NeuroLOG: sharing neuroimaging data using an ontology-based federated approach

Area of Science:

• Neuroscience
• Computer Science
• Data Management

Background:

• Neuroimaging data is often heterogeneous and distributed across various repositories.
• Managing and sharing these diverse datasets presents significant challenges for collaborative research.

Purpose of the Study:

• To design and describe the NeuroLOG middleware data management layer.
• To enable the sharing of heterogeneous and distributed neuroimaging data through a federated approach.

Main Methods:

• Implementation of a multi-layer application ontology to capture information semantics.
• Development of a derived Federated Schema to align disparate database schemata.
• Integration of a semantic search engine for querying relational data translated into a semantic representation.

Main Results:

• The NeuroLOG middleware facilitates the federation of large neuroimaging datasets.
• The system enables the exploitation of knowledge embedded within the ontology.
• Demonstrated the relevance and feasibility of a distributed approach for neurosciences data management.

Conclusions:

• A federated approach is more realistic and offers strong perspectives for multi-centric neurosciences studies compared to centralized methods.
• The NeuroLOG middleware provides a robust platform for managing and sharing complex neuroimaging data.
• Semantic technologies enhance the accessibility and utility of distributed neuroscientific data.