AI4CardioGender

The proposed project aims to develop an advanced AI-powered platform for the integration, harmonization, and analysis of cardiovascular health data with a specific focus on sex and gender differences. The platform will leverage heterogeneous sources such as HIS, LIS, PACS, and structured/unstructured clinical data, centralizing information using interoperable standards (e.g., FHIR, OMOP), overcoming fragmentation in healthcare systems.

By integrating heterogeneous hospital data sources and applying advanced AI models, the project will develop predictive tools and tailored strategies for CVD management. The project will deliver actionable insights, prototypes, and pilot cases to improve equity and outcomes in cardiovascular care.

Through cutting-edge Natural Language Processing (NLP), the system will extract clinically relevant concepts from free-text records (e.g., patient histories, progress notes), transforming them into structured data for advanced analysis. The Medallion architecture (Bronze, Silver, Gold layers) ensures data quality, traceability, and readiness for reporting and visualization.

Built on the Microsoft Fabric and OneLake ecosystem, the platform The platform will be built on a scalable and secure cloud-based ecosystem, which will  supports unified data management, team collaboration, and centralized governance, ensuring compliance with regulations such as GDPR and HIPAA via tools like Microsoft Purview.

Once standardized, the data will be enriched through advanced AI techniques, enabling the development of predictive models and decision-support tools tailored to clinical needs. the data will be enriched using AI models developed within the Fabric Data Science Lab. These models will be tailored to identify sex/gender-specific risk factors, pathways, and disease progression patterns in cardiovascular conditions. AI Enrichment features will support medical image analysis, cohort creation via natural language queries, and predictive modeling for personalized prevention and treatment strategies.

Interactive Power BI dashboards will enable clinicians and researchers to explore data and uncover indicators relevant to sex/gender-specific cardiovascular care. Generative AI capabilities will further enhance data exploration, enabling natural language interaction and automated generation of insights to support inclusive and effective clinical decision-making.