Biomedical Research Feels the BIRN

The Biomedical Informatics Research Network is pioneering the use of advanced cyberinfrastructure for biomedical research, starting with the mysteries of the human and animal brain. Shape analysis in Alzheimer's disease from the Morphometry BIRN test bed. Image Courtesy of The John Hopkins University, Center for Imaging Science
The BIRN's main goal is to get biomedical scientists collaborating like never before. Starting with four projects, or test beds, the BIRN is discovering which information technologies are wanted and needed by—and useful for—the biomedical community. The technologies help researchers share ideas, software and data; access high-performance computing resources; and deal with the explosion of data from new research techniques. "The amount of data we acquire in biomedical research now is huge and dollars are extremely tight," says Mark Ellisman, director of the BIRN Coordinating Center. "The BIRN is a way of distributing and reducing the costs by using advanced technology to make data and resource sharing easier." BIRN, a National Institutes of Health-funded initiative, already impacts the study of cognitive impairment associated with aging. Researchers use the infrastructure to analyze MRI images of human brains, looking for changes in the size and shape of certain parts of the brain that might signal the onset of Alzheimer's disease. This requires making MRI images collected from different institutions comparable, a great technical challenge. "If you went out and gave everybody digital cameras from different manufacturers and had them all shoot close-ups of the same thing, each picture would be a little different," explains Ellisman. "In order to do a large, geographically distributed population study using instruments like MRI, you need software tools that allow you to massage all the data to get them aligned in a similar way. Once they're aligned, other tools can be used to extract from these complicated 3D scenes the object you want to compare." The BIRN provides a way for scientists to share the MRI data and to access the high-performance computing resources needed to analyze the images. The second BIRN test bed, which focuses on schizophrenia, poses a similar set of challenges: integrating terabytes of similar image data from many different researchers and making sense of the data. The third test bed integrates vastly different types of data from the brains of strains of mice created to model human diseases like Alzheimer's and Parkinsons. MRI images are integrated with images from advanced laser-light microscopes and high-energy electron microscopes. The fourth, and newest, test bed focuses on the need for collaboration in non-human primate research. Project leaders hope to link genomic, imaging and behavioral data from the Yerkes National Primate Research Center in Atlanta with that from other national primate centers and the three other BIRN test beds. Since its beginning in 2001, the BIRN collaboration has grown to almost 300 people and more than 30 sites in the UK and the United States. It's the only large NIH research project with large-scale participation from the computer science community—appoximately 40% of the scientists participating in BIRN are information technology researchers and developers. Learn more at the BIRN Web site. - Katie Yurkewicz, Science Grid This Week