AI-supported breast cancer screenings result in fewer aggressive cases

One study found that using artificial intelligence in breast cancer screenings resulted in less aggressive and advanced cancer.

It is estimated that routine screening prevents around 1,300 deaths from breast cancer each year in the UK. More than one in seven people died of breast cancer in the 1990s. Today, the rate is 1 in 20, according to Breast Cancer Now.

Currently, two specialists are needed for mammogram screening, but some cancers still go undetected. But using AI to assist could mean only one expert is needed to complete the same scan safely and efficiently, and may even be more accurate.

World-leading research tested AI-assisted mammography screening on 100,000 Swedish women over two years and found cancer detection increased by almost a third, with no increase in false positives.

AI also helped reduce the rate of breast cancer diagnoses by 12 percent in the years following screening.

Retired thoracic surgeon Dr. who is fighting breast cancer. Liz O’Riordan said: Independent: “Currently every screening mammogram is read by two doctors. This takes time and in some parts of the UK women wait two to three weeks to receive their results. As a cancer patient, I know what this feels like.

“Anxiety can be crippling for women who assume the worst. If the AI ​​reading proves accurate, speeding up the process, allowing women to get results faster, and giving radiologists the freedom to perform other cancer-related tasks, such as image-guided biopsies, that can only be a good thing.”

For published randomized controlled trial Lancet magazine, 100,000 women who attended mammography screening at four sites in Sweden between April 2021 and December 2022 were randomly assigned to either an AI-assisted mammography screening or a standard double-reading process by radiologists without AI.

The AI ​​system has been trained by multiple institutions in more than 10 countries and tested with more than 200,000 exams.

It was then used to analyze mammograms and prioritize low-risk cases based on a single read by radiologists and high-risk cases based on a double read. Artificial intelligence was also used as detection support to radiologists by highlighting suspicious findings in the image.

During two-year follow-up, the AI ​​group had 16 percent less invasive, 21 percent less large, and 27 percent less aggressive subtype cancers compared to the two-specialist group.

While 81 percent of cancer cases in the artificial intelligence-assisted mammography group were detected during screening, 74 percent of cancer cases in the control group were detected.

“Our results potentially justify the use of artificial intelligence to relieve significant pressure on radiologists’ workload, allowing these specialists to focus on other clinical tasks that could reduce patient waiting times,” said study author Jessie Gommers, of Radboud University Medical Center in the Netherlands.

A trial is currently underway in the UK involving 700,000 women, 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, was launched in February last year and if successful, AI could save more lives and shorten waiting lists.

Dr O’Riordan added: “It is great to see the results of this trial showing that AI-assisted mammogram reading is comparable to the current dual-reading standard, and it is great to see that it is now being evaluated for widespread use in Sweden. I hope the results of the EDITH trial in the UK will be equally positive.”