DaRIS/Mediaflux at The Melbourne Brain Centre Imaging Unit
by Dr. Neil Killeen and Dr. Brad Moffat
The Melbourne Brain Centre Imaging Unit at Parkville (part of the National Imaging Facility) operates state of the art imaging systems in the form of Siemens PET/CT and 7-Tesla MR scanners. These systems image a diverse range of subject material including humans, anatomical samples and minerals.
Figure - A demonstration of
the powerful static magnetic field in the 7T MR scanner (the metal bulldog clip is on the end of a chain).
A primary focus is on human neurological disease and the unit is closely engaged in clinical research projects that may lead to new therapies. The Unit places emphasis on being strongly engaged with the research community to enable the best use of its systems. A key part of this is an effective means of managing and delivering data to research teams.
At the heart of the Unit’s data operations is the DaRIS (Distributed and Reflective Informatics System) data management capability. DaRIS has been developed in collaboration with Arcitecta, Monash University and the University of Queensland. DaRIS is a rich set of Mediaflux plugins (including a web-based portal). It’s particular niche is for the handling of bio-medical imaging data for research projects, although it can be used more broadly.
DaRIS receives data directly from imaging scanners. It operates in the dual role of providing the Unit with a comprehensive archive as well as providing researchers with convenient access to their data. Distributed research teams access their data from DaRIS and deposit it in compute platforms or run integrated workflows to analyze it.
The Imaging Unit’s DaRIS system utilises many of Mediaflux’s extensible framework characteristics with a DICOM engine that matches the DaRIS data model, a DICOM client for on-sending data to other DICOM servers, sinks to which (big) data can be deposited (e.g. scp and owncloud), and inter-operability with FileMakerPro servers (ensuring that critical meta-data are synchronized and accurate). DaRIS also supports a RESTful API as well as the service interfaces.
The diagram shows the busy deployment at Parkville (there
is a similar setup at Monash Biomedical Imaging [MBI]). The instruments (MBC Imaging Unit, other NIF, Royal Children’s Hospital and MBI) can send data to the appropriate DaRIS system (e.g. Parkville for parkville precinct people and MBI for Monash precinct
people). The deployment utilises other national infrastructure including RDSI storage, NeCTAR National Servers Program Virtual Machine, fast networks, HPC compute platforms, and ANDS for meta-data harvesting to RDA.
DaRIS is deployed at a number of facilities and platforms around Australia (e.g. the National Imaging Facility). For more information, see the DaRIS wiki or find the code on github.
DaRIS is now available more broadly to the Victorian community via a central VicNode service. If you are interested in using this, or operating your own DaRIS, please contact VicNode (http://www.vicnode.org.au).
For more information about the Melbourne Brain Centre
Imaging Unit see http://www.neuroscience.unimelb.edu.au/content/melbourne-brain-centre-imaging-unit
and also this newsletter article on the Unit -
http://us7.campaign-archive2.com/?u=360e08c4bfc022c47854240f1&id=11d155162a
