Last modified: Wednesday, October 10, 2012
'Google of the brain' neuroimaging project receives $2.5 million NIH grant
FOR IMMEDIATE RELEASE
Oct. 10, 2012
BLOOMINGTON, Ind. -- Indiana University Bloomington cognitive scientist Michael Jones, in collaboration with researchers at the University of Colorado, University of Texas at Austin and Washington University in St. Louis, was awarded $2.5 million from the National Institute of Mental Health to develop an automated system for large-scale synthesis of human neuroimaging data.
The four-year award will support the development of NeuroSynth.org, an online platform that is intended to be sort of a "Google of the brain" for researchers in cognitive neuroscience. The unique system will be designed to learn new concepts, draw inferences and make discoveries based on the collected sources.
"There is a vast amount of so-called 'unrealized knowledge' across a number of scientific sources -- yet-to-be discovered information that is not located in any specific article, but is rather distributed across many," Jones said. "Scientists are regularly reading distinct but related articles to make these discoveries, and NeuroSynth will attempt to simulate and scale up this knowledge discovery process, generating novel hypotheses to test with future experiments."
The explosive growth in the study of human neuroimaging has led to major advances in the understanding of normal and abnormal brain function, but the sheer volume of neuroimaging findings is consequently difficult to synthesize. A sophisticated understanding about how brain systems are associated with various psychological processes depends on a thorough understanding of the scientific work that already has been done.
The system will automatically integrate findings from tens of thousands of neuroimaging articles. To look at a concept such as "pain," for example, NeuroSynth will be able to draw on an entire body of work to reveal the brain regions that are consistently activated in all experiments that have studied pain. Researchers may also select particular brain regions, and the system will be able to compute how closely a given concept is associated with a certain region -- how often, for example, is the anterior cingulate cortex activated in studies of pain.
NeuroSynth will be completely automated to integrate new neuroimaging studies as they are published and update its beliefs about links between brain regions and psychological processes in real time.
The beta site for NeuroSynth.org gives a small sample of what is planned, explains the project and includes an FAQ section.
Jones directs the Cognitive Computing Laboratory in the Department of Psychological and Brain Sciences and Cognitive Science Program, which specializes in computational models of concept learning from natural language. The lab will spearhead development of NeuroSynth's ability to mine text and make conceptual inferences from the many sources it organizes. Jones' models of human concept learning will be built into NeuroSynth, allowing it to learn new concepts being discussed in scientific papers to link to the corresponding brain imaging data. Jones' models will also give NeuroSynth the ability to make new discoveries by integrating knowledge across thousands of articles.
The NeuroSynth tools will be freely and publicly available. When its knowledge of the neuroimaging literature has grown to a sufficiently large scale, the system will allow for powerful new discoveries in cognitive neuroscience and will enable rapid and efficient application to a broad range of clinical and basic research applications.
Jones is an assistant professor in the Department of Psychological and Brain Sciences and the Cognitive Science Program in the College of Arts and Sciences, and is an adjunct professor of Informatics in the School of Informatics and Computing. He was recently awarded the CAREER award from the National Science Foundation, the foundation's most prestigious award for elite early-career scientists. For more information, visit the IU Cognitive Computing Laboratory website.
Tal Yarkoni, University of Colorado, is principal investigator. Co-principal investigators are Jones; Tor Wager, University of Colorado; and Russell Poldrack, University of Texas at Austin.