1,144 research outputs found
Parga, and the Ionian islands; comprehending a refutation of the mis-statements of the Quarterly Review and of Lieut.- Gen. Sir Thomas Maitland, on the subject; with a report of the trial between that officer and the author.
Preface: Bosset, C.P.(de)Appendix.Dedication:Content description: TitleIllustration: (Maps ,)Pagination: PP26+530PVolumes: 1Text Genre:ProseIllustration: (χάρτες ,
Curvature Lines for Lesion Detection and Visualization in CT Colonography
In this thesis, we present an automatic polyp detection approach that integrates knowledge from flow visualization techniques. Our primary goal was to compute additional characteristic polyp features that improve the CAD performance of existing polyp detection approaches by eliminating found false-positive polyp candidates. We found that surface principal curvature directions presented discriminating patterns on surface areas that belong to colonic polyps. These could be visualized using lines of curvatures. We developed novel approaches for generating lines of curvature on the colon wall. In order to provide sufficient shape information, lines of curvature were distributed using curvature-adaptive streamline seeding and spacing strategies. Geometric features of lines of curvature were used to differentiate between true polyp detections and false positives. The visualization of the colon wall was also enhanced by visualizing lines of curvature.MediamaticsElectrical Engineering, Mathematics and Computer Scienc
Visualization of Variation and Variability
As datasets grow in size and complexity, the importance of comparison as a tool for analysis is growing. We define comparison as the act of analyzing variation or variability based on two or more specific instances of the data. This thesis explores a number of cases spread across the range of comparisons, from variability within one entity through variability between two or more entities to variability within a population. For each of these we present an exploration tool, combining interaction with high-performance visualization rendering techniques, with the aim of providing more insight into a given dataset. We explore how different aspects of an application designed for interactive visual analysis can aid the user. This concerns both their initial exploration of a new dataset, as well as their ability to drill down into their discoveries and investigate the underlying details. For instance, multiple linked views can be used to combine highly abstract general-purpose views with highly problem-domain specific views in order to allow a user to translate abstract discoveries into the specific concepts used in their profession. Interactive composition can be applied to quickly focus on areas of interest, suppressing details which may not be relevant at the moment.Computer Graphics and VisualizationElectrical Engineering, Mathematics and Computer Scienc
Direct and comparative visualization techniques for HARDI Data
DWI is an MRI imaging technique used to gain information concerning the diffusion process in tissue. Using DTI techniques, a diffusion profile can be constructed for fiber tract analysis. Recently developed HARDI techniques increase the detail to visualization on the process of diffusion. While HARDI reconstruction methods are used to model the underlying diffusion process, the HARDI signal attenuation data can be used for a better understanding of noise in DWI data. This project addresses the direct visualization of HARDI data without any intermediate processing steps between acquisition and visualization. We present new glyph shapes for direct and comparative visualization of HARDI data using the signal attenuation or ADC and a multiple linked views layout. We developed new difference metrics to create a complete comparative visualization pipeline to identify and explore areas of interest. Evaluation of our developed methods by means of a case study, indicates the techniques to be a valued addition. The comparative visualization allows for quick identification of areas of interest. The glyph representation allows for rapid exploration of local diffusion data.Computer GraphicsMediamaticsElectrical Engineering, Mathematics and Computer Scienc
A Client/Server Framework for Interactive Remote 3D Visualization of Histological Data
The amount of data involved in medical volumes, especially in the case of volumes consisting of stacks of histological sections, tends to be very large. Despite rapid advancements in computer processing power and storage- and memory capacity in the past years these volumes are simply too large to be transferred to and visualized on regular workstations. At the same time Internet connectivity is becoming common and connections are getting faster every day. In this situation a client-server remote visualization can be used to hand off visualizations to (clusters of) dedicated, well-equipped, servers which can do the heavy work and send the result, or a part thereof, back to the client when ready. In this project we developed a client-server framework for remote visualization of histological data. The framework is designed to be scalable and effectively utilize hardware resources from multiple servers for each visualization pipeline. The framework uses the concept of strategies to assign resources to each visualization and to distribute these resources over the available hardware according to predefined rules. To manage the available data we designed a data scheme in which multiple objects belonging to a certain dataset, we will define these objects as modalities, can be stored in a relational database. A key feature of our solution is that the scheme allows appropriate visualizations of the modalities to be stored along with the data in the same relational database. Additionally, as part of this project, we developed a reference client application which is able to use the framework to do remote visualization. This application was purposely designed to work on many platforms without the need for state-of-the-art hardware or software. The application currently offers support for visualizations that require viewing of images, viewing of planes extracted from volumes and viewing of three-dimensional scenes under control by a camera. Both the framework and the reference client application are designed to be extensible and will be made available to the public domain so that the software can be used freely and adapted to future demands whenever needed.Mediamatics, Computer Graphics & CAD/CAMElectrical Engineering, Mathematics and Computer Scienc
Direct Visualization of Photographic Volumes
Volume visualization of photographic data is a relatively new domain for medical visualisation. In photographic volumes, voxel colour is predetermined, making colour selection through transfer functions less important, or even unnecessary. However, voxel classi?cation from photographic data is more complex than classi?cation from scalar datasets. Materials and structures in the human body overlap in the colour-space, and are therefore not separable solely by their colour values. In addition, low contrast between some of the adjacent materials and structures in the data complicates the use of boundary-based classi?cation techniques. In the ?rst half of this work we investigate the differences between volume visualization of scalar datasets and volume visualization of photographic datasets. We study a broad range of techniques which could be extended for photographic data, or replace traditional techniques in the volume render pipeline. In addition we study normal estimation for photographic data, which suffers from the non-linear colour spaces and previously mentioned low contrast. In the second half of this work we present an example-based visualization system which does not require its user to understand the mechanics of volume visualization. Our target group are non-visualization experts who want to use volume visualization to provide insight into their data. The system uses live rendered style previews which can be combined to form a composite render style. The system supports multiple photorealistic as well as illustrative render techniques and can apply any combination of these techniques to any object in the dataset, rendering comprehensible images. The effectiveness of our system is demonstrated by experimental results on real volumetric data.MediamaticsComputer GraphicsElectrical Engineering, Mathematics and Computer Scienc
The Unified Anatomical Human: Model-based Visualization of Heterogeneous Anatomy Data
In the course of anatomical research, anatomists acquire and attempt to organize a great deal of heterogeneous data from different sources, such as MRI and CT data, cryosections, immunohistochemistry, manual and automatic segmentations of various structures, related literature, the relations between all of these items, and so forth. Anatomical variation between subjects further complicates this organization. Currently, there is no way of storing, accessing and visualizing these heterogeneous datasets in an integrated fashion. Such capabilities would have great potential to empower anatomy research. In this work, we present methods for the integration of heterogeneous spatial and non-spatial data from different sources, as well as the complex relations between these, into a single model with standardized anatomical coordinates: The Unified Anatomical Human. All captured data can then be interactively visualized in various ways, depending on the anatomical question. Furthermore, our model enables data to be queried both structurally, i.e., relative to existing anatomical structures, and spatially, i.e., with anatomical coordinates. When new patient-specific medical scans are added to the model, all available model information can be mapped to them. Using this mapping, model information can be transferred back to the new scans, thus enabling the creation of visualizations enriched with information not available in the scans themselves.Computer Graphics and Visualization GroupComputer Science - Media and Knowledge EngineeringElectrical Engineering, Mathematics and Computer Scienc
Hierarchatomy: Modelling and Automatic 2D Visualization of Branching Anatomical Structures
Current anatomical learning material usually consists of static hand drawn images accompanied by detailed descriptions. In medical education and research 3D anatomical systems are also extensively used. The challenge of these media is how can as much information as possible be conveyed in a relatively small illustration or viewport. We present a model that automatically visualizes branching anatomical structures like the vascular or nervous system in such a way that the 2D drawing space is optimally utilized. A layout algorithm based on simulated annealing optimizes the layout of the structures. The visualization is generated from an underlying data representation that manages a wide range of relations and properties of the branching structures. The illustrations of Instant Anatomy (Robert Whitaker) formed the basis of the research. The system is presented in a user friendly web application that provides the user a way to add information to the model and navigate through the structures.Computer Graphics and Visualization GroupIntelligent SystemsElectrical Engineering, Mathematics and Computer Scienc
Techniques and software architectures for medical visualisation and image processing
This thesis presents a flexible software platform for medical visualisation and image processing, a technique for the segmentation of the shoulder skeleton from CT data and three techniques that make contributions to the field of direct volume rendering. Our primary goal was to investigate the use of visualisation techniques to assist the shoulder replacement process. This motivated the need for a flexible environment within which to test and develop new visualisation and also image processing techniques with a medical focus. The Delft Visualisation and Image processing Development Environment, or DeVIDE, was created to answer this need. DeVIDE is a graphical data-flow application builder that combines visualisation and image processing techniques, supports the rapid creation of new functional components and facilitates a level of interaction with algorithm code and parameters that differentiates it from similar platforms. For visualisation, measurement and pre-operative planning, an accurate segmentation from CT data of the bony structures of the shoulder is required. Due to the complexity of the shoulder joint and the fact that a method was required that could deal with diseased shoulders, existing techniques could not be applied. In this thesis we present a suite of techniques for the segmentation of the skeletal structures from CT data, especially designed to cope with diseased shoulders. Direct volume rendering, or DVR, is a useful visualisation technique that is often applied as part of medical visualisation solutions. A crucial component of an effective DVR visualisation is a suitable transfer function that assigns optical characteristics to the data. Finding a suitable transfer function is a challenging task. We present two highly interactive methods that facilitate this process. We also present a method for interactive direct volume rendering on ubiquitous low-end graphics hardware. This method, called ShellSplatting, is optimised for the rendering of bony structures from CT data and supports the hardware-assisted blending of traditional surface rendering and direct volume rendering. This characteristic is useful in surgical simulation applications. ShellSplatting is based on the object-order splatting of discrete voxels. As such, maintaining a correct back-to-front or front-to-back ordering during rendering is crucial for correct images. All existing real-time perspective projection visibility orderings show artefacts when splatting discrete voxels. We present a new ordering for perspective projection that remedies these artefacts without a noticeable performance penalty.Electrical Engineering, Mathematics and Computer Scienc
Particle-based Visualization of Soft-Tissue: Deformation in the Human Orbit
Orbital fat is the soft tissue inside the eye socket, or orbit, that surrounds the eye muscles, eyeball, optic nerve and blood vessels. It plays an important role in the mechanical operation of the eye during eyeball rotation. Knowledge of how this orbital fat behaves is limited. It is important for surgeons to improve this knowledge to make better informed decisions before and during eye surgery. Visualization of the deformation of orbital fat can improve this knowledge. Herein lie two problems. First, we do not have accurate and detailed information on how the orbital fat deforms. Second, we do not have a visualization method that provides the necessary insight into orbital fat deformation data. We obtain deformation data by deriving it from a sequence of MRI volumes that represent the rotating eyeball. We do this by using a method from the optic ow family, which uses the changing tissue boundaries in the MRI data to locally extract deformation data. We visualize this deformation data using a particle-based method. In particle-based visualization, a large number of tiny, point-like particles are placed inside part of the volume. The positions of individual particles are affected by the deformation data. We smoothly animate the particles as they move through the orbit while the eyeball rotates, giving a detailed impression of the deformation of a subvolume of orbital fat. Demarcation of where inside the volume to place the particles is called region of interest (ROI) specification. We developed a method that uses anatomical information from the MRI volumes to specify sensible ROIs. We implemented our solutions in two software tools that work together to allow quick exploration of deformation from a sequence of MRI volumes. Experts in the fields of ophthalmology and biomechanics have expressed that our visualization software is a significant improvement over previous techniques and that the software should continue to be developed.Computer GraphicsMediamaticsElectrical Engineering, Mathematics and Computer Scienc
- …
