optical imaging, and biological microscopy have appeared in these pages. Many topical editors with specific expertise in biomedical optics have served the journal with distinction, and the Division of Biomedical Optics was launched as the fourth independent Division of the journal in 2006. Due to the outstanding quality of biomedical content throughout all of its journals and continued rapid growth anticipated in this area, this fall the OSA is launching a new peer-reviewed journal devoted specifically to biomedical optics and photonics. Biomedical Optics Express will become OSA’s principal outlet for serving the biomedical optics community with rapid, open-access publication of content related to optics, photonics, and imaging in the life sciences. The new journal will offer many features of particular interest to biomedical authors, including free color figures, movies, animations, and live reference links. Authors will also be able to take advantage of interactive science publishing (ISP) capability for online visualization of large datasets, and automated submission of published articles to PubMed Central to assist in compliance with federal open-access requirements. Although authors will initially not be charged for publishing in Biomedical Optics Express (no author fees in the first year of publication and reduced fees in the second), the open-access publishing model requires that authors and their funding sponsors eventually offset the bulk of publication costs through author fees. In order to support consolidation of biomedical content in the new journal, Applied Optics will close its Division of Biomedical Optics by the end of 2010 and encourage all authors to consider submitting their work to Biomedical Optics Express. The Division will officially close for submissions on August 31, 2010 and will attempt to publish all remaining manuscripts under review by the end of the year. Despite the closure of the Division, however, there may remain circumstances under which some authors may prefer to submit and publish selected manuscripts with biomedical content to Applied Optics. Such submissions will remain welcome, and should be submitted to the most appropriate of the other Divisions, whether Optical Technology, Information Processing, or Lasers, Photonics, and Environmental Optics. We thank the biomedical optics community for its longstanding support of Applied Optics, and join in welcoming Biomedical Optics Express to OSA’s portfolio of journals.

'Brukin2D': a 2D visualization and comparison tool for LC-MS data

BackgroundLiquid Chromatography-Mass Spectrometry (LC-MS) is a commonly used technique to resolve complex protein mixtures. Visualization of large data sets produced from LC-MS, namely the chromatogram and the mass spectra that correspond to its compounds is the focus of this work.ResultsThe in-house developed 'Brukin2D' software, built in Matlab 7.4, which is presented here, uses the compound data that are exported from the Bruker 'DataAnalysis' program, and depicts the mean mass spectra of all the chromatogram compounds from one LC-MS run, in one 2D contour/density plot. Two contour plots from different chromatograph runs can then be viewed in the same window and automatically compared, in order to find their similarities and differences. The results of the comparison can be examined through detailed mass quantification tables, while chromatogram compound statistics are also calculated during the procedure.Conclusion'Brukin2D' provides a user-friendly platform for quick, easy and integrated view of complex LC-MS data. The software is available at .

A Comparison of the Perceptual Benefits of Linear Perspective and Physically-Based Illumination for Display of Dense 3D Streamtubes

Large datasets typically contain coarse features comprised of finer sub-features. Even if the shapes of the small structures are evident in a 3D display, the aggregate shapes they suggest may not be easily inferred. From previous studies in shape perception, the evidence has not been clear whether physically-based illumination confers any advantage over local illumination for understanding scenes that arise in visualization of large data sets that contain features at two distinct scales. In this paper we show that physically-based illumination can improve the perception for some static scenes of complex 3D geometry from flow fields. We perform human-subjects experiments to quantify the effect of physically-based illumination on participant performance for two tasks: selecting the closer of two streamtubes from a field of tubes, and identifying the shape of the domain of a flow field over different densities of tubes. We find that physically-based illumination influences participant performance as strongly as perspective projection, suggesting that physically-based illumination is indeed a strong cue to the layout of complex scenes. We also find that increasing the density of tubes for the shape identification task improved participant performance under physically-based illumination but not under the traditional hardware-accelerated illumination model.

Ubiquitous interactive visualization of 3D mantle convection using a web-portal with Java and Ajax framework

We have developed a new strategy and espouse a novel paradigm for large-scale computing and real-time interactive visualization. This philosophy calls for intense interactive sessions for a couple of hours at a time at the expense of storing data on many disk drives during regular or heroic runs on massively parallel systems. We have already carried out successfully real-time volume-rendering visualization by employing hundreds of processors for a grid with over 25 million unknowns. Both Cartesian and spherical 3D mantle convection are visualized. The volume-rendered images are viewed on a large display device, with many panels holding around 13 million pixels. We will employ a software strategy involving an hierarchical rendering service, which will have as software an Ajax interface for interactive visualization of large data sets on many different platforms from desktop PC’s to hand-held devices, such as the OQO and the Nokia N-800. An option for stereo viewing is also implemented. We have installed a user interface as web application, using Java and Ajax framework in order to achieve over the Internet reasonable accessibility to our ongoing runs. Our goal is to expand the array of interactive devices, which will make it feasible to carry out ubiquitous visualization and monitoring of large-scale simulations or onsite events and to allow for collaborations across oceans.

Distributed Graphics Pipelines on the Grid

jgViz uses standard grid technologies and Chromium cluster graphics software to schedule the best available distributed graphics pipeline. Visualization of large data sets using computer graphics techniques has become a mainstream requirement for many scientists and engineers. However, achieving interactive performance when rendering a complex visualization often requires high-performance computing facilities. One solution is to exploit real-time graphics accelerators, which are used in the full range of computing devices from games consoles and set-top boxes to supercomputers and advanced training simulators. At the same time, the service infrastructure for distributed networks has produced the third generation of the connected world: the grid. jgViz, our Java-implemented grid visualization system, uses grid functionality to enable transparent access to reliable parallel graphics pipelines. Lightweight and highly portable, jgViz aims to be plug-and-play for the existing grid resource allocation mechanism

Visualization and Analysis of Multi-terabyte Geophysical Datasets in an Interactive Setting with Remote Webcam Capabilities

Visualizing and analyzing datasets in the geosciences is becoming increasingly complicated as their volumes are growing explosively. This poses a challenging problem for researchers who must sift through terabytes of data to discover useful relationships inside the information. There is a great need for geophysicists to interactively explore their data sets. Conventional visualization systems lack adequate bandwidth and rendering capabilities necessary for the largest data sets. CAVE and Powerwall display devices are necessary for researchers to explore their data sets in an immersive setting. We describe a utilitarian system targeted specifically at the cost-effective interactive exploration of data sets tens of terabytes in size and harness this system for visualization and analysis of geophysical simulations. Webcams can be used as a steering device to track a local region of interest, which is useful for remote visualization of large data sets. This system will be employed as a web-service under the auspices of Narada-Brokering, while using webcam technologies to enable remote visualization for collaborating researchers. Webcams can be incorporated in a point-to-point network for rapid exchange of information and quickly announcing natural disasters, such as tsunamis, landslides and earthquakes.

Visualization of large data sets by mixing Tcl and C++ interfaces to the VTK library

This article describes a technique to build applications for the visualization of large data sets using VTK library and components written in Tcl and C++. VTK is a software system for computer graphics, visualization and image processing. VTK library is written in C++ however it provides interfaces to the scripting languages Tcl, Python and Java. Though it is possible to write a whole VTK application in a scripting language like Tcl, it is more suitable, for efficiency reasons, to implement some functionality in a compiled language like C/C++ for instance. This article presents in through details how to access VTK objects from different languages and how to mix Tcl and C++ components in one application.

Distributed and collaborative visualization of large data sets using high-speed networks

We describe an architecture for distributed collaborative visualization that integrates video conferencing, distributed data management and grid technologies as well as tangible interaction devices for visualization. High-speed, low-latency optical networks support high-quality collaborative interaction and remote visualization of large data.

The importance of locality in the visualization of large datasets

AbstractMany scientific phenomena in large high‐resolution datasets such as the U.K. Ocean Circulation and Advanced Modelling (OCCAM) ocean model are better discovered through visualization than by algorithmic analysis: it is often more straightforward to see a feature than it is to characterize it numerically. Using traditional rendering techniques, the size of modern datasets presents a challenge for even high‐end graphical supercomputers, and the cost of such hardware limits its availability for day‐to‐day analysis. We present an architecture that brings visual analysis to the desktop by exploiting consumer‐grade graphics hardware in order to provide initial interactive exploration and Web services to enable finer‐grained analysis and interoperability with traditional visualization tools. Copyright © 2006 John Wiley & Sons, Ltd.

Fishing for Biomarkers: Analyzing Mass Spectrometry Data with the New ClinProTools™ Software

Recently, applications of mass spectrometry in the field of clinical proteomics have gained tremendous visibility in the scientific and clinical community. One major objective is the search for potential biomarkers in complex body fluids like serum, plasma, urine, saliva, or cerebral spinal fluid. For this purpose, efficient visualization of large data sets derived from patient cohorts is crucial to provide clinical experts an interactive impression of the data quality. Additionally, it is necessary to apply statistical analysis and pattern matching algorithms to attain validated signal patterns that may allow for later applications in sample classification. We introduce the new ClinProTools bioinformatics software, which performs all major steps of profiling, screening, and monitoring applications in clinical proteomics. ClinProTools is the data interpretation software of the mass spectrometry-based ClinProt solutions for biomarker analysis. ClinProTools performs data pretreatment, visualization, statistics, pattern determination, pattern evaluation, and classification of spectra. This article will focus on ClinProTool's powerful and intuitive visualization options for clinical proteomics applications.

Large volume visualization of compressed time-dependent datasets on GPU clusters

We describe a system for the texture-based direct volume visualization of large data sets on a PC cluster equipped with GPUs. The data is partitioned into volume bricks in object space, and the intermediate images are combined to a final picture in a sort-last approach. Hierarchical wavelet compression is applied to increase the effective size of volumes that can be handled. An adaptive rendering mechanism takes into account the viewing parameters and the properties of the data set to adjust the texture resolution and number of slices. We discuss the specific issues of this adaptive and hierarchical approach in the context of a distributed memory architecture and present corresponding solutions. Furthermore, our compositing scheme takes into account the footprints of volume bricks to minimize the costs for reading from framebuffer, network communication, and blending. A detailed performance analysis is provided for several network, CPU, and GPU architectures—and scaling characteristics of the parallel system are discussed. For example, our tests on a eight-node AMD64 cluster with InfiniBand show a rendering speed of 6 frames per second for a 2048 × 1024 × 1878 data set on a 1024 2 viewport.

Multidetector CT of pancreatic adenocarcinoma: diagnostic advances and therapeutic relevance.

Detection and staging of pancreatic malignancies remains a challenge for radiologists. Considering the poor prognosis of pancreatic adenocarcinoma, accurate preoperative staging is the key to a possibly curative surgical treatment. Contrast-enhanced helical CT has been the most commonly used for evaluation of pancreatic cancer in many institutions, although it suffers from many limitations. With the fast pace of advances in multidetector CT (MDCT), and the beginning clinical implementation of 16-row scanners, improvements in spatial resolution in the z-axis with near-isotropic imaging provide exquisite multiplanar reconstructions of pancreatic anatomy. This article provides an overview of current MDCT technique and protocols for assessment of pancreatic adenocarcinoma, and describes new 3D-display methods for effective visualization of large data sets provided by MDCT.

The limits of realism: perceptions of virtual landscapes

Communicating planning results within expert groups or to local citizens is crucial to an efficient planning process. In the planning and design disciplines such as landscape and urban planning, recent digital 3D-visualizations have gained increased recognition. However, the validation of simulations of virtual landscapes, in terms of their degree of realism (R’Degree) has so far been neglected in research. This study concentrates on the question whether, how, and to which degree the real visually perceived landscape, represented through photographs, can be validly represented by means of virtual landscapes. The study area comprises the communities of Schwyz and Ingenbohl–Brunnen situated on Lake Lucerne in Central Switzerland. From a modelling point of view, landscapes are highly complex structures. Instead of manually modelling the virtual environment, which is the traditional CAD-approach, a GIS-based approach is pursued. This is the prerequisite for the efficient visualization of large data sets. The validity of the created virtual landscape is tested in an empirical study in which test persons are asked to order a set of real images and variations of the corresponding computer-generated images. In the experiment, approximately 75% of the test persons assigned the highest possible value (very high degree of realism) to one or more scenes of simulated landscapes. In order to achieve an even higher degree of realism, more and very detailed 3D-object-data and accompanying texture information would be necessary.