AI technology can identify 42 per cent of breast cancers that develop between routine mammograms by flagging women at highest risk, researchers have found.

The study analysed 134,217 screening mammograms using a deep learning model known as Mirai, which estimates breast cancer risk from imaging data, tumour features and breast density.

Researchers from the University of Cambridge and Addenbrooke’s Hospital tested the tool on UK screening data collected between 2014 and 2016.

They identified 524 “interval cancers” — cases diagnosed between regular screening appointments.

The algorithm’s risk scores predicted 42.4 per cent of these interval cancers among women ranked in the highest 20 per cent for risk, equating to an additional detection rate of 1.7 cancers per 1,000 women screened.

Interval cancers tend to be more advanced when diagnosed, often larger or more aggressive than cancers found during routine screenings.

Co-author Fiona J. Gilbert is professor of radiology at the University of Cambridge and honorary consultant radiologist at Addenbrooke’s Hospital.

Gilbert said: “Interval cancers generally have a worse prognosis compared with screen-detected cancers because they tend to be either larger or more aggressive.

“That’s why it’s important to minimise the number of interval cancers that you have in any screening programme.”

The tool was most accurate for cancers developing within a year of screening, performing less well for those emerging after 12 to 36 months.

It was also less effective for women with extremely dense breast tissue — tissue with more glands and fibrous material than fat — which can make tumours harder to detect.

However, Mirai still outperformed conventional risk assessment tools.

Lead researcher Joshua W. D. Rothwell, an M.B.B.S./Ph.D. student at the University of Cambridge, said the findings could help identify women who need closer monitoring or additional imaging.

Rothwell said: “Our results suggest that further workup of mammograms within the top 20 per cent of scores could yield 42.4 per cent of interval cancers, meaning that Mirai could be used to identify women for supplemental imaging or a shortened screening interval.

In the UK, around 2.2 million women undergo breast screening each year through the triennial programme, which invites women for mammograms every three years.

The technology could refine this process by identifying those who would benefit from extra imaging such as MRI or contrast-enhanced mammography — an X-ray technique that uses dye to highlight areas of concern.

Gilbert said: “If we called back 20 per cent of women for supplemental imaging, we’d have to find the capacity to offer contrast-enhanced mammography or MRI to 440,000 women.

“Personalised breast cancer screening depends on accurately assessing an individual’s risk of developing breast cancer within a specific timeframe.

“We can use supplemental imaging and adjust screening frequency based on a woman’s breast density and likelihood of developing breast cancer within a short timeframe.”

Next steps include comparing commercially available predictive tools, conducting cost-effectiveness studies and running trials to identify which women would benefit most from additional imaging.

“Identifying women at increased risk of developing breast cancer is a complex, multifactorial problem,” Dr Gilbert said.

“The goal is to accurately identify the women most likely to have an interval cancer while minimising the amount of additional imaging required.”