Research Indicates AI Could Reduce Time Spent Examining Breast Cancer Scans by Half

According to a recent study conducted in Sweden, the implementation of artificial intelligence technology has the potential to reduce the workload of radiologists by nearly 50% when it comes to analyzing routine scans for indications of breast cancer.

The trial’s interim results were considered promising, but the authors cautioned that more research is needed before AI can be used for breast cancer screening on a larger scale.

While increasingly persuasive chatbots like ChatGPT have fueled speculation about possible future applications for AI, one area where the technology has already shown adeptness is in reading medical scans.

With many countries suffering from a shortage of radiologists, there are hopes that artificial intelligence could speed up and refine the time-consuming analysis of routine scans.

This can have a particularly large effect on breast cancer. According to the World Health Organization, more than 2.3 million women were diagnosed with cancer in 2020 alone, and it caused 685,000 deaths.

Regular screening is essential to identify early signs of cancer. In Europe, women between the ages of 50 and 69 are advised to have a mammogram every two years, and two radiologists analyze the result.

The Swedish study looked at 80,000 women who underwent mammograms at four locations in southwestern Sweden between April 2021 and July last year.

– “Very important” –

Their scans were randomly assigned to be analyzed by either an AI-assisted system or two human radiologists who served as a control group.

An artificial intelligence algorithm read the scans and predicted the cancer risk out of ten. A radiologist then checked its predictions.

The AI-backed system detected 20 percent more cancers, the study found, which turned out to be an extra case for every 1,000 women screened.

When it comes to false positives — when a mammogram is first deemed suspicious but later dismissed — both the AI-powered system and two humans detected the same percentage: 1.5 percent.

And radiologists’ workload was reduced by 44 percent in the AI group because only one person was needed to read the scans instead of the usual two.

“The biggest potential of AI right now is that it could give radiologists less of a burden from reading too much,” said Kristina Lang, a radiologist at Lund University in Sweden and lead author of the study.

But Lang said the “promising interim safety results” were not “alone enough to confirm that AI is ready for use in mammography screening,” he said in a statement.

It will take another two years before research can say whether the use of artificial intelligence leads to so-called to a reduction in interval cancers, which are detected between routine checkups, the researchers warned.

Stephen Duffy, a professor of cancer screening at Queen Mary University of London who was not involved in the study, pointed out that the AI algorithm may have overdiagnosed certain forms of early breast cancer called ductal carcinoma in situ.

Nevertheless, he praised the “high-quality research” and said reducing the burden on radiologists’ time was a “significantly important issue in many breast screening programs.”