February 21, 2020
We have been proud to sponsor ‘The Future of Health’ series from the Royal Institution – a series of future-focused lectures and discussions with experts from across the healthcare field. February’s topic was ‘The Future of Diagnostics’, held at the Blizard Institute at Queen Mary’s University Hospital. Journalist Danny Buckland attended the event and shares his insights in this second blog on the series.
When technology and science combine to secure vital diagnostic evidence from 13,000 retinal images in the blink of an eye, it would be hard to argue against a future dominated by AI.
The evidence is powerful.
An algorithm, trained on a huge data set, proved the far reaching capabilities of machine learning when it accurately predicted across a range of factors – including tagging blood pressure to within a millimetre of mercury – when put to the test on a new data set.
The research, by Poplin et al, is a clear illustration of a diagnostic future where diseases could be intercepted before they take hold and even register any symptoms.
How society deploys the power of AI across medicine was probed in the Future of Diagnostics – the second of four fascinating Future of Healthcare debates – held at the Blizard Institute at Queen Mary University of London.
Sir Harpal Kumar, head of Johnson & Johnson Innovation EMEA and former CEO of Cancer Research UK, told the audience that technology came with possibilities, challenges and questions.
“The importance of diagnostics cannot be overstated and it is going to be of ever greater importance in coming years,” he said.
“It is not inconceivable to imagine a world in the not too distant future where we will use diagnostics with the ability to diagnose the possibility of someone getting a condition and intervening before they do, so they never develop the disease.”
The event, organized by the Royal Institution, drew an audience from across medical technology, science and academia.
Professor Alastair Denniston, a consultant ophthalmologist at University Hospitals Birmingham who is leading research into the use of health data research and artificial intelligence, introduced Poplin’s data as an example of a rich diagnostic potential.
The algorithm’s performance is way more than an impressive exercise in number-crunching as the data and its accuracy delivers extra information to help decode the intricacies of conditions that are difficult to spot early and treat adequately.
It also holds promise for scientists to define why more women than men suffer macular degeneration, according to Professor Denniston, who is based at the University of Birmingham and the NIHR Biomedical Research Centre for Ophthalmology at Moorfields Eye Hospital/UCL.
It is not inconceivable to imagine a world in the not too distant future where we will use diagnostics with the ability to diagnose the possibility of someone getting a condition and intervening before they do.Sir Harpal Kumar, head of Johnson & Johnson Innovation EMEA
The implications of refining technology and delivering it in settings outside clinics and hospitals, such as high streets and rural areas, holds transformative potential around the world; around 31 million people – 300,000 in the UK – are blind or living with sight issues.
But society’s relationship with technology needs to be nuanced. “Increasingly, we will go to platforms of new technology,” added Professor Denniston. “Does this mean we will totally replace doctors and healthcare professionals? I would argue not.
“There will still be a role for human doctors and healthcare professionals because diagnosis is not just a label. It is a fork in the road and when I get there I want someone to be there with me and I want them to be human. I want to know they know what it is like to be there too and that is where the perfect algorithm will not be enough.”
It is a textured view echoed by Professor Rebecca Fitzgerald, who is based at the Medical Research Council unit at the University of Cambridge, who celebrated future potential but emphasised the need not to over-diagnose and spread anxiety.
However, extra layers of diagnostic information from technology and genetics could identify people at risk of developing diseases and allow early interceptions. Around 55,000 people will survive five or more years if their cancer diagnoses are made at Stage 1 or 2, rather than at advanced stages, according to the UK government’s Long Term Plan for cancer survival.
Professor Fitzgerald also revealed the benefits of diagnostic technology that can be achieved with the Cytosponge process, an ingenious application to detect early oesophageal cancer. Patients, who would normally be put through a lengthy and uncomfortable endoscope procedure, need only swallow a pill attached to a string which expands into a textured sponge in the stomach and can then be drawn back through the oesophagus to collect cells that can be tested for cancer.
The Cytosponge test, which takes around ten minutes, is now in advanced clinical trials and Professor Fitzgerald’s team believes it will allow hundreds of patients to receive early, less invasive and less toxic treatments to cure their conditions.
“Cancer can start developing ten years before it manifests itself, which means we have a great opportunity to detect these changes early and intercept them,” she said. “We might even be able to intercept them in a person who feels completely well before it causes them a problem that is symptomatic.”
Both speakers agreed that technology is leading towards having more diagnostic tests at GP surgeries and High Street pharmacies. Technology and the strategic application of data and diagnostics are clearly potent weapons for healthcare to meet the mounting global healthcare challenges.
The event was supported by Medtronic and was the second of four debates on the Future of Healthcare.
Image credit: Queen Mary University of London