Section site map: News

• News Home
• Media Press Room
• User/System News

2005 News Stories

• GADU/GNARE Uses TeraGrid For Protein Sequence Analysis
• TeraGyroid: Gyroids on the TeraGrid
• TeraShake: Simulating a Big Shake in Southern California Basins
• Seismic Modeling and Oil Reservoir Simulations with TeraGrid
• Improving Groundwater Cleanup Decision-Making with TeraGrid
• Harnessing TeraGrid to the Sky: Understanding Dark Energy
• AMANDA and TeraGrid: Exploring the Violent Universe
• TeraGrid Science Gateways: NanoHUB
• Identifying Brain Disorders with TeraGrid

News

Identifying Brain Disorders with TeraGrid

Through the Biomedical Informatics Research Network (BIRN) - a consortium of 15 universities and 22 research groups - scientists are investigating how to identify structural differences in patients' brains as a way to help clinicians diagnose pathologies such as Alzheimer's disease and dementia. TeraGrid resources facilitate computation and large-scale analyses of patient data acquired and pooled across collaborating sites.

The Center for Imaging Science (CIS) at Johns Hopkins University and other BIRN researchers collaborated on a processing pipeline for seamless analysis of shape data. Hippocampal data from 45 subjects in three categories - Alzheimer's, Semantic Dementia, and control subjects - was scanned via high-resolution MRI. CIS researchers then aligned and processed - using the Large Deformation Diffeomorphic Metric Mapping (LDDMM) tool - on the TeraGrid Itanium IA-64 clusters at NCSA and SDSC. LDDMM creates a mathematical description that allows quantitative characterization of differences in brain-structure shapes.

The researchers ran comparisons of 45-by-45 matrices. This large computational task used an average of 70 Itanium nodes for three weeks (over 30,000 processor hours) and generated four terabytes of data. The resulting data was then uploaded into the Storage Resource Broker (SRB) at SDSC for sharing among institutions and further analysis. BIRN researchers were able to successfully identify the three categories of subjects as collective groups. As BIRN learns how to use their tools to reliably identify different subject groups, they are paving the way for resources to aid clinicians in better identifying brain disorders, which will have major benefits in patient health.

Figure 1: The Large Deformation Diffeomorphic Metric Mapping tool computed shape analysis of scanned brain structures on the TeraGrid, enabling researchers to successfully distinguish Alzheimer's from other brain conditions. Image credit: BIRN, CIS at Johns Hopkins University.

More News Releases

Top