News
►The ADRD (Alzheimer's Disease and Related Disorders) Conclusions Report from the outGRID - ITU workshop held in Geneva on 21st February is now available, presenting the discussion on current context on ADRD research and funding, current problems and challenges, conclusions and follow-up actions.
►A Brain Imaging graduate course is organised by Karolinska Institute for registered students from 19-23 March 2012. Topics of courses will focus on Measurement of Brain Morphology using Magnetic Resonance Images and will be given by various professors. More
►Login planned downtime: the 02/19/2012 from 9:30AM to 2PM. Major VOMS upgrade.
►The EGI Community Forum in conjuction with the 2nd EMI Technical Conference is taking place on 26-30 March in Munich.
►The next N4U Management Meeting is taking place on 21-23 May in Amsterdam.
What was Accomplished with neuGRID
A Success Story
neuGRID delivered an e-Science platform for researchers of Alzheimer's disease, to accelerate their image analysis pipelines through distributed computing infrastructures.
By making available a vast database of 3D brain scans and software tools to identify Alzheimer's disease markers, neuGRID aimed to :
1. accelerate research in Alzheimer's and other neurodegenerative diseases
2. facilitate early detection of memory loss and
3. test the effectiveness of various drug treatments.
The neuGRID project implemented from 2008-2011, was devoted to the development of an online network to provide neuroscientists across Europe with a vast 3D brain scans database and powerful yet accessible computational tools to identify Alzheimer's disease markers.
Extracting markers such as the thinning of the brain cortex on a large patient group could take up to five years, if done on a single mono core computer. The efficiency of the neuGRID infrastructure was proved in September 2009 when neuGRID successfully extracted a sophisticated Alzheimer's disease marker in a record time of two weeks from the largest ever dataset of brain scans: 6,235 brain scans of 715 patients representing a total of 1,300,000 images were processed. Using neuGRID, the analysis was crunched by a factor of 130 compared to a local execution: a total of 5 CPU years was computed in less than 2 weeks. Some 200 CPU cores deployed in 4 European countries were mobilized, moving the brain scan analysis process from a single computer to a highly efficient cloud-computing system.