Researchers at three IBM labs have donated to the open-source community a package of advanced software that can help predict the transmission of diseases across countries and around the globe, so that public health authorities can make better decisions on how to handle such outbreaks.
The tool, known as STEM (Spatiotemporal Epidemiological Modeler), was made available for free download June 8. It will aid researchers and public health officials in understanding and planning more efficient responses to health crises and ultimately provide new tools for protecting population health.
The STEM software is now available for use through the Eclipse Open Healthcare Framework Project, hosted at the Eclipse Foundation, the nonprofit foundation that guides the 6-year-old Eclipse open-source community.
“This [software] could have been helpful to a great extent in cases like the recent one about the young man [Andrew Speaker] who flew on the airlines with tuberculosis, or the Severe Acute Respiratory Syndrome [SARS] scare in Southeast Asia a few years ago,” Dr. Alan Louie, a health care software analyst with Health Industry Insights in Framingham, Mass., told eWEEK.
“We would have been able to chart exactly where the diseases were moving, much faster and way ahead of time, in order to put quarantines in place and stop the movement early on.”
SARS was first reported in Asia in February 2003. Over the next few months, the illness spread to more than two dozen countries in North America, South America, Europe and Asia before the global outbreak was contained. According to the World Health Organization, 8,098 people worldwide became sick with SARS during the outbreak. Of these, 774 died.
STEM might be able to save lives in cases like SARS and others. The software represents nearly three years of research spanning the globe, with scientists from IBMs Almaden, Haifa and Watson labs contributing to its creation, an IBM spokesperson said.
The software enables the fast creation of epidemiological models—in graph form—for how an infectious disease is likely to spread over time and geography.
STEM, which runs on any operating system, creates a graphical representation of the spread of a disease based on a variety of parameters, such as population, wind patterns, geographic and macro-economic data, road maps, airport locations, travel patterns—even bird migratory routes around the world—the spokesperson said.
“STEM will allow public health officials to model the spread of a disease, much like modeling a storm or hurricane—it allows us to produce a public health weather map for the spread of a particular disease,” said Joseph Jasinski, IBM distinguished engineer and program director of Healthcare and Life Sciences.
“Until now, it has been difficult to simulate health crisis scenarios on a global scale. STEM gives us the power to do that.”
Graphs of all the pertinent information are overlaid and used to represent the spatial aspects of the model, Daniel Ford, the lead IBM programmer on the project, told eWEEK.
“While we were putting this all together, we realized that this [STEM] would enable a collaborative community in the disease modeling area,” Ford said. “If you ever attend a disease modeling conference, you quickly realize that there isnt a lot of collaboration between groups—they dont really share anything. This is really a reflection of the funding sources.”
Ford said he and his team saw an opportunity where “we could come into that kind of disorganization and provide some leadership by providing a tool that would allow people to come together and collaborate by building different parts of the models and putting them together on a common platform.”
That meant it had to be open source, Ford said. Naturally, they turned to the open-source Eclipse project—which originally was created within IBM to build open-source development tools in 2001—to supply the platform.
Eclipse is now an independent, nonprofit organization, but it still has close ties to IBM; many of its active members are IBM employees.
A basic epidemiological model framework will be provided to software developers, who can customize and configure the models based on their specific requirements. These models, which involve multiple populations and interactions between diseases, can help public health experts develop more effective preparedness plans.
“Potentially, you might have public policy decisions made based on information provided by STEM,” analyst Louie told eWEEK. “You might run multiple different kinds of scenarios in parallel with STEM, and each scenario might be some slightly different view [of the problem].
“An example would be the hoof-and-mouth disease outbreak among cows in the UK in 2000,” Louie said. “Like a million or so cows were killed, costing about $12 billion. … When a report of an infected cow came in, they would go and kill every cow within a certain radius of that farm. The problem is: Whats the right radius? Should it be 3 kilometers, 5 kilometers … should it be something else? There are a lot of consequences in these things.”
If you make the radius too big, then you kill too many cows and affect tourism, Louie said. If you make it too small, you might not be able to contain the disease, he said.
“Its really a situational awareness-and-response system that weve released to the community,” Louie said.
The person most likely to use this software would be a programmer working with a disease control or other public health office at the state or federal level, Louie said.
IBMs donation is also intended to help facilitate collaboration between governments, scientific researchers and other players in the public health community, Ford said. Users will have the ability to share the customized epidemiological models that they create in addition to the plug-ins they build using Eclipse, he added.
STEM is one of the key technologies being utilized in the Global Pandemic Initiative, a collaborative effort formed by IBM and more than 20 major worldwide public health institutions to help prevent the spread of infectious diseases.
The tool is also compatible with the standards-based interoperable health care infrastructure developed by IBM and can query clinics, hospitals, lab systems and other information sources for anonymized data by disease, which enables a complete picture of the health of a population with real-time data.
To download STEM free of charge, go here (134MB file).