Professors at the University who specialize in neuroscience and computer science may be able to help victims of brain disorders get diagnosed earlier with the help of a new computer grid.
A group of University professors, along with Electrical Geodesics Inc. and IBM, are creating a computer grid to speed and improve the diagnosis of several disorders, such as epilepsy, stroke and depression. The grid is called the Integrated Cognitive Neuroscience, Informatics and Computation grid, also known as the ICONIC grid.
EGI Chief Executive Officer Don Tucker and a University psychology professor said researchers at the University Neuroinformatics Center received a $1 million grant from the National Science Foundation to build the grid, and IBM donated the grid’s computers.
He said this project will bring together research of the brain with the technology of computer science.
“Grid computing technology from IBM will play an important role in helping EGI provide doctors and researchers with on-demand access to critical patient data,” Tucker said. “We believe grid computing not only brings performance capacity, but allows the patient security and accountability required for critical medical applications.”
Tucker added that the grid is a new way of doing large-scale computing because a single computer does not have the capability to do such imaging projects.
Computer and information science Professor Allen Malony said the grid will use the University network and will connect different computers using a large file system.
“With the multiple architectures provided by IBM’s products, we can conduct research on ways of optimizing the medical informatics demands for high-performance computing,” Malony said. “At the same time, we can evaluate all of these architectures running the Linux operating system, simplifying our system administration and improving our reliability.”
Malony also said the computers are specifically used for addressing computational processing problems that arise in cognitive neuroscience research.
Tucker and Malony said the idea for the ICONIC grid stemmed from theoretical discussions about the importance of computational methods for understanding the brain and the difficult computational problems that arise while gathering brainwave data.
Malony said they need to integrate different tools and technologies to solve the “difficult computational problems.” He added that it is very important to put the data into a form that neuroscientists can understand, which is what the ICONIC brain grid does.
But Malony said he hopes the work of the ICONIC brain grid will become a part of the problem-solving process, not just a tool.
“I hope we continue to evolve as the ICONIC grid is being developed,” Malony said.
Malony said the ICONIC grid is still in the installation process, but he hopes it will be running by the end of spring term.
Tucker said the grid is made of three computer clusters on campus of up to 16 computers each: one in the Lewis Center for Neuroimaging, one in Deschutes Hall and one in the Riverfront Research Park where EGI and the Neuroinformatics Center are located.
He added that the grid allows rapid sharing of database information between the clusters and the organization of information is more efficient and protected through the grid.
“Every computer has to play well together,” Tucker said.
The University isn’t the only institution experimenting with computer grids. According to http://www.gridcomputingplanet.com, IBM is working with the University of Texas at Austin to build the largest university grid computing project in the nation.
The article states that the purpose of this grid is to connect computing resources across the campus, including personal computers and high-end supercomputers. Those connected to the grid will be able to access massive computer power for simulations, data-sharing and data-intensive calculations in scientific areas.