196. An Introduction to AI Implementation in the UK National Health Service

Abstract:

Artificial intelligence has enormous potential to improve healthcare, but successfully translating AI into routine clinical practice requires much more than developing accurate algorithms.

This webinar will provide an introduction to AI implementation within the UK NHS, using practical examples from medical imaging to explore the evidence required for clinical translation, from validation through to deployment. We will discuss why technically successful AI tools may fail to achieve adoption if they do not address genuine clinical needs or integrate effectively into existing workflows.

Finally, we will present our approach to AI governance within an NHS Trust, including the development of an AI Hub and governance framework to support the safe evaluation and implementation of AI technologies. The session will provide practical insights into the challenges and opportunities of bringing AI from research into clinical care.

About the authors:

Jonathan Taylor is a Consultant Clinical Scientist at Sheffield Teaching Hospitals and an Honorary Lecturer at the University of Sheffield. His role spans Nuclear Medicine, where he practises as a Medical Physics Expert, and the Sheffield 3D Lab, a centralised hospital service for advanced image quantification, visualisation and 3D modelling (www.sheffield3dlab.com). 

He completed a PhD in 2019, focusing on machine learning in Nuclear Medicine. Since then, he has led the development, evaluation and clinical deployment of multiple AI applications in imaging, ranging from automated MUGA processing and chest X-ray classification to MRI kidney volumetry. His current work includes contributing to Sheffield’s AI and Imaging governance group, helping to shape how new digital tools are evaluated, implemented and monitored across the trust. 

Nationally, Jonathan is co-chair of the Institute of Physics and Engineering Medicine’s AI Group, working with colleagues across the UK to advance safe and effective adoption of AI in healthcare. 

Steven Wood (Head of Scientific Computing and Informatics – Sheffield Teaching Hospitals / University of Sheffield)

Managing of a group of 5 registered and 5 research associates, and I am responsible for the overall strategic development of the group. This primarily involves the assessment and deployment of new technologies in the Imaging directorate, and wider trust, to facilitate both the efficiency and quality of the services offered. An integral part of this role is an active involvement in research projects which have the potential to be deployed in live clinical service within the Trust. Since 2015 I have also had the role of Research Informatics Lead for the entire business, working closely with other departments to deliver on the hospitals research and innovation agenda.
In addition to supporting the research activity of others the group also has an international reputation in informatics research projects which currently focus on:

• Natural Language Processing for health records
• De-Identification and linkage of health data for secondary use
• Large scale medical imaging infrastructure for research
• Application of Artificial Inelegance and Machine Learning on real-world health data