Biomedical Engineering Reference
In-Depth Information
Fig. 5.4
Segment visualization in YaDiV [
5
], showing intracranial air after brain surgery
improvements are often not evaluated. During the design process, visualization
researchers should use the scientific method (phenomena observation, hypothesis
formulation, results prediction and evaluation) to prevent creating a toolset that is
detached from the biomedical meaning [
27
].
Evaluation of Visualization Tool Design
Considering visualization as a technology, evaluation and validation are vital in order
to provide effectiveness and efficiency. Wijk explains in [
28
] an economic model of
visualization. The cost associated with using a visualization system depends on the
user and the initial development. The user cost is related to the learning time required
to use the new tool, to convert the user data to the system's format and to interpret
the presented results. The high price of the initial development costs (i.e. to have a
novel idea and to develop it) is the major factor. The researcher's process that tries
to provide a visualization tool to a scientist in order to help him in the understanding
of complex data is described in [
28
]. The state of the art review and the development
of a new idea is not enough to achieve success. Thus, it is crucial to prepare a
study on advantages and limitations before starting to work on a new visualization
method. Wijk also states that “visualization is not good by definition” [
28
]. Designers
should avoid visualizing information that could be extracted by automated analysis
of data, and do not require direct human interpretation. Reusing existing visualization
software can also be a viable solution.
5.4 Multiscale Visualization and Interaction
Multiscale visualization is necessary due to the division of science into domains, each
investigating the nature on a specific spatiotemporal scale and working with its own
type of data. This specialization can even prevent scientific progress if phenomena
