Using AI in breast cancer screenings results in fewer aggressive and advanced cancers, a study has revealed.
It is estimated that routine screening prevents around 1,300 deaths from breast cancer each year in the UK. In the 1990s, more than one in seven people died from breast cancer. Today it is one in 20, according to Breast Cancer Now.
Two specialists are currently needed per mammogram screening, yet some cancers still go undetected. But using AI to help could mean just one specialist is needed to complete the same screening safely and efficiently – and it could even be more accurate.
The world-leading trial tested the AI-supported mammography screening on 100,000 Swedish women over two years, and found cancer detection increased by almost a third without any increase in false positives.
AI also helped to reduce the rate of breast cancer diagnosis by 12 per cent in the years following a screening.
Dr Liz O’Riordan, a retired breast surgeon who has battled breast cancer, told The Independent: “At the moment, every screening mammogram is read by two doctors. This takes time, and in some parts of the UK women are waiting two to three weeks to get their results. As a cancer patient, I know only too well how that feels.
“The anxiety can be crippling for women who are thinking the worst. If AI-reading is proven to be accurate, speeds up the process so women get their results quicker, and frees up radiologists to do other cancer-related tasks like image-guided biopsies, it can only be a good thing.”
For the randomised control trial, which was published in The Lancet journal, 100,000 women who were part of a mammography screening between April 2021 and Dec 2022 at four sites in Sweden were randomly assigned to either an AI-supported mammography screening or to a standard double reading by radiologists without AI.
The AI system was trained and tested with more than 200,000 examinations from multiple institutions across more than 10 countries.
It was then used to analyse the mammograms and triage low-risk cases to single reading and high-risk cases to double reading performed by radiologists. AI was also used as detection support to the radiologists, in which it highlighted suspicious findings in the image.
During the two-year follow-up, there were 16 per cent fewer invasive, 21 per cent fewer large, and 27 per cent fewer aggressive sub-type cancers in the AI group compared to the group with two specialists.
In the AI-supported mammography group, 81 per cent of cancer cases were detected at screening, compared to 74 per cent of cancer cases in the control group.
Study author Jessie Gommers, from the Radboud University Medical Centre in the Netherlands, said: “Our results potentially justify using AI to ease the substantial pressure on radiologists’ workloads, enabling these experts to focus on other clinical tasks, which might shorten the waiting times for patients.”
There is currently an ongoing trial of 700,000 women in the UK testing how AI tools can be used to catch breast cancer earlier.
The cutting-edge trial, called Early Detection using Information Technology in Health, or EDITH, launched in February last year and if successful, AI could save more lives and cut waiting lists.
Dr O’Riordan added: “It’s great to see the results of this trial that show AI-supported mammogram reading is comparable to the current standard of double reading, and that it is now being considered for widespread use in Sweden. I hope the results of the UK EDITH trial will be just as positive.”
