D-ATLAS (Diabetes Atlas) - How it works

What is the D-ATLAS?
The D-ATLAS (Diabetes Atlas) maps the prevalence and costs associated with type 2 diabetes.

Press Release:
The D-ATLAS: A Tool Supporting the Fight Against Diabetes

To View the Atlas, register here:
http://www.z-atlas.com/content.php?sec=diabetes

Overview
The Diabetes Atlas was built using a combination of ESRI's ArcGIS Server JavaScript API and Microsoft's .NET Framework. ArcGIS Online data is used to show the street data in the Atlas. The Disease information is displayed using ArcGIS Server MultiLayer Cached Services. Finally, the demographic boundaries are displayed using ArcGIS Server 9.3.1 Optimized Map Services and GraphicsLayers.

Managing the Cache
Although fused caches perform the best, we needed to be able to turn layers on and off. Consequently, we decided to use multilayer services that are cached. Now that 9.3.1 is out, we are looking more towards migrating these services to optimized map services. This would eliminate the struggles to manage large cached datasets. However, we've found that performance at the national level is still slower with optimized map services due to the amount of information (or polygons) that need to be rendered on the server. Using the multilayer services that are cached draws the national maps quickly by grabbing the images directly from the cached folders.

Printing:
Printing on the web is never fun. Consequently, it has by far been the biggest challenge. We have gone through many versions of printing during the lifetime of the Atlases. Now, we are using the print sample found in DS2009: Developing Advanced Applications with the ArcGIS JavaScript API. We modified the code to work with a hidden "whiteout layer" from the main page. A definition query on the "whiteout layer" is updated as users change the regions. Then when map is printed, the whiteout layer's visibility is set to true. The whiteout layer covers the non-selected areas showing only the specific region the user wants to print.

Managing the Atlases:
All the information displayed in the Atlases are managed from a SQL Sever database. The database stores information at the Disease, Disease Year, and Demographic levels. Consequently, we can easily update information in the Atlases (Legend Images, Legend Titles, Active and Inactive Services, methodology, etc) without having to touch the code. For example, if we need to update the cache for a service, we can simply set the service to inactive in the database, and it will not show up in the Atlas while the cache is being built.

Screen Grabs:
D-Atlas:


Print Page: