Thomas Schiemann

High quality rendering of attributed volume data

For high quality rendering of objects segmented from tomographic volume data the precise location of the boundaries of adjacent objects in subvoxel resolution is required. We describe a new method that determines the membership of a given sample point to an object by reclassifying the sample point using interpolation of the original intensity values and searching for the best fitting object in the neighbourhood. Using a ray-casting approach we then compute the surface location between successive sample points along the viewing-ray by interpolation or bisection. The accurate calculation of the object boundary enables a much more precise computation of the gray-level-gradient yielding the surface normal. Our new approach significantly improves the quality of reconstructed and shaded surfaces and reduces aliasing artifacts for animations and magnified views. We illustrate the results on different cases including the Visible-Human-Data, where we achieve nearly photo-realistic images.

Creating a high-resolution spatial/symbolic model of the inner organs based on the Visible Human.

Computerized three-dimensional models of the human body, based on the Visible Human Project of the National Library of Medicine, so far do not reflect the rich anatomical detail of the original cross-sectional images. In this paper, a spatial/symbolic model of the inner organs is developed, which is based on more than 1000 cryosections and congruent fresh and frozen CT images of the male Visible Human. The spatial description is created using color-space segmentation, graphic modeling, and a matched volume visualization with subvoxel resolution. It is linked to a symbolic knowledge base, providing an ontology of anatomical terms. With over 650 three-dimensional anatomical constituents, this model offers an unsurpassed photorealistic presentation and level of detail. A three-dimensional atlas of anatomy and radiology based on this model is available as a PC-based program.

Vascular Shape Segmentation and Structure Extraction Using a Shape-Based Region-Growing Model

A new, practical, and efficient approach is proposed for 3D vascular segmentation and bifurcation structure extraction. The method uses a combination of mathematical morphology, region-growing schemes, and shape features in addition to greyscale information. By an extension of math-morphological operations within bounded space of vascular shape, smooth and natural region-growing and sensitivity-controllable bifurcation detection were realized. The algorithm was implemented in the interactive segmentation and visualization software package VOXEL-MAN and validated with clinical data of X-ray CT angiography and MRA.

A 'virtual body' model for surgical education and rehearsal

A general digital model of human anatomy is very helpful both in supporting the process of anatomical segmentation and as a reference system for simulating surgical situations or even rehearsal of interventions. This article describes the data structure and implementation of such a model. Neither superhero nor crash-test dummy, Voxel-Man is an attempt to combine in a single framework a detailed spatial model enabling realistic visualization with a symbolic model of the human body. We show that although a general model does not correspond in detail to an individual patient, it does provide a variety of novel features for surgical education and training.

Exploring the Visible Human using the VOXEL-MAN framework

In principle the Visible Human data sets are an ideal basis for building electronic atlases. While it is easy to construct such atlases by just offering the possibility of browsing through the 2D slices, constructing realistic 3D models is a huge project. As one rather easy way to establish 3D use, we have registered the Visible Human data to the already existing 3D atlas VOXEL-MAN/brain. This procedure enables one to lookup anatomical detail in an atlas based on radiological images. Concerning the segmentation problem, which is the prerequisite for a real 3D atlas, we have developed an interactive classification method that delivers realistic perspective views of the Visible Human. As these volume based methods require high-end workstations, we finally have developed a multimedia program that runs on standard PCs and uses Quicktime VR movies.

Spatial Knowledge Representation for Visualization of Human Anatomy and Function

Computerized presentation of knowledge about human gross anatomy and function is presently being developed along two major tracks. Computer graphics on one hand provides more and more powerful tools for volume visualization. Knowledge engineering on the other hand provides more and more sophisticated data structures, while it does not care much for proper visualization of such knowledge. This paper presents a new approach of representing medical knowledge with a volume based data structure. This method has major advantages for true three-dimensional visualization of anatomical and functional knowledge, as well as for presenting knowledge of different domains simultaneously. The basic ideas and a pilot implementation for the human skull and brain are presented. Remaining problems and possibilities for future extensions and improvements are discussed.

Simulation of cardiac excitation patterns in a three-dimensional anatomical heart atlas

Computerized anatomical atlas systems enable interactive investigation of digital body models. Here we present a three-dimensional atlas of the human heart, based on image data provided in the Visible Human Project. This heart atlas consists of multiple kinds of cardiac tissues and offers unlimited possibilities for its visual exploration. A temporal dimension is added to the underlying heart model by simulation of cardiac excitation spreading. For this purpose a second generation cellular automata algorithm is adapted to the excitation kinetics of cardiac tissue. The presented system is shown as a successful method for the visualization-based investigation of cardiac excitation.

Segmentation of the Visible Human for High Quality Volume Based Visualization

This article describes a combination of interactive classification and super-sampling visualization algorithms that greatly enhances the realism of 3-D reconstructions of the Visible Human data sets. Objects are classified on the basis of ellipsoidal regions in RGB space. The ellipsoids are used for super-sampling in the visualization process.

A Realistic Model of the Inner Organs from the Visible Human Data

The computer-based 3D models of the human body reported to date suffer from poor spatial resolution. The Visible Human project has delivered high resolution cross-sectional images that are suited for generation of high-quality models. Yet none of the 3D models described to date reflect the quality of the original images. We present a method of segmentation and visualization which provides a new quality of realism and detail. Using the example of a 3D model of the inner organs, we demonstrate that such models, especially when combined with a knowledge base, open new possibilities for scientific, educational, and clinical work

Interactive 3D-Segmentation of Tomographic Image Volumes

Segmentation is a prerequisite for 3D visualization of image volumes. It has turned out to be extremely difficult to formalize for automatic computation. We describe an interactive segmentation method that circumvents this difficulty by using low level segmentation tools, which are interactively controlled by a human user via 3D display. Segmentation tools implemented so far are simple thresholding and morphological operations. The method has been implemented on a workstation under UNIX using an X-Window interface based on the OSF/MOTIF toolkit. It is shown with examples from different applications that this simple approach delivers good results in only a short amount of time.