A unique multi-stakeholder workshop on Artificial Intelligence (AI) and Healthcare was organised with the High-Level Expert Group on AI (AI HLEG) in December 2018 at the European Commission. With the goal to “imagine a healthcare system where AI drives human outcomes to its full potential, for all European citizens, in a safe and ethical way”, participants were invited to an ambitious discussion that looked towards the future.
This goal was taken from the AI HLEG's broader mission: to develop feasible policy recommendations for AI uptake in Europe, and to position Europe as a global leading innovator in ethical, secure and human-centric AI.
Participants were key healthcare-stakeholders, including thought-leaders from pharmaceuticals, academia, healthcare practitioners, European and national healthcare regulators, insurers, and AI vendors.
The workshop was designed to investigate these key stakeholders’ perspective on AI in healthcare, which is now being mapped and transformed into specific recommendations for the European Commission and EU Member States. The discussion focused on a key set of barriers and enablers for AI adoption in Europe:
- Data access
- Physical infrastructure (including all enablers for data infrastructures, compute infrastructures)
- Skills and education (including public engagement strategies)
- Research (including all enablers for research strategies)
- Innovation (including sandboxes, innovative strategies, startups, SMEs)
- Policy and regulation (including ethics, policy and regulation, sustainability)
- Funding & Investment
The workshop’s initial findings show that the healthcare industry is particularly concerned with three key areas:
- Data access: how can Europe ensure that data is accessible across borders and stakeholders? => Recommendations included new data repositories and synthetic data test bed collections.
- Interoperability: how can data be viable across different systems of data gathering and labelling? => Recommendations included measures to harmonise regulation between member states, universal Electronic Healthcare Records (EMRs), and common data models.
- Patient control over data: how can we ensure that patients are aware and in control of their data? => Among the key recommendations received; was the supply of European citizens with data donor cards, enabling them to intentionally give access to their health data for AI model training purposes.
In the coming months, together with Saskia Steinacker, we will lead the EU Commission’s High-Level Expert Healthcare sub-group on AI to develop specific recommendations that will support the uptake of AI in healthcare.
Based on extensive research and analyses, these recommendations will be published in 2019.
Are you a healthcare stakeholder? Would you like to contribute to the European Commission healthcare deep dive research on AI? Please contribute your thoughts through this survey