Alexandra Baer

Perceptual evaluation of ghosted view techniques for the exploration of vascular structures and embedded flow

This paper presents three controlled perceptual studies investigating the visualization of the cerebral aneurysm anatomy with embedded flow visualization. We evaluate and compare the common semitransparent visualization technique with a ghosted view and a ghosted view with depth enhancement technique. We analyze the techniques’ ability to facilitate and support the shape and spatial representation of the aneurysm models as well as evaluating the smart visibility characteristics. The techniques are evaluated with respect to the participants accuracy, response time and their personal preferences. We used as stimuli 3D aneurysm models of five clinical datasets. There was overwhelming preference for the two ghosted view techniques over the semitransparent technique. Since smart visibility techniques are rarely evaluated, this paper may serve as orientation for further studies.

Hardware-accelerated stippling of surfaces derived from medical volume data

We present a fast hardware-accelerated stippling method which does not require any preprocessing for placing points on surfaces. The surfaces are automatically parameterized in order to apply stippling textures without major distortions. The mapping process is guided by a decomposition of the space in cubes. Seamless scaling with a constant density of points is realized by subdividing and summarizing cubes. Our mip-map technique enables arbitrarily scaling with one texture. Different shading tones and scales are facilitated by adhering to the constraints of tonal art maps. With our stippling technique, it is feasible to encode all scaling and brightness levels within one self-similar texture. Our method is applied to surfaces extracted from (segmented) medical volume data. The speed of the stippling process enables stippling for several complex objects simultaneously. We consider application scenarios in intervention planning (neck and liver surgery planning). In these scenarios, object recognition (shape perception) is supported by adding stippling to semi-transparently shaded objects which are displayed as context information.

A Survey of Perceptually Motivated 3D Visualization of Medical Image Data

This survey provides an overview of perceptually motivated techniques for the visualization of medical image data, including physics‐based lighting techniques as well as illustrative rendering that incorporate spatial depth and shape cues. Additionally, we discuss evaluations that were conducted in order to study the perceptual effects of these visualization techniques as compared to conventional techniques. These evaluations assessed depth and shape perception with depth judgment, orientation matching, and related tasks. This overview of existing techniques and their evaluation serves as a basis for defining the evaluation process of medical visualizations and to discuss a research agenda.

Hardware-Accelerated Illustrative Medical Surface Visualization with Extended Shading Maps

In this paper, we introduce a new framework for the illustrative visualization of medical surface data. In most visualization frameworks, only light intensity is used to determine the surface shading. The analysis of medical textbooks reveals more complex shading approaches. The parameters of these approaches are mapped to different Shading Maps, which may be weighted and flexibly combined. We discuss the use of high-level attributes to simplify the specification. The resulting Shading Mapis used as a lookup to determine the final intensity at a certain area. For this purpose, the rendering is accomplished on GPU by using OpenGLs Framebuffer Objects. This framework may be useful for interactive educational systems or for medical record printings.

Comparative Evaluation of Feature Line Techniques for Shape Depiction

This paper presents a qualitative evaluation of feature line techniques on various surfaces. We introduce the most commonly used feature lines and compare them. The techniques were analyzed with respect to the degree of realism in comparison with a shaded image with respect to the aesthetic impression they create. First, a pilot study with 20 participants was conducted to make an inquiry about their behavior and the duration. Based on the result of the pilot study, the final evaluation was carried out with 129 participants. We evaluate and interpret the trial results by using the Schulze method and give recommendations for which kind of surface, which feature line technique is most appropriate.

Curvature- and Model-Based Surface Hatching of Anatomical Structures Derived from Clinical Volume Datasets

We present a texture-based method to hatch anatomical structures derived from clinical volume datasets. We consider intervention planning, where object and shape recognition are supported by adding hatching lines to the anatomical model. The major contribution of this paper is to enhance the curvature-based hatching of anatomical surfaces by incorporating model-based preferential directions of the underlying anatomical structures. For this purpose, we apply our method on vessels and elongated muscles. Additionally, the whole hatching process is performed without time-consuming user interaction for defining stroke direction and surface parameterization. Moreover, our approach fulfills requirements for interactive explorations like frame coherence and real time capability.

Sketching 2D Vessels and Vascular Diseases with Integrated Blood Flow

In this paper, we present a sketch-based interface, which allows medical doctors to illustrate different vascular diseases and treatment methods as well as fluid behavior. With this sketching interface, we provide the physician with an effective tool to illustrate different medical cases, which is important in the complex field of vascular diseases with respect to patient education. We use techniques from sketch-based interfaces and GPU-based computational fluid dynamics by considering usability aspects. We provide a concept and a prototypical implementation whose usability is tested with quantitative and qualitative methods. Additionally, we interviewed a physician to assess the benefits of the tool with respect to patient education.

A Sketch-Based Interface for 2D Illustration of Vascular Structures, Diseases, and Treatment Options with Real-Time Blood Flow

We present a sketching interface, which enables physicians to illustrate various vascular structures, diseases, and treatment options with integrated blood flow. This sketch-based interface provides medical doctors with an effective tool to illustrate different medical scenarios and support patient education. This work integrates methods from sketch-based interfaces and GPU-supported computational fluid dynamics. The usability of the prototype was assessed qualitatively and quantitatively. Additionally, we performed a structured interview with a physician to evaluate the benefits with respect to patient education. The results of the evaluation confirmed the usability of the prototype as well as the usefulness to support physicians during the process of patient education.

Importance-driven structure categorization for 3D surgery planning

We present an importance-driven categorization approach to automatically gather all currently required structures for the surgery planning process. Therefore, we analyzed common demands for tumor intervention planning and integrated domain knowledge to enable a determination of the relevant structures for various surgical questions. The categorization of structures in focus, focus-relevant and context is defined and initiated by the question. Our method uses the structures specific meta data and geometric information to determine an importance value for each structure automatically. This importance value encodes the structures priority for the current question and defines the structures category. Furthermore, this value can be used to define a structure-specific visual style to generate expressive 3D surgery planning visualizations.

Hardwaregestütztes Stippling von medizinischen Oberflächenmodellen

In dieser Arbeit wird ein texturbasiertes Verfahren zur hardwaregestutzten Stippling-Darstellung vorgestellt, welches auf beliebige medizinische Oberflachen angewendet werden kann. Unter Verwendung von Shadern werden Stippling-Texturen auf die Oberflachen aufgebracht und eine echtzeitfahige 3D-Visualisierung generiert. Ferner ermoglicht das Verfahren eine Stippling-Darstellung der Oberflachen ohne notwendige Vorverarbeitungsschritte und unabhangig von der Komplexitat des Objektgitters. Die Anforderungen an traditionelle Stippling-Illustrationen sowie an interaktive NPR-Darstellungen, wie Frame-Koharenz und Skalierbarkeit, wurden berucksichtigt und in der Umsetzung integriert.