In recent years, high-resolution displays have become increasingly important to decision makers and scientists because large screens combined with a high pixel count facilitate content rich, simultaneous display of computer-generated imagery and high-definition video data from multiple sources. Tiled displays are attractive due to their extended screen real estate, scalability, and low cost. LCD panels are usually preferred over projectors because of their superior resolution. One of the drawbacks of LCD-based tiled displays is the fact that users sometimes get distracted by the screens' bezels, which cause discontinuities in rendered images, animations, or videos. Most conventional solutions either ignore the bezels and display all pixels, causing objects to become distorted, or eliminate the pixels that would normally fall under the bezels, causing pixels to be missing in the display of static images. In animations, the missing pixels will eventually reappear when the object moves, providing an experience that is similar to looking through a French window. In this paper, we present a new scalable approach that leads neither to discontinuities nor to significant loss of information. By projecting onto the bezels, we demonstrate that a combination of LCD-based tiled displays and projection significantly reduces the bezel problem. Our technique eliminates ambiguities that commonly occur on tiled displays in the fields of information visualization, visual data analysis, human-computer interaction, and scientific data display. It improves the usability of multimonitor systems by virtually eliminating the bezels. We describe a setup and provide results from an evaluation experiment conducted on a 3 x 3 and on a 10 x 5 tiled display wall.

High Content Screening (HCS) is a method for analyzing large numbers of cells typically arranged in arrays. Multi-well assays and micro-pallet arrays can be used to single out and harvest cells with desired features. However, visual inspection of large time series of images, each containing hundreds of wells or tens of thousands of micro-pallets is time-consuming and prone to detection errors.We have developed cell Analyst, a publicly available software package for automated identification and counting of cells. In this article we outline the image processing methods behind cell Analyst and show how this software can be applied to conduct highly specific cell counting.

This paper describes an efficient image transformation method based on histogram information and some prior knowledge of tissue expression in different modalities for registration of images from different image aquisition methods, such as CT and MRI. The metric defined in this paper represents a computationally efficient alternative to traditional, mutual information based methods. The optimization step in the registration algorithm can be carried out either by using a normalized cross-correlation function or as a simple image subtraction, which makes it suitable for hardware implementations.

Interactive virtual reality applications commonly require two key technologies: multiple degree-of-freedom input devices, and 2D or 3D displays. The industry has developed a vast variety of devices for a growing consumer market. Consumer magazines regularly publish test reports for new devices. These reports are often focused on the gaming community, which is typically the driving force behind new product development. Although many lessons can be learned from the gaming industry, the scientific community is generally focused on other criteria, such as precision, degrees of freedom, and user tracking. It is expected that some of these criteria, which are currently in the state of research, will eventually be incorporated into products for a mass market, just like consumer graphics cards and certain input devices did in the past.

Large-scale biomedical data sets of macromolecular structures such as DNA and proteins describe highly complex biomolecular entities which often consist of thousands of atoms and residues in large 3D strands of amino acids. Various types of abstract representations are used to display these data sets, ranging from traditional ball-and-stick models to feature-presenting cartoon models. Despite these abstractions, even the most simplified representations still generate a large amount of geometry data. Useful features are often difficult to recognize in the sheer amount of detail produced by the rendering software. Comparative visualization of various structures in any type of abstract representation is helpful for understanding the relation between function and structure, and prediction of properties of newly discovered proteins which might have an impact on drug design and better treatment options for diseases. We present a case study that uses high performance workstations and a virtual reality display system to process large amounts of geometry data for real-time 3D exploration, superimposition, and interactive navigational tasks in a virtual reality environment.

Coronary heart disease (CHD) is the number one killer in the United States. Although it is well known that CHD mainly occurs due to blocked arteries, there are contradictory results from studies designed to identify basic causes for this common disease is. To find out more about the true reason for CHD, virtual models can be employed to better understand the way the heart functions. With such a model, scientists and surgeons are able to analyze the effects of different treatment options, and ultimately find more suited ways to prevent coronary heart diseases. To investigate a given model, appropriate navigation methods are required, including suitable input devices. For the visualization, graphics cards originally designed for gaming applications are used; so, it is a just natural transition to adapt gaming input devices to a visualization system for controlling of the navigation. These devices are usually well designed with respect to ergonomics and durability, yielding more degrees of freedom in steering than two-dimensional input devices, such as desktop mice. This poster describes a visualization system that provides the user with advanced control devices for navigation enabling interactive exploration of the model. Force-feedback and sound effects provide additional cues.

Glial cells express a large variety of bio-electric and bio-chemical phenomena. A quantitative analysis shows that the brain consists of about 10 to 50 times more glial cells than neurons. Similar to neurons, glial cells implement voltage-dependent and ligand-gated membrane channels. Neurotransmitter receptors associated with ion-selective membrane channels have been described in several types of macroglial cells. Leech giant glial cells provide a good example for studying structural and physiological properties of this cell type. Confocal laser-scanning microscopy generates images from cutting (i.e. focus) planes of the object, in a way that, by varying the focus plane, the object can be resolved and scanned serially at several levels. Artifacts, such as transparent, low contrast, and blurred or fuzzy structures from neighboring slices must be filtered in order to reveal the capillary structure of the cell. With the microscope, the object cannot be rotated by ninety degrees, perpendicular to the viewing direction. Once scanned, the viewing direction cannot be altered any more. Computer graphics allows projections from any perspective angle, so that the cell can be studied in full detail. A visualization system should provide a flexible environment for interactively exploring the cell. Therefore, we have extended our InVIS (Interactive Visualization) system, which has been originally designed for medical applications (CT and MRI), to confocal microscopy. We describe the structure of our system and the addition of new features.