A groundbreaking study by researchers in the UK and Australia has revealed that machine learning software can provide prognoses which are significantly more accurate than those delivered by existing approaches.
The research shows that artificial intelligence can successfully predict how patients with ovarian cancer will respond to different treatments, raising the prospect of breakthroughs in personalised cancer treatment.
While existing treatments classify patients based on the kind of cancer they have and how advanced it is, the machine learning programme developed by the researchers instead categorises patients by the texture of tumours.
The software, developed by researchers at Imperial College London and the University of Melbourne, has been trialled at Hammersmith Hospital, a part of Imperial.
Professor Eric Aboagye, lead author and professor of cancer pharmacology and molecular imaging at the London university, said “the long-term survival rates for patients with advanced ovarian cancer are poor despite the advancements made in cancer treatments”.
“There is an urgent need to find new ways to treat the disease,” he added. “Our technology is able to give clinicians more detailed and accurate information on the how patients are likely to respond to different treatments, which could enable them to make better and more targeted treatment decisions.”
Professor Andrea Rockall, co-author and honorary consultant radiologist at Imperial College Healthcare NHS Trust, said AI has the “potential transform the way healthcare is delivered”.
“Our software is an example of this and we hope that it can be used as a tool to help clinicians with how to best manage and treat patients with ovarian cancer,” she added.
The programme analysed CT scans and tissue samples from more than 350 women with ovarian cancer between 2004 and 2015, in an attempt to identify four characteristics known to influence survival rates: structure, shape, size and genetic makeup. Based on this analysis, the patients were then assigned a score indicating the severity of the cancer. According to the study, the software was four times more accurate at predicting deaths than existing approaches.
Health Minister Nicola Blackwood added: “We must embrace this type of technology to enable clinicians to provide the best possible care on the NHS which is personalised to individuals.”
The study, published in Nature Communications, is just the latest to raise hopes that artificial intelligence could usher in a new era of healthcare. According to a landmark study published in August last year, neural networks developed by researchers at Google’s UK AI division, DeepMind, were as capable as world-class clinicians at detecting more than 50 eye diseases.
The research marked the first time DeepMind’s AI had been applied to healthcare data and lent weight to the idea that medics and intelligent machines will one day work side-by-side.