According to preliminary analysis from a study being undertaken in Sweden, the use of artificial intelligence (AI) in breast cancer screening has been indicated to be “safe” and could significantly reduce radiologists’ workloads.

The study, which is still ongoing, is being undertaken to assess the clinical safety of an AI-supported screen-reading protocol compared with standard screen reading by radiologists following mammograms — X-rays of the breast.

Previous studies looking at whether AI can help to accurately diagnose breast cancer in mammograms were carried out retrospectively, assessing scans that had already been looked at by radiologists.

However, interim results of this latest study have been published by the group of researchers in the peer-reviewed medical journal The Lancet, with the lead author being Kristina Lång at Lund University.

The researchers held a randomised, controlled trial involving roughly 80,000 women from Sweden with a median age of 54-years-old. Around half (40,024) of the participants’ mammograms were assessed by two radiologists, while the other half (39, 996) were assessed by the AI-supported tool followed by interpretation by radiologists.

In total, 244 women from AI-supported screening were found to have breast cancer, while 203 women from standard screening were found to have cancer, suggesting that AI-aided detection can identify cancer at a “similar rate” to two radiologists.

Additionally, the trial found that the workload for the radiologists working as part of the AI-supported group was reduced by 44%, seeing as only one radiologist was needed as opposed to the normal two required for double reading.

Recommended

• Scotland to Benefit from Faster Radiology Reporting
• NHS Scotland IT Glitch Causes 1,800 Women to Miss Breast Cancer Screenings
• Tech Experts Warn Against Using AI in Breast Cancer Screenings

While “The greatest potential of AI right now is that it could allow radiologists to be less burdened by the excessive amount of reading,” Lång also said that the initial results “are not enough on their own to confirm that AI is ready to be implemented in mammography screening,” according to The Guardian.

Further, Lång said that “The study was conducted on a single site in a Swedish setting. We need to see whether these promising results hold up under other conditions, for example with other radiologists or other AI algorithms. There may be other ways to use AI in mammography screening, but these should preferably also need to be investigated in a prospective setting.”

Last month, the World Health Organisation highlighted that breast cancer is the most prevalent cancer around the world, with over 2.3 million women diagnosed with it in the year of 2020.