Rutgers wins NSF award to develop online index of brain fMRI
Rutgers University imaging researchers have been awarded a $2 million National Science Foundation grant to develop a new system for indexing and comparing functional MRI brain studies.
The project, funded by the NSF's Cross-Cutting Initiative in Information Technology Research, will support the development of software that allows researchers and clinicians anywhere to use the Internet to find, store, and share related sets of medical images.
The starting point for Rutgers' computer experts, neuroscientists, and engineers is a Napster-like file-sharing system called RUMBA, which allows data sharing over the Internet. Brain-imaging archives omit important information about spatial localization and temporal dynamics, but fMRI can produce a dynamic series of images that map brain activity by showing the pathways brain signals follow, said principal investigator Paul Kantor, Ph.D., a professor of information sciences at Rutgers. The ability to view and interact with this information is one of the project's goals, he said.
"We are interested in indexing these images by context rather than metadata or descriptions of patients' conditions," Kantor said. "The records will stay at the home institution, but the images will be made available to researchers and clinicians on the Internet."
The new technology may help clinicians deal with imaging data so complex that they are nearly impossible to interpret without previous knowledge or handy reference. For a hypothetical patient with a disorder in the anterior cingulate region of the brain, the Rutgers project will give researchers Internet access to thousands of comparative studies to determine how the anterior cingulate, temporal lobe, and other prefrontal regions of the brain interact, said co-principal investigator Steven Hanson, Ph.D.
"Clinicians would begin focusing on patterns or sequences of areas associated with particular mental disorders, such as depression or schizophrenia, rather than local regions of the brain that might distinguish one patient from another in some static sense, but not predict onset or severity," he said.
Patients with these and other mental disorders, including Alzheimer's disease and autism, could benefit from the new technology. Better information translates into better diagnoses, according to Kantor.
"I don't like the idea that the physician who is taking care of me has access to nothing but what is in his mind. I think that's an outdated model," he said.
Original studies are coded with a subject ID to address patient privacy concerns, Hanson said. Information shared over the Internet is rigorously protected under university institutional review board regulations.
