Developing a Global Imaging Laboratory based on Grid/Cloud computing to help develop drugs for Alzheimer's disease
In order to reduce development time for new effective drugs, and thus improve the quality of life and reduce the social costs of disease, it is necessary to identify markers of disease or biological parameters that allow accurate and early diagnosis of the disease and its progression. According to the latest estimates of EU, the cost of brain diseases in Europe in 2004 was of €386 billion and the global prevalence of Alzheimer's disease is predicted to quadruple to 106 million by 2050. The neuGRID e-Infrastructure can process the world's largest Alzheimer's disease imaging database in ten days, instead of five years. That equates to approximately 6,500 brain MRI scans consisting of over 1.6 million images related to more than 700 patients with Alzheimer's disease and Mild Cognitive impairment. This is tremendously effective because the analysis that can be done today in neuGRID would require more than five years to be run on a single computer.
What has the project achieved?
The Consortium has developed the core e-Infrastructure (neuGRID) required to develop disease markers on extra large brain imaging datasets. Core grid-based applications have been integrated of acknowledged utility for early diagnosis and disease marker development. Preliminary harmonization of the infrastructure has been done with two similar e-Infrastructures in North America (LONI at UCLA in Los Angeles, and CBRAIN in Canada), as a preparatory step to full integration. US and Canadian funding agencies have been aligned to the EC in view of coordinated calls aiming to build a global imaging laboratory based on Grid/Cloud computing. Proactive actions of US funding agencies have been elicited to raise appropriate funds to develop the LONI-based US node of the global infrastructure.
The whole initiative involves four EC-funded projects: neuGRID (2008-11), outGRID (2009-12), DECIDE (2010-12), and N4U (2011-15). The neuGRID Consortium has built the core of the infrastructure (computational nodes, connectivity, middleware, and a set of core applications for image processing). The DECIDE consortium has implemented onto neuGRID a popular tool for the detection of a diagnostic marker for Alzheimer's disease based on PET imaging (gridSPM). The outGRID Consortium has promoted the interchange of technical information of neuGRID with LONI and CBRAIN partners, developed a common strategic agenda, and carried out concertation activities (personal meetings, workshops, public events, media broadcasts) at the global level. N4U will develop the user-facing services of neuGRID and a managing structure that will make it sustainable and profitable.
Why does this matter?
Alzheimer's affects 5 million people worldwide and it is estimated that 13.5 million individuals will manifest clinical AD by the year 2050. The reason why Alzheimer's develops in the brain is being elucidated, and a few drugs aiming to delay or altogether arrest its progression are in an advanced testing phase. Early diagnosis and disease markers to test drugs quickly and efficiently are critical success factors.
Extra large brain imaging datasets are lately available that will allow to develop disease markers. The community of imaging scientists working on Alzheimer's disease need new powerful environments to perform experiments on such datasets.
Who is involved?
Four academic partners leaders in imaging neuroscience and based in Brescia (IT), Amsterdam (NL), Stockholm (SW), and Genève (CH), representing the community working in the fields of Alzheimer's and other neurodegenerative diseases, inflammatory diseases including multiple sclerosis, and psychiatric diseases including bipolar disorder and schizophrenia. Four technical partners based in Archamps (FR), Paris (FR), Bristol (UK), and Sophia-Antipolis (FR), leaders in the field of biomedical image processing and grid computing for biomedical applications. A partner with expertise in dissemination, based in Milano (IT). Two international partners based in Los Angeles (USA) and Montreal (CAN), home of infrastructures (LONI and CBRAIN) homologous to neuGRID.
What is the European added value?
The European-wide approach is clearly mandatory to an initiative that is aiming to change the way science is carried out in a strategic field such as clinical neurosciences. The neuGRID Grid-based e-infrastructure aims building a research environment where a large clinical and image dataset that is being collected Europe-wide can be exploited by a large neuroscience community through the application of sophisticated brain analysis algorithms. The neuGRID e-infrastructure is so far the first and only European effort aiming to offer a distributed working environment to computational neuroscientists. neuGRID has been developed for the front runner community of neuroscientists working in the field of Alzheimer's Disease and other neurodegenerative diseases (i.e. NS-NDD), but designed generically so as to accommodate new services for other biomedical communities in the short term. neuGRID provides integrated and reusable services ranging from simple access to distributed images and clinical data, to advanced imaging pipelines execution and monitoring, to quality control interfaces for experiment outputs and data clean-up, as well as collaboration facilities enabling scientists to exchange and team-up in their daily research work.
How much money has the EU invested in this?
The EU contribution amounts to € 9 240 million overall.