Social Media and Scientific Discovery - (Oak Ridge National Laboratory)

Mobility and risk

Social networks may play a role in understanding relationship between population migration, cancer

	Georgia Tourassi and Songhua Xu are using information found on social networks to study the cancer risks associated with migration. (hi-res image)

Public forums on social media may enable Oak Ridge National Laboratory researchers to study environmental cancer risks virtually, bringing the scientific community closer to understanding the impacts of modern population migration patterns on cancer risk.

Over the next four years, Georgia Tourassi and Songhua Xu in ORNL's Biomedical Science and Engineering Center will use such information to try to answer a question that has puzzled experts for decades: What environmental factors change the risk of various cancers when people move from one geographic region to another?

The ORNL researchers hope the endless supply of case studies available online and in newspapers will enable them to develop a framework that will help epidemiologists narrow future studies.

"There is a general movement to see how we can use social networks to not only help epidemiologists discover and monitor the spread of infectious diseases, but also answer a large range of epidemiological questions specifically related to cancer," Tourassi said. "Plenty of studies strongly suggest that many answers we get through expensive clinical trials are similar to those we get quickly by mining online media. If we demonstrate that social media can be used to answer epidemiological questions, we will have set the stage for this line of research in the cancer scientific community."

Epidemiology studies and clinical trials on cancer are typically time consuming, expensive and complex, and major breakthroughs have been stalled for the past few years. This lull has prompted the National Institutes of Health to fund 38 research teams who proposed high-risk creative projects that aim to answer 24 of the most provocative questions in cancer research.

The ORNL research duo's four-year grant is worth more than $1.6 million and will allow them to design cyber informatics tools that can search for, read through, and translate large amounts of online information.

Xu, an expert on web intelligence and online contents mining, will tailor the programs to filter out reliable stories on breast and lung cancer — creating a tool that acts like an army of computer analysts constantly collecting and processing information. Linking these stories with publicly available environmental data and mining them using artificial intelligence will allow the ORNL team to search for associations between changes of migration-influenced environmental factors and cancer risk. Each phase of the project will be reviewed by the Oak Ridge Institutional Review Board to assure proper protections are in place for information that is mined and used in their research.

Collaboration with clinical specialists, cancer environmentalists and biostatisticians from Brown University and the University of North Texas will help ORNL researchers interpret the associations they discover.

"ORNL will have a very critical role to play when it comes to this type of research because of its unique computing resources and scientific capabilities," Tourassi said. "This research fits perfectly with DOE's mission of scientific discovery from big data, ensuring that its resources are put to good use to advance the public good."

NIH's Provocative Questions initiative program aims to engage a diverse range of scientists in an exercise that will define and solve perplexing questions in cancer research. For fiscal year 2012 $18 million were distributed to research teams.— Jennifer Brouner, Dec. 11, 2012