Published online December 5, 2002, 10.1148/rg.e12
(Radiographics. 2003;23:e12-e12.)
© RSNA, 2003
Standardized Evaluation of CT Angiography with Remote Generation of 3D Video Sequences for the Detection of Intracranial Aneurysms
Bernd F. Tomandl, MD,
Peter Hastreiter, Dr Ing,
Sabine Iserhardt-Bauer, Dipl Inf,
Niels C. Köstner,
Miriam Schempershofe,
Walter J. Huk, MD,
Thomas Ertl, PhD,
Christian Strauss, MD and
Johann Romstock, MD
1 From the Departments of Neuroradiology (B.F.T., N.C.K., M.S., W.J.H.) and Neurosurgery (P.H., C.S., J.R.), University of Erlangen-Nuremberg, Germany; and the Visualization and Interactive Systems Group, University of Stuttgart, Germany (S.I-B., T.E.). Presented as a scientific exhibit at the 2001 RSNA scientific assembly. Received September 18, 2002, revision requested October 22, revision received and accepted October 23. Address correspondence to B.F.T. (e-mail: tomandl@neuroradiologie-erlangen.de).
 |
Abstract
|
|---|
Computed tomography (CT) angiography is a well-known imaging technique commonly applied to both the detection and therapy planning of intracranial aneurysms. For this purpose, current studies predominantly focus on three-dimensional (3D) representations of CT angiographic volumes obtained with varying visualization approaches on different computers. Interactive manipulation performed by users individually is an important prerequisite for data analysis. However, this leads to inconsistent and barely reproducible 3D visualization results. Furthermore, the quality of any 3D representation depends on the applied visualization strategy (eg, maximum-intensity projection, shaded-surface display, direct volume rendering). To overcome these limitations, the authors present a novel method for standardized visualization of CT angiographic volumes, consisting of three steps: (a) transfer of the image data to a remote high-end graphics workstation, (b) automatic 3D visualization with high-resolution direct volume rendering, and (c) consecutive video generation performed according to a standardized protocol. The recorded video sequences are transferred for evaluation to a local desktop computer. In the experimental setup, high-quality videos based on 3D visualizations were produced in less than 60 minutes per patient. Although aneurysms above the skull base are usually visualized with excellent quality, the analysis of aneurysms at the skull base is still difficult.
© RSNA, 2002
Index Terms: Aneurysm, CT, 17.12112, 17.12115, 17.12116, 17.12117 • Aneurysm, intracranial, 17.73 • Cerebral blood vessels, CT, 17.12112, 17.12115, 17.12116, 17.12117 • Computed tomography (CT), three-dimensional, 17.12117
|
Introduction
|
|---|
Sudden onset of severe headache is the typical leading symptom in patients with subarachnoid hemorrhage. The cause of subarachnoid hemorrhage is a ruptured aneurysm in 75%ndash;85% of cases, nonaneurysmal perimesencephalic hemorrhage in about 10% of cases, and a variety of rare conditions such as arteriovenous malformations in about 5% of cases (1). Computed tomography (CT) is the first step in the examination of these patients. If subarachnoid hemorrhage is confirmed, it is necessary to detect the source of bleeding for appropriate therapy. While digital subtraction angiography (DSA) is still the most sensitive tool for the detection of intracranial aneurysms and other vascular malformations, many studies have shown the usefulness of CT angiography for this purpose. The reported sensitivity of CT angiography has been reported to be in the range of 70%ndash;96% (2–7), depending on the size and location of the aneurysm (8). In most of these studies three-dimensional (3D) visualizations of the intracranial vessels were used to asses the value of CT angiography. In this technique, the section images are transferred to a workstation and 3D visualization of vascular structures is performed. The aim of 3D visualization is to extract the vascular structure data from the volume. These methods always lead to loss of information, regardles of the applied algorithm. Although the suppression of brain tissue surrounding the intracranial vessels is a prerequisite for providing an open view of the intracranial arteries in a 3D representation, it is possible that important information such as calcification or thrombosed parts of an aneurysm is not seen. Therefore, it is mandatory to accurately study the section images before 3D visualization is performed (Fig 1).
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Comparison of section images from CT angiography to 3D representation in a patient with severe subarachnoid hemorrhage and aneurysms of both internal carotid artery (ICA) bifurcations. The left aneurysm is calcified. (a) Axial section image at the level of the circle of Willis shows an aneurysm at the bifurcation of the right ICA (arrow). The basal cisterns are filled with subarachnoid blood (arrowheads). For a better comparison with the 3D image,the view is from above. (b) Axial section image at the level of a calcified aneurysm of the left ICA bifurcation (arrow). The extent of the intramural calcification is clearly demonstrated (arrowheads). Again, the view is from above. (c) Superoinferior 3D visualization with direct volume rendering shows the two aneurysms (arrows) in relation to the intracranial arteries and the skull base. The intramural calcification within the left aneurysm is demonstrated, but its extent is not clearly seen (arrowhead). Note that brain tissue and subarachnoid blood are not demonstrated within the 3D representation because the selected thresholds do not include the voxels containing this information.
|
|
View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Comparison of section images from CT angiography to 3D representation in a patient with severe subarachnoid hemorrhage and aneurysms of both internal carotid artery (ICA) bifurcations. The left aneurysm is calcified. (a) Axial section image at the level of the circle of Willis shows an aneurysm at the bifurcation of the right ICA (arrow). The basal cisterns are filled with subarachnoid blood (arrowheads). For a better comparison with the 3D image,the view is from above. (b) Axial section image at the level of a calcified aneurysm of the left ICA bifurcation (arrow). The extent of the intramural calcification is clearly demonstrated (arrowheads). Again, the view is from above. (c) Superoinferior 3D visualization with direct volume rendering shows the two aneurysms (arrows) in relation to the intracranial arteries and the skull base. The intramural calcification within the left aneurysm is demonstrated, but its extent is not clearly seen (arrowhead). Note that brain tissue and subarachnoid blood are not demonstrated within the 3D representation because the selected thresholds do not include the voxels containing this information.
|
|
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c. Comparison of section images from CT angiography to 3D representation in a patient with severe subarachnoid hemorrhage and aneurysms of both internal carotid artery (ICA) bifurcations. The left aneurysm is calcified. (a) Axial section image at the level of the circle of Willis shows an aneurysm at the bifurcation of the right ICA (arrow). The basal cisterns are filled with subarachnoid blood (arrowheads). For a better comparison with the 3D image,the view is from above. (b) Axial section image at the level of a calcified aneurysm of the left ICA bifurcation (arrow). The extent of the intramural calcification is clearly demonstrated (arrowheads). Again, the view is from above. (c) Superoinferior 3D visualization with direct volume rendering shows the two aneurysms (arrows) in relation to the intracranial arteries and the skull base. The intramural calcification within the left aneurysm is demonstrated, but its extent is not clearly seen (arrowhead). Note that brain tissue and subarachnoid blood are not demonstrated within the 3D representation because the selected thresholds do not include the voxels containing this information.
|
|
Little is known about the influence of postprocessing methods such as maximum-intensity projection (MIP), shaded-surface display, and direct volume rendering (9–11). However, it can be assumed that the same CT angiography data may result in different detection rates when various visualization strategies, computer platforms, and graphics hardware are used (8,12–15). In addition, the user of the workstation must be familiar with several manipulation tools. These comprise the tools that position objects such as clipping planes, define thresholds, and adjust so-called transfer functions for color and opacity values. Rather than investigating the value of CT angiography for the evaluation of intracranial aneurysms, available studies have been analyzing specific systems. In general, these have consisted of the investigation modality, such as CT or magnetic resonance (MR) angiography, as the source of the data; the available workstation, including rendering software; and finally, an investigator steering the 3D visualization process. Figure 2 illustrates the typical workflow in such a system.
Reproducible protocols containing information about the section thickness, the injection rate of contrast medium, and the reconstruction of the section images are usually standard in these investigations. A standardized visualization method that provides reproducible 3D representations of CT angiographic data requires the use of a unique rendering approach that produces automatic and user-independent results. For this purpose, we present a Web-based method using a remotely controlled high-end graphics computer for image processing after transfer of the CT angiographic data via the Internet. This method results in a modified user-independent workflow (Fig 3) that supports a more objective evaluation.
|
Materials and Methods
|
|---|
CT Angiography
The CT studies were performed with a multisection spiral CT scanner (Somatom VZ; Siemens Medical Solutions, Forchheim, Germany). One hundred milliliters of nonionic iodinated contrast material (Ultravist 300; Schering, Berlin, Germany) was injected with a power injector at a flow rate of 4 mL/sec. The bolus tracking method (16) integrated within the scanner was used for bolus timing, with a region of interest placed within the ICA below the skull base. The spiral scan was started when 100 HU was reached within the region of interest. A collimation of 4 x 1 mm with a table feed of 2.7 mm was used for the spiral scan, which covered the area from the first cervical vertebra to the corpus callosum. The image reconstruction was performed with a section thickness of 1.25 mm, an increment of 0.5 mm, and a field of view of 120 mm2. Thus, the circle of Willis and the posterior inferior cerebellar arteries were completely included within the volume. The resulting voxel size was 0.23 x 0.23 x 0.5 mm3. The applied narrow field of view of 120 mm2 results in an excellent in-plane resolution and reproduces all relevant information. Additionally, lateral parts of the skull are already eliminated, making postprocessing easier.
Generation of the Video Sequences
"Flight Path"
Before the development of a software tool for the generation of videos from automatic 3D visualizations, a "flight path" was designed to analyze the relevant vascular structures in a comprehensive and efficient way. At the start, an overview shows the entire vessel configuration contained within the CT angiographic volume. Next, a detailed inspection of all locations in which intracranial aneurysms most likely occur is performed. As a basis for the design of this flight path, data from the literature were used and supplemented with experience gained from the analysis of data from 50 patients studied with both CT angiography and DSA in our department within the previous 2 years.
More than 80% of aneurysms are located within the anterior part of the circle of Willis (17). Typical locations are the anterior communicating artery and the bifurcations of the middle cerebral arteries (MCAs). In addition, aneurysms are frequently found in the ICA in regions in which smaller vessels branch off or at the intracranial bifurcation. The tip of the basilar artery and the region in which the posterior inferior cerebellar arteries branch off are typical locations for aneurysms in the posterior circulation.
The flight path was developed on the basis of the results from interactive 3D visualizations we performed with every CT angiographic volume obtained in our 50 patients. We used a volume-rendering technique that included the following steps: (a) For a good overview representation of the relevant vessels, superimposed structures such as the straight sinus and adjacent venous vascular structures must be removed. For this purpose, a clipping plane is applied parallel to the clivus at a distance of about 3 cm (Fig 4). (b) The best overview of the vertebral and basilar arteries is obtained from a posteroanterior direction parallel to the clivus (Fig 4c). (c) For a comprehensive understanding of the MCA bifurcation, which is characterized by considerable anatomic variation, multiple views from various directions are necessary.
View larger version (158K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. (a) 3D visualization of intracranial arteries from a posterosuperior view parallel to the clivus. The basilar artery (small arrow) is not visible because of overlying venous structures (large thin arrow). An aneurysm is seen within the left MCA bifurcation (large thick arrow). (b) Venous structures are eliminated by a clipping plane (yellow dotted line) parallel to the clivus. Left posterolateral view. (c) Application of the clipping plane allows an open view of the basilar artery (arrow) in the posterosuperior direction.
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. (a) 3D visualization of intracranial arteries from a posterosuperior view parallel to the clivus. The basilar artery (small arrow) is not visible because of overlying venous structures (large thin arrow). An aneurysm is seen within the left MCA bifurcation (large thick arrow). (b) Venous structures are eliminated by a clipping plane (yellow dotted line) parallel to the clivus. Left posterolateral view. (c) Application of the clipping plane allows an open view of the basilar artery (arrow) in the posterosuperior direction.
|
|
View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. (a) 3D visualization of intracranial arteries from a posterosuperior view parallel to the clivus. The basilar artery (small arrow) is not visible because of overlying venous structures (large thin arrow). An aneurysm is seen within the left MCA bifurcation (large thick arrow). (b) Venous structures are eliminated by a clipping plane (yellow dotted line) parallel to the clivus. Left posterolateral view. (c) Application of the clipping plane allows an open view of the basilar artery (arrow) in the posterosuperior direction.
|
|
For analysis of the entire CT angiographic data set, the direction of view during volume rendering moves from posteroanterior to left and right lateral. To allow a closer inspection, the volume is enlarged and the movement to lateral views is repeated. After the initial evaluation of the presented approach with use of only a single video of the whole CT angiographic volume, additional movie sequences based on three overlapping subvolumes extracted from the entire image data set were recorded (Fig S1). The first subvolume showing central areas is used to depict the anterior communicating artery, both ICAs, and the basilar tip. Further on, two lateral subvolumes contain the left and right bifurcations of each MCA, respectively. Each subvolume is rotated by 180°, resulting in a transition from a posterior view to a frontal view and vice versa. A similar flight is performed after turning the subvolumes by 90° around the vertical axis, providing lateral views of both sides. A typical set of video sequences is demonstrated in Fig S1. (To reduce the video file size for downloading, the frame rate and size of the videos were markedly reduced. The original quality is demonstrated on the screen shots presented here. The original format was 640 x 480 pixels, the average file size of the main video was 53 MB, and the file size of the subvolumes was about 32 MB.)
Data Transfer and Automatic Visualization
The original DICOM (digital communications in medicine) data from the CT scanner were sent from a local PC to a remote graphics workstation. For convenient transfer of the data, a Web-based user interface providing a file selection box was developed. The interface is a Web page that can be accessed from any computer connected to the Internet. In our setup, the local computer was located in Erlangen, Germany, and the remote visualization workstation was in Stuttgart, Germany. The computers were linked via a fast Internet connection incorporated in the German research network (100 Mb/sec), and about 3 minutes were required to upload a CT angiography dataset consisting of about 150 slice images (75 MB). The visualization of the volume data was performed on a high-end graphics workstation (Onyx Infinite Reality3; SGI, Mountain View, Calif) equipped with two processors (MIPS and R12000; 400 MHz) and 2 GB of main memory. The additional 2 x 256 MB of 3D texture memory of the graphics subsystem supports a substantial number of trilinear interpolation operations, ensuring direct volume rendering at frame rates (18)] of at least 12 frames per second. Volume rendering is a method for 3D visualization that allows defining groups of voxels by thresholds dependent on their Hounsfield unit values. Thereby the groups of voxels are assigned to colors and so-called "opacity." Low-opacity voxels result in transparent objects, and high-opacity voxels result in opaque objects. In the case of CT angiography, it is useful to define at least two groups of voxels: The first group consists of voxels containing the vascular information defined by enhancement due to contrast media, typically in the range of 100–300 HU. A red color was chosen for this group of voxels. The second group of voxels contains the bone structures of the skull, with values above 300 HU. For these, a white color was chosen along with a low opacity, allowing the intracranial vessels to be seen even when bone overlaps them. For the important assignment of these color and opacity values, a previously presented automatic strategy (19,20) was applied, ensuring meaningful 3D representations of the intracranial vasculature and skull base. The actual 3D visualization is then performed with direct volume rendering based on 3D texture mapping that follows the predefined flight path (21). Thereby, the camera position is sampled in steps of small degrees, resulting in smooth transitions between consecutively rendered images. At the end of the visualization process, the entire series of images is automatically converted to a digital video. The applied video format is initially chosen within the user interface (eg, MPEG, AVI). The movie files of the entire volume and the three subvolumes are then made available on the remote Web server. An information e-mail is automatically sent when the videos are ready for download, and they are accessible exclusively to the user who submitted the original data. Addressing additional security issues, the presented Web-based approach permits access only to registered persons using password verification. An automatic logout process is launched if the Web service is not contacted for a certain period of time. Encoding with HTTPS (hypertext transfer protocol secure) ensures safe transfer of the image files.
|
Initial Clinical Experience
|
|---|
The data from 50 patients with 63 intracranial aneurysms in different locations known from DSA were analyzed with the presented remote volume-rendering approach. The main purpose was to show the feasibility of the method and to analyze whether the approach was able to demonstrate aneurysms at typical locations. In this initial analysis, aneurysms located above the skull base were demonstrated reliably, independent of the location. Even the smallest aneurysm with a diameter of about 2 mm, located at the distal ICA, was clearly seen on the video sequences (Fig S2). Difficulties were seen with aneurysms located near the skull base. One aneurysm of the intracavernous part of the ICA, with a diameter of 3 mm, was not depicted on the videos because it was hidden by the anterior clinoid process (Fig S3). With an interactive analysis of the section images by means of multiplanar reconstruction of oblique planes and a thin-section MIP technique, this aneurysm was demonstrated (Fig S3c). Large aneurysms involving the skull base were visualized only incompletely; it was not possible to see their extent and their origin (Fig S4). Analysis of the section images more easily allowed visualization of the origin of these aneurysms.
|
Discussion
|
|---|
Since methods for standardized and reproducible 3D visualization are, to our knowledge, still not available, the real value of CT angiography for the detection and evaluation of intracranial aneurysms remains unclear, despite the wealth of studies focusing on this topic. The use of any commercially available workstation requires intensive familiarization with both the techniques of visualization and the diverse manipulation tools available. The actual value of these workstations strongly depends on the individual experience of the investigator (22). For a meaningful delineation of the target structures, correct adjustment of transfer functions and efficient interactive manipulation of the volume data are necessary to make the vascular structures, including aneurysms, clearly visible. The final step consists of the documentation, including screen shots and individual video sequences. Studies performed thus far have evaluated specific systems, and the results depend on the experience of individual users. Thus, the results cannot readily be compared between studies. The neuroradiologist or neurosurgeon performing 3D visualization should not have to become involved in the actual rendering process but rather focus on the assessment and evaluation of the images or videos produced by an automatic system. To enable comparisons of scientific studies, it is essential to eliminate every external influence as much as possible and to reduce the number of free parameters to a minimum. The proposed approach allows the use of any remote high-end graphics workstation. Thus, multicenter studies become possible, and a blind rating of the 3D visualization results is readily performable if other institutions are asked for evaluation. This process also contributes to continuous improvement of the system. An important advantage of the presented approach is that the remote visualization server supports updates of hardware and software. Thereby, the system allows an objective evaluation of new 3D visualization strategies and new protocols for data acquisition in comparison with known standards.
The presented example of the intracavernous aneurysm (Fig S3) demonstrates that although the CT angiographic data contained the information, the applied visualization strategy was not sufficient to depict the lesion. The same is true for aneurysms of the ICA involving the skull base. Thus, it would be useful to find clinically applicable methods for an improved representation of aneurysms located within the cavernous sinus and those involving the skull base. For this purpose, special postprocessing techniques that suppress irrelevant bone structures are required.
Two solutions to this problem have been proposed. First, the acquisition of two data sets, one before and one after the administration of contrast medium, allows digital subtraction in the images in the context of CT-DSA (3,23). The main disadvantages of this strategy are that patients must lie absolutely still during scanning and they are exposed to increased radiation. Second, automatic segmentation of bone structures was suggested (24). This approach is technically even more challenging, since a threshold-based elimination of bone is difficult because of partial-volume effects. As mentioned before, a prerequisite for 3D visualization of intracranial arteries is the suppression of surrounding tissue by means of defined thresholds. This necessarily results in a distinct loss of information within the 3D representation. Therefore, 3D imaging cannot replace careful analysis of the underlying source images.
Intracranial aneurysms are the cause of subarachnoid hemorrhage in 75%–85% of cases. Currently, despite a high sensitivity for the detection of intracranial aneurysms, CT angiography is not able to definitively exclude very small aneurysms and other vascular lesions such as dural arteriovenous fistulas. Therefore, negative findings at CT angiography must be confirmed with DSA.
|
Conclusion
|
|---|
The automatic generation of video sequences of the cerebral vasculature based on direct volume rendering performed on a remote high-end graphics workstation is a viable approach for standardized and reproducible 3D visualization. The proposed technique could be used to initiate multicenter studies that may allow a more objective evaluation of intracranial aneurysms with CT angiography. Despite the high quality of modern volume-rendering systems, 3D representations alone cannot replace the careful inspection of source images containing the complete information in a volumetric CT data set.
|
Footnotes
|
|---|
Abbreviations: DSA = digital subtraction angiography, ICA = internal carotid artery, MCA = middle cerebral arteries, MIP = maximum-intensity projection, 3D = three-dimensional.
|
References
|
|---|
- van Gijn J, Rinkel GJ. Subarachnoid haemorrhage: diagnosis, causes and management. Brain 2001; 124:249-278.[Abstract/Free Full Text]
- White PM, Teasdale EM, Wardlaw JM, Easton V. Intracranial aneurysms: CT angiography and MR angiography for detection—prospective blinded comparison in a large patient cohort. Radiology 2001; 219:739-749.[Abstract/Free Full Text]
- Imakita S, Onishi Y, Hashimoto T, et al. Subtraction CT angiography with controlled-orbit helical scanning for detection of intracranial aneurysms. AJNR Am J Neuroradiol 1998; 19:291-295.[Abstract]
- Ogawa T, Okudera T, Noguchi K, et al. Cerebral aneurysms: evaluation with three-dimensional CT angiography. AJNR Am J Neuroradiol 1996; 17:447-54.[Abstract]
- Strayle-Batra M, Skalej M, Wakhloo AK, Ernemann U, Klier R, Voigt K. Three-dimensional spiral CT angiography in the detection of cerebral aneurysm. Acta Radiol 1998; 39:233-238.[Medline]
- Kato Y, Katada K, Hayakawa M, et al. Can 3D-CT angiography surpass DSA in diagnosis of cerebral aneurysm? Acta Neurochir 2001; 143:245-250.
- Young N, Dorsch NW, Kingston RJ, Markson G, McMahon J. Intracranial aneurysms: evaluation in 200 patients with spiral CT angiography. Eur Radiol 2001; 11:123-130.[CrossRef][Medline]
- Korogi Y, Takahashi M, Katada K, et al. Intracranial aneurysms: detection with three-dimensional CT angiography with volume rendering—comparison with conventional angiographic and surgical findings. Radiology 1999; 211:497-506.[Abstract/Free Full Text]
- Calhoun PS, Kuszyk BS, Heath DG, Carley JC, Fishman EK. Three-dimensional volume rendering of spiral CT data: theory and method. Radiographics 1999; 19:745-764.[Abstract/Free Full Text]
- Blank M, Kalender WA. Medical volume exploration: gaining insights virtually. Eur J Radiol 2000; 33:161-169.[CrossRef][Medline]
- Tomandl BF, Hastreiter P, Rezk-Salama C, et al. Local and remote visualization techniques for interactive direct volume rendering in neuroradiology. Radiographics 2001; 21:1561-1572.[Abstract/Free Full Text]
- Napel S, Marks MP, Rubin GD, et al. CT angiography with spiral CT and maximum intensity projection. Radiology 1992; 185:607-610.[Abstract/Free Full Text]
- Liang EY, Chan M, Hsiang JH, et al. Detection and assessment of intracranial aneurysms: value of CT angiography with shaded-surface display. AJR Am J Roentgenol 1995; 165:1497-1502.[Abstract/Free Full Text]
- Choyke PL, Yim P, Marcos H, et al. Hepatic MR angiography: a multiobserver comparison of visualization methods. AJR Am J Roentgenol 2001; 176:465-470.[Abstract/Free Full Text]
- Tsuchiya K, Katase S, Yoshino A, Hachiya J, Yodo K. Preliminary evaluation of volume-rendered three-dimensional display of time-of-flight MR angiography in the diagnosis of intracranial aneurysms. Neuroradiology 2001; 43:633-636.[CrossRef][Medline]
- Kirchner J, Kickuth R, Laufer U, Noack M, Liermann D. Optimized enhancement in helical CT: experiences with a real-time bolus tracking system in 628 patients. Clin Radiol 2000; 55:368-373.[CrossRef][Medline]
- Day AL, Morcos JJ, Revilla F. Management of aneurysms of the anterior circulation. In: Youmans JR, eds. Neurological surgery: a comprehensive reference guide to the diagnosis and management of neurosurgical problems. Philadelphia, Pa: Saunders, 1996; 2:1273-1274.
- Cabral B, Cam N, Foran J. Accelerated volume rendering and tomographic reconstruction using texture mapping hardware. In: Kaufman A, Krueger W, eds. Symposium on volume visualization. New York, NY: ACM SIGGRAPH, 1994; 91-98.
- Hastreiter P, Tomandl B, Eberhardt K, Ertl T. Fast analysis of intracranial aneurysms based on interactive direct volume rendering and CT-angiography. Proceedings of MICCAI Heidelberg, Germany: Springer, 1998; 660-669.
- Rezk-Salama C, Hastreiter P, Scherer J, Greiner G. Automatic adjustment of transfer functions for 3D visualization. In: Girod B, Greiner G, Niemann H, Seidel HP, eds. Vision, modeling and visualization. Sankt Augustin, Saarbrücken, Germany: Infix, 2000; 157-364.
- Iserhardt-Bauer S, Rezk-Salama C, Ertl T, Hastreiter P, Tomandl B. Automated 3D video documentation for the analysis of medical data. In: Handels H, Horsch A, Lehmann T, Meinzer HP, eds. Bildverarbeitung für die medizin 2001: algorithmen systeme anwendungen. Heidelberg, Germany: Springer, 2001; 409-413.
- Udupa JK. Three-dimensional visualization and analysis methodologies: a current perspective. Radiographics 1999; 19:783-806.[Abstract/Free Full Text]
- Gorzer H, Heimberger K, Schindler E. Spiral CT angiography with digital subtraction of extra- and intracranial vessels. J Comput Assist Tomogr 1994; 18:839-841.
- Moore EA, Grieve JP, Jager HR. Robust processing of intracranial CT angiograms for 3D volume rendering. Eur Radiol 2001; 11:137-141.[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
M. M. Lell, K. Anders, M. Uder, E. Klotz, H. Ditt, F. Vega-Higuera, T. Boskamp, W. A. Bautz, and B. F. Tomandl
New Techniques in CT Angiography
RadioGraphics,
October 1, 2006;
26(suppl_1):
S45 - S62.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. F. Tomandl, N. C. Kostner, M. Schempershofe, W. J. Huk, C. Strauss, L. Anker, and P. Hastreiter
CT Angiography of Intracranial Aneurysms: A Focus on Postprocessing
RadioGraphics,
May 1, 2004;
24(3):
637 - 655.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
