CBRAIN

CBRAIN. CBRAIN is a web-based platform designed for the collaborative analysis of large-scale neuroimaging data. It provides researchers with a unified interface to access high-performance computing resources and a suite of neuroinformatics tools. The platform was developed to address the computational challenges inherent in projects like the International Consortium for Brain Mapping and the BigBrain project. Its architecture enables distributed processing across data centers, facilitating large, multi-site studies in neuroscience.

Overview

CBRAIN acts as a middleware portal that connects researchers in fields like cognitive neuroscience and neurology to remote computational infrastructure. It abstracts the complexity of accessing heterogeneous resources such as Compute Canada and the Canadian Brain Imaging Research Platform. The system is particularly aligned with the needs of consortia like the Canadian Institutes of Health Research-funded Co-Add and international efforts such as the Alzheimer's Disease Neuroimaging Initiative. By centralizing data and tool access, it accelerates the pace of discovery in studies of Parkinson's disease and autism spectrum disorder.

Architecture and Design

The platform employs a distributed, service-oriented architecture built primarily with the Ruby on Rails framework. Its core components include a central server that manages user authentication, data provenance, and job scheduling to remote Bourne shell execution sites. Data management leverages technologies like DataCite for persistent identification and integrates with systems like Brain-Imaging Data Structure to ensure standardization. The design emphasizes interoperability with existing cyberinfrastructure, allowing seamless computation on clusters managed by Compute Québec and the Texas Advanced Computing Center.

Functionality and Features

Key functionalities include a graphical interface for launching processing pipelines, real-time monitoring of job status on systems like SciNet, and tools for data visualization. It supports a wide array of neuroimaging software packages, including FMRIB Software Library, Statistical Parametric Mapping, and FreeSurfer. Features like automated data format conversion and integration with the LONI Pipeline facilitate complex analyses. The platform also provides robust data security and sharing mechanisms, compliant with standards from the International Neuroinformatics Coordinating Facility, enabling collaborative projects across institutions like the Montreal Neurological Institute and Hospital.

Development and History

The initial development of CBRAIN began around 2009 at the McGill University Centre for Integrative Neuroscience, led by researchers including Alan Evans and Tristan Glatard. Its creation was driven by the computational demands of the Human Brain Project and the Mouse Brain Architecture Project. Early funding and support came from CANARIE and the National Institutes of Health. The platform has evolved through several major versions, with contributions from the Jülich Research Centre and the INCF, expanding its capabilities to support global brain initiatives like the Brain Research through Advancing Innovative Neurotechnologies Initiative.

Applications and Use Cases

CBRAIN has been deployed in numerous large-scale neuroimaging studies and consortia. It is a core platform for the Prevent-AD cohort study and the Open MEG Archive, facilitating the analysis of electroencephalography and magnetoencephalography data. Use cases include processing diffusion tensor imaging for the Human Connectome Project and managing data for the ENIGMA Consortium. The platform also supports research in pediatric neurology through projects at the Hospital for Sick Children and has been used in clinical studies linked to the Mayo Clinic and Stanford University.