THE Connect podcast: AI offers a new view on healthcare
From proteins to the health of a nation, AI has tremendous scope for impact in the healthcare section
Sponsored by Queen's University Belfast
Sponsored by
There is “tremendous scope” for AI to have an impact in healthcare across a wide range of people and diseases, said Iain Styles, a professor in the School of Electronics, Electrical Engineering and Computer Science at Queen’s University Belfast, on an episode of the THE Connect podcast.
“If we can get AI to uncover patterns and structures in data that tells us something about the health of the nation and perhaps the health of individuals, there’s a potentially enormous social benefit,” Styles said.
Privacy around health data has been a major concern, but many important areas of research do not infringe on privacy, said Styles, who is also the director of the Centre for Intelligent Sustainable Computing at Queen’s. For example, AI can help identify drug targets, new candidate molecules for therapeutics, and understand the mechanisms of disease “all without compromising patient privacy”, Styles said.
Such research requires interdisciplinary collaboration, something that Styles is actively promoting. “It is critically important that we have people from the biomedical sciences working with people from the computational sciences and that we don’t let either side go it alone,” he said. Styles advocates a model for training scientists in which students are co-supervised by academics from multiple disciplines, an approach that is at the heart of a new Northern Ireland-based doctoral programme in AI for bioscience.
In a long-term collaboration with University of Birmingham professor of chemistry Helen Cooper, Styles is helping to develop new technologies to identify proteins and protein complexes that are implicated in diseases. “As the technology evolves, we will be able to tell what molecules are involved in the disease, where they are in the native tissue context,” he said. “That will feed into the development of new therapeutic agents, and then we’ll be able to use the same technology to understand where the drugs go and what they do when they get there.”
Researchers are only beginning to understand the implications of AI in healthcare, he said. For example, AlphaFold, developed by DeepMind, is an AI programme that predicts protein structures. Its developers won the 2024 Nobel Prize in Chemistry. “It really showed the potential of AI technologies in undercovering the fundamental processes of life.”
A major challenge is in “recruiting fair and balanced patient cohorts”, Styles said. “AI models trained in Northern Ireland would not necessarily be readily applicable to other parts of the UK, and vice versa, because their demographics are quite different.” There is a push to be more open in labelling data without disclosing the data itself, Styles said. “Tackling biases is critically dependent on that openness.”
For young researchers, healthcare AI is a vibrant field with the opportunity to have an impact outside of academia, he said. “It’s very possible to have a successful career through the standard academic career path” but there’s also scope outside of academia. “We will see very exciting new businesses and industries coming out of this,” he said.