"HealthVis" is actually two separate visualization prototypes designed to facilitate the exploration of geo-referenced health statistics.

The U. S. Centers for Disease Control and Prevention, National Center for Health Statistics (NCHS) recently commissioned a study designed to pinpoint specific conceptual and implementational issues in interface design for Geographic Visualization (geovisualization) and exploratory data analysis (EDA) of relationships between risk factors and mortality rates. This page reports on one component of that project, the implementation of some exploratory data analysis operations in ArcView GIS (ESRI, Redlands CA).

The software prototype provides a number of interactive methods for exploring relationships between risk factors and mortality rates and how they are distributed in space. The use of geographically referenced mortality data to detect disease "hot spots" can be traced, at least, to Dr. John Snow's 1854 map of cholera deaths in London, which allowed him to hypothesize that a particular water pump was the source of the epidemic. While the use of traditional static maps for cluster identification continues to be important, with a major new atlas of mortality in the U.S. just published, dynamic exploratory data analysis and visualization techniques have the potential to further enhance detection of "hot spots".

Scatterplot brushing, interactive data classification, focusing, and representational methods for multivariate display, were incorporated into our system. These techniques can help the analyst identify disease "hot spots" and facilitate data exploration that may lead to hypothesis about causal links between mortality and potential risk factors.

Both ArcView ® and Macromedia Director ® provide fairly robust object oriented scripting languages. The ArcView prototype was chosen to explore multivariate data-driven analysis of geographic data. However, ArcView provides poor support for animation or the development of custom user controls (such as slider bars). Director, on the other hand, emphasizes animation, and is highly flexible in facilitating design of user controls. Its weaknesses lie in the area of data-driven analysis of geographic representations. As a result, ArcView hosts the data-driven analysis and provides images of geographically referenced data that Director can animate and manipulate with a more flexible user interface.

If you would like to try out HealthVIS you can download it for free. You don't need any special software to use HealthVIS. Please select one of the following options.

Windows 95/ NT (3.4 MB)

PowerPC Mac (5.9 MB)

Animated GIF. Although you can't interact with this tutorial demonstration, this GIF is much smaller (270k).

[PLEASE NOTE: HealthVIS is a prototype and is provided "as is"]

If you would like to learn more about the ArcView component of HealthVis, go here.

Further discussion of the Director side of this research can be found here.

For a characterization of HealthVis use in data exploration, see our IEEE Information Visualization paper (PDF Format).

 

The goals of this project were to construct a prototype interface that could be employed in empirical testing of representation and interaction techniques applied to analyzing multiple geo-referenced variables at several time intervals. Because of the limitations of the software packages available at the project outset, we chose to develop two separate modular prototypes.

Figure 1: Linked Brushing.
As an analyst interacts with the data in the 1992 scatterplot, the same data entities are highlighted in the 1989 scatterplot and on the map. This allows the relationship between statistical space and geographic space to be explored.

---

Figure 2a and 2b: Interactive Focusing
The use of a moveable slider on the scatterplot divides the data into two halves, which is simultaneously displayed on a linked map display. As shown above in the two images, this technique can be used to isolate outliers and anomalies in the data.

---

Figure 3a and 3b: Dynamic Classification
By adding more axes to the scatterplot the user now has the ability to create dynamic classifications, exploring the relationship between health variables both in statistical space and geographic space.