‘A visionary outlook’

Dominic Cushman, deputy director for AI at NHS England, gave the first keynote address to delegates, exploring the government’s strategic priorities for AI in healthcare and highlighting how it is transforming the NHS to become more efficient, patient-centred and future ready.

Dominic highlighted the importance of developing a “visionary outlook” for the future of the NHS – one where AI enhances timely access to care, improves survival rates through early diagnosis and ensures personalised treatment plans. He also emphasised the importance of collaboration across government, healthcare, academia and industry to ensure AI’s benefits are realised responsibly and equitably. “I call on the audience – healthcare professionals, policymakers and innovators – to actively participate in this transformation, acknowledging their critical role in shaping the future of AI in healthcare,” he concluded.

A common theme that emerged across sessions was that AI was not a “silver bullet” or a “magic wand” that would fix the multitude of problems facing the UK healthcare system – but that, if implemented properly at scale, it could help to ease the challenges pervading the NHS.

‘We’ve moved on from AI being theoretical’

Dr Tracy O'Regan, professional officer for clinical imaging at the SoR, added: “One of the main things that’s coming through is collaboration and building that. We’ve moved on from it being theoretical, and not something that’s achievable, to something that’s now pragmatic. How do we implement things? How do we ensure it’s implemented fairly? What’s the learning that’s needed? Where’s the funding? Where’s the research?

“It’s great to see the number of radiographers and sonographers who were here and presenting. I chaired the session on global healthcare inequalities, and it’s been great to see the power of the organising committee and AI to raise that and aim to do something about it.”

In the global healthcare inequalities session, Dr Ana Luisa Neves, clinical senior lecturer in digital health and director of the Global Digital Health Unit at Imperial College London, discussed some of the issues relating to developing AI in healthcare at scale. She demonstrated various case studies and pilots that had been established at trusts throughout the UK – how their attempts had gone and what challenges they had run into.

One of the primary issues facing healthcare organisations, particularly the NHS, is a general lack of digital maturity. This impacts the speed of establishment, the usefulness of data collected and success in understanding the benefits of implementing AI. Ana explained: “There isn’t enough evidence around the clinical costs and benefits of these technologies. Therefore, they need to follow an evidence generation plan. 

“The market forces drive towards AI imaging – so why is access limited currently? The first factor is digital maturity and infrastructure, but another is the availability and type of data also dictates the type of innovation available. Bias in data can be quite problematic, and there is ultimately a chasm between what can happen in a research lab versus its translation on the clinical frontline. 

“One of the things we see is the development of quite a few solutions that don’t take into account the entire clinical pathway. Using AI to identify and predict certain types of cancer through scans means verifying the results, which can actually lead to more biopsies, and more time in the operating theatre.”

‘Working on change and improvement’

Even putting aside these issues, AI faces significant issues facing usage – for example, the problem of post-market surveillance. 

How can hospitals prove AI effectiveness – in lung cancer screening – if AI predicts lung cancer potentially years into the future? Tracking that kind of patient data runs into issues of patient safety and privacy issues, but that doesn’t mean we shouldn’t try, argued speakers at the ‘Training in AI’ session.

Tom Welton, president of the SoR, said: “I’m really glad we’re getting away from the ‘robots and computers’ ethos, and towards how it integrates into healthcare. How can we shape what the NHS will be moving forward? Knowing that AI provides that real opportunity to achieve what we need to achieve – fewer people on waiting lists, reaching more people with better care – proves we’re on the cusp of something here.

“One of the sessions, run by Tracy O’Regan, focused on healthcare inequalities across the globe, and the stark concerns around AI and putting that into context. We need to be aware of how digital technologies can affect different backgrounds with very different levels of access internationally, and that really stuck with me today.

“It’s been a positive event; lots of people around, lots of radiologists, clinicians and chief operating officers who are looking towards AI diagnostics and therapeutics to work on change and improvement.”

Read more