Visualization in Medicine

Visualization is one of the rapidly developing areas of scientific computing. The cost of high performance computing has become increasingly more affordable in recent years. This has promoted the use of scientific visualization in many disciplines where the complex data sets are rich in quality and overwhelming in quantity. In medicine, more than cost, advancement of the physics of imaging has become a major influencing factor that has spurred the growth of image computing and visualization. The increasing cost of health care has created an awareness and demand for investigating safe and cost-effective approaches to practice diagnosis and deliver treatment. This trend has created a need for powerful ways of delivering information to physicians at every step in patient care delivery that might help the physicians to understand the problem better and faster and stay closer to the truth without leaving any room for potential instrument or human error. Traditionally, the practice of medicine requires information to be handled in a variety of ways, such as by touch, sound, appearance, and smell, in a manner comparable to a craft. The craftsmanship involved in medicine becomes more obvious when it comes to therapy or treatment procedures where the physician's hand-eye coordination is the final step that decides the outcome of what may be considered an extensive and expensive investigation. From an engineering standpoint, these aspects pose new challenges, since most of the information is qualitatively rich while quantitatively difficult to characterize. Thus, visualizing the information that describes the nature of the underlying "functional" source becomes vital in medicine. Functional visualization is a particularly effective approach because multiple functional characteristics can be mapped to different visual cues that facilitate the interpretation of multidimensional information and correlation of qualitative and quantitative information simultaneously. Some examples are three-dimensional perspective realism for representing spatial relationships effectively, animated displays for representing temporal information, and other forms of visual cues such as hues and textures for representing various quantitative or functional information. Although information visualization is one aspect of the problem, the paradigm of interaction between the physician and the information is an equally significant one for craftsmanship needs.

It is the rapid development and culmination of technology

Copyright © 2000 by Academic Press.

All rights of reproduction in any form reserved.

that meets such needs in graphics computing, virtual interaction, and tactile feedback systems that has created a synergy of research between physicians and engineers that has led to the development of new frontiers in visualization in medicine. The impacts these advancements are creating in biomedical visualization are many, and they can be broadly classified to different areas based on the application such as clinical diagnostic visualization, image guided therapy, procedure training simulation, pretreatment planning, and clinical hypothesis validation.

The chief purpose of this chapter is to introduce visualization concepts to the reader. Given that such concepts change in unison with technology, it is also important to present here the underlying problems, through case examples in medical visualization. This provides a better appreciation of these concepts from an application point of view and will allow the reader to follow the evolution of the concepts in the future.

1.1 The Genealogy of Visualization Visualization Phenotypes

Based on their visual characteristics resulting from their interaction with the user environment, visualization in medicine can be grouped into three major classes: illustrative visualization, investigative visualization, and imitative visualization (Fig. 1). The extent to which they embrace the technology of visualization can be considered as marginal, high, and maximum, respectively.

Illustrative Visualization. Illustrative visualization developed over the past two decades from attempts to separate visualization into two distinctive processes: extraction of information and its presentation. Fast processing is desirable but not critical, whereas quality and accuracy are essential. The concepts of illustrative visualization form the basis of the other two classes of visualization. Illustrative systems do not rely on interactive data manipulation but present information that may have been carefully extracted from the data by other means.

The initial expectations of biomedical visualization based on this approach soon reached a plateau. As stated in an NSF grant announcement: "In spite of enormous advances of computer hardware and data processing techniques in recent decades, the amazing and still little understood ability of human beings to 'see the big picture' and 'know where to look' when presented with the visual data is still well beyond the computer's analytic skills.'' Although trying to characterize humans' ability to perform such processing is a challenging scientific investigation in itself, it also became imperative in biomedical research to provide rapid visualization systems that take advantage of this human ability. This presented the motivation for investigative visualization systems.

Investigative Visualization. Investigative visualization tries to focus on the explorative aspect of visualization. These techniques generally do not aim to provide a strict analytical solution; instead, they aim to provide a visual solution that the human eye might be able to interpret better. In general these approaches do not require the detailed knowledge about the data that analytical approaches require. The explorative aspect of this class of visualization has been gaining appeal because it is essential for the clinical application of new imaging methods. Some of the concepts in this class are interactive or real-time volumetric visualization, dynamic visualization, multimodality registration, functional (multidimensional) visualization, and navigational visualization. The emphasis is usually on speed, because interactive ability is essential to make these visualization tools useful in practice.

FIGURE 1 Visualization pathways.

Imitative Visualization. This class of visualization attempts to imitate reality through visualization. The imitation can be a visual perception as in virtual reality type systems or functional imitation as in simulation and modeling. The challenge is to provide a realistic simulation in a virtual environment. In medicine, such a challenge arises naturally in the area of pretreatment planning and training for interventional and intraoperative surgical procedures. Both the speed and quality of visualization are critical here, and so is the paradigm of interaction. Some of the technology needed for such visualization applications may not be available today, but there is rapid progress in this area.

1.2 Visualization Genotypes

In medicine, because of the inherent complexity in visualizing the information in the data, different concepts of visualization evolved as the technologies that could enable them became available. For the purpose of this discussion, the evolution of different concepts of visualization in medicine can be grouped into several generations.

First generation systems are essentially ID waveform displays such as those that appear in patient monitoring systems.

Second generation systems perform 2D image processing and display. Contours and stack of contour lines that can represent the three-dimensional form of the data also were developed during this period.

Third generation systems generally involve 3D image processing and visualization. Isosurfaces, contour surfaces, shell-rendering and volume-rendering techniques were developed in this generation.

Fourth generation systems process multidimensional data such as dynamic volume data sets, sometimes called 4D data. In general the fourth dimension can be any other dimension associated with volume data.

Fifth generation systems are virtual reality type visualization systems, which combine multidimensional data with three-dimensional (i.e., six degree of freedom) interaction.

Next generation systems represent concepts under development, such as sensory feedback techniques where the user interacting with the structures could feel the physical properties of the material and obtain valuable "visual-sensory" information in simulation type visualization systems.

The main focus of this chapter is on investigative and imitative visualization. Clinical research examples from medicine and biology are provided to illustrate the concepts of visualization and their significance.

Was this article helpful?

0 0

Post a comment