Image Processing Reference
In-Depth Information
ablation therapy. These types of reformations andmappings entail that more effort has
to be put into carefully designing simplified, usually 2D, representations of complex
3D data, as opposed to, for example, the relatively straight-forward projection of
volume data. The resultant visualization, if done right, requires less or no interaction
and by definition avoids a number of problems inherent in 3D representations [ 74 ].
23.3.7 Illustrative Visualization in Medicine
For a long time, it was not possible to apply illustrative visualization techniques in
practice due to performance constraints. With advances in graphics hardware and
algorithms, such as GPU raycasting [ 45 ], it is now feasible from a computational
standpoint. Now that computational problems have been largely solved, illustrative
visualization approaches have to be finetuned and evaluated for diagnostic and treat-
ment planning purposes. Recent examples of such work include the simulation of
crepuscular rays for tumor accessibility planning [ 37 ] and multi-modal illustrative
volume rendering for neurosurgical tumor treatment [ 58 ].
Illustrative medical visualization becomes increasingly important when visualiza-
tions become more complex and multi-modal, integrating functional (measured and
simulated) information, anatomical information and, for example, surgical instru-
ments. Illustration techniques enable visual representations to be simplified intelli-
gently by the visualization designer, whilst still communicating as much information
as possible. An example of this is the work of Zachow et al. [ 77 ] on the visualization
of nasal air flow simulation where boundary enhancement was used as an illustrative
technique to convey the simulated flow and the anatomy simultaneously.
23.3.8 Hyper-Realism
Analogous to the case of illustrative visualization, the rapid development in graphics
hardware and algorithms has now enabled the interactive rendering of medical imag-
ing datasets with physically-based lighting [ 44 ]. Figure 23.4 shows an example of
such a visualization. These techniques make possible the simulation of an arbitrary
number of arbitrarily shaped and textured lights, real shadows, a realistic camera
model with lens and aperture, and so forth, all at interactive rates.
These techniques enable not only photo-realism, but also a technical form of
hyper-realism in art, where it is possible to enhance visualizations with additional
realistic detail in order to better convey information. Whilst there are strong indica-
tions that, for example, global illumination and shadows can have a positive effect
on task performance in normal volume rendering [ 48 ], the possibilities and value of
hyper-realistic effects in medical visualization need to be explored.
 
Search WWH ::




Custom Search