Researchers discover 'window on the brain' that could help treat disorders
High performance computing process leads to creation of a new company
EUGENE, Ore. -- A discovery by scientists at the University of Oregon's Neuroinformatics Center (NIC) promises to help physicians "see" electrical activity in the brain more clearly, opening the way to improved medical care for epilepsy and other brain disorders.
The process, which recently won a major international award, provided the genesis for a new company, Cerebral Data Systems. The company's debut was announced today (June 20).
University of Oregon psychologist Don Tucker says signals of electrical activity in the human brain become distorted when they pass through the skull, much like light through frosted glass. Measurements of brain activity using electrodes attached to the skull, such as an EEG (electroencephalograph), produce "a smeared picture of brain activity," according to Tucker.
Such distortion hinders treatments that require precision in locating brain electrical activity. To correct for the distortion, physicians must approximate the source of electrical signals by analyzing and averaging EEG readings collected over extended periods of time, or open the skull and attach sensors directly to the inner cranium.
University scientists have used high-performance computing to eliminate the distortion caused by the skull and have developed a computational algorithm to translate data generated by an EEG into a three-dimensional model that pinpoints the location of the activity. A physician or diagnostician then could use this "functional image" to make a medical decision.
Co-developers of the process are Allen Malony, NIC director; Serge Turovets, an NIC computational physicist; and Adnan Salman, a University of Oregon computational science doctoral student. The university has filed for a preliminary patent on the process.
"This new computational advance gives us a window on the brain that could be a breakthrough in the precision of localizing the brain's electrical activity," says Tucker. And that, he says, could be helpful in the treatment of many brain disorders, such as epilepsy and stroke.
"An epileptic seizure is like a storm in the brain," says Tucker, with electrical charges spreading so quickly that it's not obvious where they start. He says surgical techniques exist that could be used to help remedy epileptic seizures, if the epicenter could be determined. "Physicians have only a narrow window of time to determine the location of a stroke and administer clot-busting chemicals," he says.
Tucker and Malony, along with attorney Ann Bunnenberg, are principals in Eugene-based Cerebral Data Systems (CDS). The new company will function as a subsidiary business unit of Electrical Geodesics (EGI), a private entity owned by Tucker and Bunnenberg, with partial ownership by the University of Oregon. It's a structure that Bunnenberg says is as innovative as the product.
"The partnership will expedite the commercialization of new computational neuroscience technologies developed by the NIC by applying procedures required by the regulated medical industry during the preliminary stages of research," she says. The parallel process will create an environment that will bring new discoveries to the market quickly without shortchanging the rigor required to develop a safe and reliable product, she says.
Tucker says CDS also will facilitate application of NIC discoveries to advance computation and telemedicine services pioneered by EGI. He says CDS eventually could employ as many as 100 people.
NIC director Malony says the next research steps will make the process faster and more reliable, integrate it with software developed by the NIC and make it available for use in real applications. The research will be conducted in cooperation with several teaching hospitals, including those at the University of Washington and Harvard University Medical School.
Using high-performance computation to solve the distortion problem recently drew acclaim from the international computer science community.
The researchers' paper describing the discovery, "Computational Modeling of Human Head Conductivity," was one of only three (out of 200 presented) to win an award of excellence, at the Fifth International Conference on Computational Science, held in May in Atlanta, Ga. Malony says their discovery received recognition because it "was not an easy problem to solve and required a good merger of physics, neuroscience and computer science." The paper's authors included Salman, Turovets, Malony, computational physicist Jeff Eriksen and Tucker.
The high performance computations were made possible by a $1 million grant from the National Science Foundation received by the NIC in 2003. With the grant, NIC founders Malony and Tucker established the Integrated Cognitive Neuroscience, Informatics and Computation Grid, a high-performance computing system dedicated to the analysis of imaging data on brain structure and function. The grid harnesses the collective processing power of 100 computers and makes the translational nature of this new process possible.
Electrical Geodesics is a direct spin-off from research done at the Brain Electrophysiology Laboratory in the University of Oregon's Department of Psychology. EGI licenses technology from the university and designs, produces and sells electrophysical neuroimaging equipment and related software. EGI also employs undergraduates, graduate students and postdoctoral fellows, serving as a training ground in the field of cognitive neuroscience.
Source: Eurekalert & othersLast reviewed: By John M. Grohol, Psy.D. on 21 Feb 2009
Published on PsychCentral.com. All rights reserved.