Extended vision for oncology: A perceptual science perspective on data
visualization and medical imaging
Todd S. Horowitz, Ronald A Rensink
In B.W Hesse, D.K. Ahern, E. Beckjord (Eds). Oncology Informatics: Using Health
Information Technology to Improve Processes and Outcomes in Cancer Care
(pp. 287-303). Elsevier. 2016. [web]
Recent years have seen an explosion in
the use of medical informatics, the application of computing power to medicine.
This development has created great opportunities for improving the science and
practice of oncology and cancer care, as illustrated in many other chapters in
this volume. In principle, getting more information faster should let us make
better, more informed decisions. However, any such system must face a critical
bottleneck: the human being who must make sense of this information. An
oncology informatics system typically presents information visually—for example,
in the form of a dashboard, a graph, or a digital mammogram. But the nature of
this bottleneck—the limitations of the human observer—is often not well
understood, and so often is not adequately considered when designing a medical
informatics system. This can cause problems. In this chapter, we start with the
view that any system for processing information is incomplete unless the human
observer—the person supposed to explore, analyze, and ultimately use the
information—is taken into account. We do not pretend that this is a new
insight. There is a considerable body of work on improving the effectiveness of
systems to convey quantitative information visually. To date, however, this has
been a fairly empirical tradition. In contrast, we suggest here that oncology
informatics could be substantially improved by drawing on basic research in
visual perception to offer a better understanding of how best to present data
and images. To this end, we begin with an introduction to the functional
properties of the human visual system. We follow this with a general discussion
of how knowledge of vision science can be applied to the field of
visualization, which is concerned with developing effective ways of conveying
information visually. Then, to illustrate how this approach can be applied to
the design of improved informatics systems, we discuss in detail two specific
examples: the effective display of quantitative data, and the effective display
of medical images.