Fourth AIDAVA General Assembly Meeting on AI-powered Data Curation & Publishing Virtual Assistant in Tallinn, Estonia

The North Estonian Medical Centre in Tallinn hosted the 4th General Assembly of the AIDAVA project from 14 to 16 October. Researchers, clinicians, business analysts and representatives from all project partners gathered together with members of the AIDAVA Ethics and Sustainability Advisory Board to mark an important milestone in reviewing the project's achievements and setting the direction for future development.

"AIDAVA is at the forefront of integrating cutting-edge tools and methods, such as human-in-the-loop data quality checks, to deliver high-quality results to its stakeholders. The team's commitment was evident in Tallinn as we saw how the project is reshaping data management and analysis for critical healthcare applications," said Michel Dumontier from the coordinating institution, the Maastricht University Institute of Data Science.

Remzi Celebi, AIDAVA's technical coordinator, said: “The AIDAVA team has achieved the first major milestone of the project. We have developed and evaluated the first generation of our prototype. While we may still have some issues to address, the prototype already incorporates several innovative features, such as automation of data curation from various sources, data quality checks, and the generation of FHIR IPS from the integrated health record. We are confident that, together, AIDAVA will continue to push the boundaries of innovation and improve the prototype”.

The project, which has just entered its third year, has made significant progress:

• 35 out of 69 deliverables have been completed and approved, with 5 out of 11 milestones successfully achieved.
• 24 deliverables have been approved by the European Commission approval following a successful interim review.
• Six face-to-face workshops and numerous work package specific meetings have been held over the past year.
• Several publications and a strong presence at international conferences have established AIDAVA as a prominent player in the field of health data curation.

The recent evaluation of the first generation of the AIDAVA prototype at clinical sites provided valuable feedback to guide further development of the second generation of the prototype which will start evaluation in January 2026. As the project progresses, it will continue to enhance its data curation capabilities by developing a set of AI-based tools, while assessing the impact of evolving regulatory frameworks, including the Data Act, the Medical Device Regulation (MDR) and the new AI Act.

At the General Assembly, the Sustainability Advisory Board chaired by Terje Peetso and co-led by experts Isabelle de Zegher, and Svetla Boytcheva focused on exploring AIDAVA's potential value propositions with long-term goals and market needs. In addition, the Ethics and Sustainability Advisory Board emphasised the need for equitable access to digital health tools and high ethical standards to promote inclusivity and responsibility. The technical workshops, chaired by Remzi Celebi, focused on the integration specification and on a structured approach for testing the different sub-components of AIDAVA enabling the automation of the curation process in the backend; the team also explored the possibility of developing synthetic data in a privacy compliant manner, reinforcing AIDAVA's commitment to quality and privacy.

Michaela Kargl from the Medical Faculty of the University of Graz points out: "AIDAVA is making progress in revolutionising digital health tools by integrating innovative AI technologies with a focus on ethics and patient-centricity. We are tackling challenges head-on while ensuring that our tools are transparent, compliant and future-proof”.

The Assembly reaffirmed AIDAVA's commitment to four core principles:

1. Patient-Centric Data Control - Empowering patients by giving them control over their health data, at their level of knowledge and readiness, creating high-quality, interoperable records that benefit patients, providers and the healthcare system as a whole.
2. Optimised Data Reuse - Taking a "curate once, use many times" approach to maximise data quality at the patient level, enabling aggregation and reuse at the population level in clinical care, research and policy making.
3. Advanced Data Sharing Standards - Moving beyond traditional point-to-point data exchange by adopting a many-to-many data exchange standard, based on ontological principles, to connect existing standards and facilitate seamless data transformation for analytics.
4. AI-Driven Automation in Data Curation - Leveraging AI technologies, including large language models (LLMs) and machine learning, to automate complex data curation processes. Applications include natural language processing, entity linking, de-duplication and knowledge-based quality checks.

"This meeting has set the course for the next phase of AIDAVA's development. We remain committed to delivering state-of-the-art AI-based solutions that meet the patients’ needs for integrated care through integrated health records, in an increasingly fragmented healthcare environment," concluded Isabelle de Zegher, AIDAVA's Clinical Coordinator.