The EuCanImage consortium has unveiled a secure and federated large-scale European cancer platform designed to significantly enhance the potential of artificial intelligence (AI) in oncology. This innovative platform curates and annotates new data resources from over 25,000 individuals, making them publicly accessible to address unmet clinical needs.
Key Features of the Platform:
- Curated and Annotated Imaging Data: Comprehensive imaging data is meticulously curated and annotated, linked to biological and health records.
- Multi-Scale AI Solutions: Integrates organ-level, molecular, and clinical predictors to create detailed cancer fingerprints.
- Trusted Data Management: Employs secure, well-established data infrastructures and curation tools to enhance AI cancer imaging.
AI Modeling
EuCanImage is developing a suite of radiomics methods and AI algorithms to build advanced integrative AI models using extensive imaging and non-imaging data. These resources are available through the AI Virtual Research Environment, a portable computational space that supports the development and validation of AI tools.
- Cancer Image Feature Extraction and Selection Tools
- Machine-Learning Pipeline for Predictive Modeling
- AI Development Platform for Running Experiments
Researchers can run EuCanImage AI tools in a secure, private execution environment. The user-friendly web interface allows for easy dataset uploads or imports from EuCanImage Data Repositories. A pilot installation is currently operational at the Barcelona Supercomputing Center.
This platform represents a significant leap forward in the application of AI to oncology, providing robust tools and data resources to drive groundbreaking research and clinical advancements.
