Artificial intelligence based data curation: enabling a patient-centric European health data space

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Data Governance

Background:

• The European Health Data Space (EHDS) Regulation aims to enhance health data sharing but has limitations regarding individual focus and secondary data generation burdens.
• Current methods for creating secondary datasets from heterogeneous patient data are inefficient and complex.

Purpose of the Study:

• To address limitations in the EHDS by developing an AI-based virtual assistant for automated health data integration and transformation.
• To create interoperable, longitudinal health records for individual patient benefit and efficient secondary data generation for research and policy.

Main Methods:

• Development of an AI-powered virtual assistant to automate the integration and transformation of individual patient health data.
• Implementation of a 'curate once at patient level, use many times' data management strategy.
• Formal description of data source content by data holders to enable data integration.

Main Results:

• Promising preliminary results in automating the integration and transformation of heterogeneous patient data after 15 months.
• Successful conceptualization of a patient-centric EHDS framework.
• Identification of recommendations to facilitate secondary data generation within the EHDS.

Conclusions:

• The AIDAVA project demonstrates a viable approach to overcoming EHDS limitations through AI-driven data management.
• The developed system facilitates a paradigm shift towards patient-centered health data utilization and efficient secondary data creation.
• Recommendations are proposed for a more patient-centric and data-efficient EHDS.