Computer Aided Diagnosis: State-of-the-Art and Application to Musculoskeletal Diseases - 3D Multiscale Physiological Human - page 278

the view through a real endoscope. The virtual camera is moved along a path in the
center of the relevant structure. The calculation of this path requires segmentation
and center-line extraction of the relevant structure. The relevant structure may be
visualized by means of surface or volume rendering.
Virtual endoscopy can be applied with a variety of different goals. These goals
depend largely on the clinical questions of the original real endoscopic procedure
being mimicked by virtual endoscopy. Some of the goals are aimed at reducing
costs in the clinical routine, a topic of almost permanent relevance in modern health
care; others are aimed at improving the intervention quality by reducing the risks of
complications. Another possible goal is the reduction of the training costs for medical
specialists. Several technical issues need to be addressed for virtual endoscopy. The
first issue concerns the rendering of the virtual endoscopy view, and the second
issue relates to the navigation of the virtual camera through the visual representation
of the respective body cavity. Furthermore, the CAD system user-interface and the
functionality must be carefully integrated to provide a seamless workflow.
Hong et al. [ 52 ] pointed out that in the context of such a system, three different
navigation options are available: automatic navigation, manual or free navigation, and
guided navigation. Automatic or planned navigation relies on a predefined camera
path through the representation of the respective body cavity. This camera path must
specify positions and viewdirections (orientations) of the virtual camera. Afterwards,
a fly-through is computed based on that path.While this option offers a good overview
of the target area, it requires the refinement of the camera path, and the subsequent
regeneration of the fly-through, to capture details that were previously not sufficiently
visible. The second possibility is manual or free navigation , where the camera is
transformed freely. This navigation paradigm is particularly popular for viewing
computer graphics models from the outside. For virtual endoscopy applications,
however, free navigation poses severe difficulties due to the high complexity of
many body cavities. Furthermore, the lack of collision avoidance mechanisms and
the difficulties of adding those to a free navigation system worsen this issue. The
best option for virtual endoscopy combines navigation flexibility with guidance and
is hence called guided navigation [ 53 ]. It combines a set of constraints that guide the
user to a predefined target area.
The most important application is virtual colonoscopy, in which the virtual endo-
scope is moved through the colon to detect polyps (which are often a presage of
colon cancer). Computer support for virtual colonoscopy (path planning and navi-
gation support) is provided by a number of manufacturers and has matured over the
last years.
12.3.2 Follow-Up Features
For some chronic illnesses, frequent monitoring provided by health care staff is an
integral part of the treatment plan (e.g., blood pressures, MRI scan, …) [ 54 , 55 ]. All