Website for the Turing AI Fellowship programme at the University of Oxford
TURING AI WORLD FELLOWSHIP
TURING AI World Fellowship
The TURING AI World Fellowship is a five-year programme funded by UKRI EPSRC, focused on the ultrasound multi-modal video- based Human-machine collaboration.
As artificial intelligence (AI) starts to be deployed in clinical practice, there are big questions about how this will change decision-making in healthcare. AI for full automation of tasks has received most attention, whereas what is arguably more useful in healthcare is for AI to enable humans and machine to share tasks and decision-making. This shifts the view of AI as an automation technology that replaces human skill, to one that empower individuals; with the AI acting as a teacher, peer or assistant depending on context. However, building human-machine collaboration systems for healthcare is hard, in part as the human decision-making in healthcare can be complex, and understanding of how to do this is in its infancy.
The research team collaborates with clinical partners across two research themes, compromising four work packages, strengthening engagement between academia and industry within the healthcare and medical imaging sectors.
WP1 Single & Multi-modal video-based human-machine collaboration
WP2 Federated Learning for Healthcare Research Collaboration
WP3 Interdisciplinary AI – Understanding Human Skill in Healthcare Settings
WP4 AI Scientist Career Mobility Scheme
Overview of Research Themes
Human-AI Collaboration
The overarching goal of this project is to investigate AI-driven human–machine collaboration, specifically focusing on systems that can determine when and how to appropriately defer to either the human expert or the machine. This work also extends to the design of collaborative frameworks for shared human–machine tasks.
Federated Learning
A second scientific aim is motivated by a practical challenge in data-hungry healthcare learning-based research which is that data governance rules can prevent sharing of data between research sites. A technique called federated learning has recently been proposed to model real-world problems from de-centralised heterogenous data.
