Poster by: Neil Killeen, Jason Lohrey, Wilson Liu, Steve Melnikoff and Gary Egan.

The contemporary E-research paradigm engages scientists in multi-disciplinary, distributed, heterogeneous, dynamic and collaborative environments. Correspondingly, the software tools required to meet these demands must be increasingly sophisticated to provide sufficient adaptability and flexibility.

Fulfilling these over-arching requirements, a Neuro-informatics system that provides for the reception, management and distribution of Neuroimaging data and meta-data has been developed. The foundations of the system that enable collaborative E-research are:
• Adaptive (reflective) “Web 2.0” and Java interfaces
• Meta-data are moved through the system as XML
• The interfaces are driven by a generic ‘subject-centric’ Data Model
• The Data Model captures the Research method and embodies strong semantics
• The ability to handle arbitrary (but specified) meta-data and content
• The ability to federate resources
• Project- and role-based access permissions
• The repository layer is built with the Mediaflux Digital Asset Management platform
• A modular and service oriented architecture (SOA)

This system manages data for over 30 research projects, containing Magnetic Resonance (MR) and Microscopy imaging data (human and animal) and the accompanying meta-data. As well, it operates with a variety of Research methods.The key components of this system, useful experiences and lessons learnt from the development, deployment and operations will be related.

One of the most important goals of any modern informatics system is to minimize development time through rapid deployment and re-use of existing components. Because of its modular and reflective architecture, and use of modern web technologies, the web-based interface can be straightforwardly leveraged to rapidly develop other specialized interfaces such as the ‘Abstract Explorer’ for the Human Brain Mapping 2008 Conference.

Another important aspect of modern E-research is to recognize that systems no longer exist in isolation. The development of loosely coupled systems, built from a heterogenous collection of independently designed and implemented components is probably the greatest challenge. Because the system described above features an interface exposed in a SOA and a Data Model with strong semantics, it is well-placed to participate in such systems and is, in fact, a loosely-coupled system by itself.