Coordinator : Michel Dojat PhD, HDR

Project manager : Michael Kain

Local administrative coordination: INRIA

 

 

The Information Analysis Management, IAM hub, was set up to provide a database management and information processing support for imaging population for Human and Animal studies. The aim is to build an infrastructure (hardware and software) accessible to users of national in vivo platforms, to store, manage and process large sets of clinical and preclinical data and associated metadata. 

The node gathers French experts in medical image processing from the regional nodes of FLI with additional experts from INRIA (Rennes), CNRS (ICUBE, CRMBM, CREATIS), CEA (NeuroSpin, MIRCen)  and INSERM (Brest, Nancy, Grenoble), under the scientific coordination of INRIA.

The project’s implementation followed two stages. The first one, 2013 to 2018, was devoted to the building of image analysis and data management solutions allowing the interoperability between heterogeneous and distributed storage solutions implementing raw and meta-data indexing (e.g. through the use of semantic models or ontologies). Four national existing storage solutions have been selected CATI (CEA, Paris-Sud), SHANOIR (INRIA, Rennes), MediBase (CNRS, Strasbourg) and Archimed (INSERM, Nancy). Three national existing solutions for image processing and workflow management have been selected: VIP (CNRS, Lyon), MedInria (INRIA, Rennes) and BrainVisa (CEA, Paris-Sud).

IAM node set up a steering committee defines the action priorities and orientations of the node. A technical manager (M. Kain) recruited on purpose since 2014 has supervised the developments ensured by a team comprising up to 15 software engineers distributed in six different centers. In order to fulfill the maximum types of uses, with the development of generic solution, the IAM node has set up three working groups:

• The first one has worked on interoperability issues, especially the interoperability between the data repository solutions.
• The second one has worked on image processing issues in order to provide with versatile solutions for standalone, or massive parallel computing
• The last one has worked on the transfer of the environment and expertise existing in the clinical imaging field to the emerging field of pre-clinical imaging.

Currently, during the operational stage, the software platform is transferred to an external partner by means of a contract. The partner will offer users, a facility for integration and management of data processing plug-ins building, description and sharing of workflows, and workflow deployment on high performance computing facilities to process large databases. The contract for this stage is under negotiation.

Research teams involved in IAM project  :

INSERM U650 LaTIM, Neuroimagerie Fonctionnelle et de Perfusion (GIN  E. Barbier team, U1260 INSERM UGA), CREATIS, Empenn (ex-VisAGeS, U1228), IADI U947), CIC CIT IT801, LSIIT/ICUBE, CEA-Neurospin, ICM, MIRCen, CRMBM.