CBRAIN: A web-based, distributed computing platform for collaborative neuroimaging research

Abstract

The Canadian Brain Imaging Research Platform (CBRAIN) is a web-based collaborative research platform developed in response to t the challenges raised by data-heavy, compute-intensive neuroimaging research. CBRAIN offers transparent access to remote data sources, distributed computing sites, and an array of processing and visualization tools with in a controlled, secure environment. Its web interface is accessible through any modern browser and uses graphical interface idioms to reduce the technical expertise required to perform large-scale computational analyses. CBRAIN's flexible meta-scheduling has allowed the incorporation of a wide range of heterogeneous computing sites, currently including nine national research High Performance Computing (HPC) centers in Canada, one in Korea, one in Germany, and several local research servers. CBRAIN leverages remote computing cycles and facilitates resource-interoperability in a transparent manner for the end-user. Compared with typical grid solutions available, our architecture was designed to be easily extendable and deployed on existing remote computing sites with no tool modification, administrative intervention, or special softwarehhardware configuration. As October 2013, CBRAIN serves over 200 users spread across 53 cities in 17 countries. The platform is built as a generic framework that can accept data and analysis tools from any discipline. However, its current focus is primarily on neuroimaging research and studies of neurological diseases such as Autism, Parkinson's and Alzheimer's diseases, Multiple Sclerosis as well as on normal brain structure and development. This technical report presents the CBRAIN Platform, its current deployment and usage and future direction. © 2014 Sherif, Rioux, Rousseau, Kassis, Beck, Adalat, Das, Glatard and Evans.