13
th
AGILE International Conference on Geographic Information Science 2010 Page 1 of 10
Guimarães, Portugal
Three-dimensional Visualization, Interactive Analysis and
Contextual Mapping of Space-time Cancer Data
Pierre Goovaerts (BioMedware, Inc)
INTRODUCTION
The increase in computational power and storage capacity of computers, combined with the
growing availability of geocoded data, has increased dramatically the amount of information
processed in health studies, making it difficult to understand, to explore, and to discover interesting
patterns within the data. The major difficulty in the analysis of health outcomes is that the patterns
observed reflect the influence of a complex constellation of demographic, social, economic, cultural
and environmental factors that likely change through time and space, and interact with the different
types and scales of places where people live. Thus, there is a large heterogeneity in the temporal and
spatial scales of investigation, leading to the utilization of a wide range of statistical methods and
visualization techniques in most studies of health outcomes.
Despite the significant work accomplished in health data visualization and analysis this last
decade, spatial and temporal data are still displayed in separate views and so current software do not
capitalize on the human visual processing engine to extract knowledge from the spatial
interconnectedness of information over time and geography. In addition, common approaches to
disease mapping too often focus on the display of disease rates for political or administrative units
and lack information on the local context of cancer burden, which is critical to facilitate the
interpretation of these maps by local communities and engage their participation to prevention and
control activities. This paper reviews common approaches for the space-time visualization of health
data and explores solutions for the 3D interactive visualization of health outcomes in a combined time
and geography space and contextualization of disease maps through the incorporation of familiar
markers, such as highways, rivers or topographic details.
VISUALIZATION OF SPACE-TIME DATA: COMMON SOLUTIONS
Most visualization techniques for analyzing health outcomes only display information along a
single spatial or temporal dimension. For example, time-focused scheduling charts such as Lifelines
(Plaisant et al., 1998) or Microsoft (MS) Project display attributes of events over the single dimension
of time. Geographic Information System (GIS) products, such as GeoDA (Anselin et al., 2006) or
ESRI ArcView, show events in the single dimension of space on a map. In each case, only a thin slice
of a multidimensional picture is represented.
The following visualization techniques (Kapler and Wright, 2005), which are all available in
TerraSeer’s Space-Information Intelligence System (STIS, Avruskin et al., 2004), are commonly used
to overcome the limitations of one-dimensional displays:
1. Temporal animation: slideshow controlled by a time slider. Such dynamic representations
are well suited to convey the general trend in health outcomes over time. However, this
technique relies on limited human short term memory and quickly reaches its limit when
longer time series have to be visualized: users simply cannot retain all changes in the visual
representation and the animation takes too long for the user to remember its course. Even to
achieve simple comparison of different time steps users would be compelled to browse
through the animation over and over again (Tominski et al., 2005).