AI:HealthData

Making health records clear, useful, and trustworthy with AI

Health records contain a wealth of clinically valuable information, but a large proportion of these data, notably patient journal notes, are not structured, hindering their usability, potentially leading to suboptimal care. AI:HealthData investigates AI methods that extract clinically relevant entities, relations and events from health records and organise them into knowledge graphs, supported by data management techniques for efficient retrieval and analytics. As a first application, AI:HealthData focuses on generating contextual and trustworthy patient summaries for general practitioners, saving time and improving care.

Lab directors:

• Associate Professor Charles Vesteghem
• Associate Professor Daniele Dell’Aglio