Image Display Techniques

A major advantage of magnetic resonance angiography (MRA) is the ability to acquire a three-dimensional (3D) data set of the vessel tree. Any technique, be it time-of-flight (TOF) or phase contrast, is usually applied in 3D format to obtain a series of sl

  • PDF / 1,706,776 Bytes
  • 9 Pages / 547.08 x 765.36 pts Page_size
  • 36 Downloads / 253 Views

DOWNLOAD

REPORT


Image Display Techniques JOHAN VAN CLEYNENBREUGEL and GERHARD LAUB

to display. Therefore, this matter is fairly technical. To the radiologist reading the images, differences in dis9.1 Surface-Based and Volume-Based Methods 129 play results obtained by either technique may even be 9.2 Exoscopic Techniques: MIP, Volume Opacity unnoticeable in some cases. Rendering, and Surface-Based Rendering 131 Surface-based methods, such as shaded surface 9.3 Reformatted Techniques 133 display (SSD) or isosurface generation (e.g., based 9.4 Virtual Endoscopy 136 References 136 on "marching cubes:' LORENSEN and KLINE 1987) hinge on thresholding the image volume in order to obtain a binary classification in the vascular object and background. Therefore such techniques are more 9.1 appropriate for image volumes with a relatively high Surface-Based and signal-to-noise ratio. For example, in X-ray based Volume-Based Methods CT angiography, SSD is reported to be the preferred method (KALENDER and PROKOP 2000). In MRA A major advantage of magnetic resonance angiogra- imaging, contrast-enhanced MRA is a typical candiphy (MRA) is the ability to acquire a three-dimen- date for applying surface-based methods. An examsional (3D) data set of the vessel tree. Any technique, ple is shown in Fig. 9.1. Additionally, this figure also be it time-of-flight (TOF) or phase contrast, is usually demonstrates the functionalities to be expected from applied in 3D format to obtain a series of slices or current image reading systems. Indeed, by reprea volume data set which is interpreted as a series senting the 3D MRA volume as a 3D scene (VAN of thin contiguous slices or partitions. One can look CLEYNENBREUGEL et al. 1996), the vascular structhrough such individual slices or calculate multipla- tures are useful to interactively access (re)slices at nar reconstructions from interesting vessel regions. appropriate locations. For example, in a presumably While this approach is straightforward and simple to stenotic region, a reslice perpendicular to the vesuse, it is very difficult for the observer to obtain a sel's centerline can be included in the 3D scene whencorrect spatial perception of the vascular structure. ever needed. In analogy to the term MPR (multiplaDuring the period 1990-2000, methods developed nar reformatting), this reslicing facility is sometimes from image processing and computer graphics called CPR (curved planar reformatting). matured into so-called "visualization techniques". As Contrary to surface-based methods, volume-based a result, interactive real-time 3D reading of an image methods do not require a binary decision with respect volume has become common practice. In particular, to object and background. Volume-based methods spatial perception of vascular structures can be read- produce a 3D scene visualization by shooting rays ily obtained. Visualization techniques are usually from a given viewpoint (the center of a virtual classified as either surface-based or volume-based. camera) through the 3D image volume in order to This distin