European project puts Artificial Intelligence at the service of health in areas like cancer

To facilitate the generalised access to anonymised medical data towards developing tools to support preventive healthcare and clinical decision-making. This is the main objective of the new European project entitled PHASE IV AI (Privacy Compliant Health Data As A Service For AI Development), featuring INESC TEC; the initiative focuses on high-impact diseases, especially on some types of cancer

With a budget of €8.5M over three years, PHASE IV AI aims to facilitate the access to privacy-compliant data and high-quality computing among European researchers, entrepreneurs, and healthcare industrialists by 2026, facilitating the reduction of time-to-market for data-driven innovation, thus increasing their competitiveness.

Artificial Intelligence (AI) has been a crucial tool in the healthcare sector. However, the processing of large amounts of data, e.g., clinical reports and medical images, quickly and thoroughly, faces certain barriers. On the one hand, the issue of data storage and distribution, and on the other hand, the anonymisation of said data – which would allow the creation of more comprehensive and effective instruments. Developing reliable AI systems requires large amounts of data for concept formation and validation, and using secure, privacy-compliant data is critical to unlocking the full potential of AI and data analytics.

According to Hélder Oliveira, INESC TEC researcher, professor at the Faculty of Sciences of the University of Porto, and scientific coordinator of the project, “PHASE IV AI aims to innovate current data synthesis methods, trying to move towards a more generalised approach to generating synthetic data for the medical sector, including the design of new metrics to evaluate the quality of data generated”. Within the scope of the project, the team will define “protocols and services that allow access to privacy-preserving health data (real-world evidence), through distributed computing, focusing on the development of AI models and algorithms”, explained Artur Rocha, researcher at INESC TEC and technical coordinator of the project.

INESC TEC’s participation in PHASE IV AI focuses on two distinct areas, thus combining the competences of two research Centres – Telecommunications and Multimedia and Human-Centered Computing and Information Science. The researchers in telecommunications and multimedia will focus mainly on AI algorithms for artificial data generation. The other researchers will focus on the development of software technologies, which will allow access to privacy-preserving data through distributed computing.

The work developed within the scope of this project – which, as already mentioned, focuses on high-impact diseases, especially some types of cancer – relies on clinical data environments from several European regions, including Scandinavia, Central Europe, Western Europe, and the United Kingdom. In addition to a wide variety of data, this ensures a broader validation of the tools and services developed in various work environments and professional groups.

The project kick-off meeting took place on October 25 and 26, in Helsinki – hosted by the project leader, the University of Turku.

Besides INESC TEC, PHASE IV AI features 19 more partners from countries like Finland, Belgium, Spain, United Kingdom, Luxembourg, Switzerland, Austria, Italy, and Turkey.

The researchers mentioned in this news piece are associated with INESC TEC.