More cases of breast cancer are found when AI is used in screenings, research shows

The use of artificial intelligence in breast cancer screening increases the chance of detecting the disease, researchers have found, in what they say is the first real test of the approach.

Numerous studies have suggested that AI can help medical professionals detect cancer, whether it’s identifying abnormal growths on CT scans or signs of breast cancer on mammograms.

However, many studies are retrospective – meaning AI isn’t involved from the start – while studies that take the opposite approach often have small sample sizes. Importantly, larger studies do not necessarily reflect real-world use.

Now researchers say they have tested AI for the first time in a nationwide screening program, showing its benefits in a real-world environment.

Professor Alexander Katalinic, a co-author of the study from the University of Lübeck in Germany, said: “We could improve the detection rate without increasing the harm to the women who participate in breast cancer screening,” adding that the approach could also reduce the risk of breast cancer. workload of radiologists.

Katalinic and his colleagues analyzed data from 461,818 women in Germany who underwent breast cancer screening between July 2021 and February 2023 as part of a national program aimed at asymptomatic women aged 50-69.

In all women, the scans were examined independently by two radiologists. However, for 260,739 women, at least one of the experts used an AI tool to support them.

The AI ​​tool not only visibly labels scans it considers non-suspicious as ‘normal’, but also provides a ‘safety net’ warning when a scan it considers suspicious is assessed as non-suspicious by the radiologist. In that case, the tool also highlights the area of ​​the scan that it suggests needs further investigation.

A total of 2,881 of the women took part in the study, which was published in the journal Nature Medicinewas diagnosed with breast cancer. The detection rate was 6.7% higher in the AI ​​group. However, after taking into account factors such as the age of the women and the radiologists involved, the researchers found that this difference increased, with the rate being 17.6% higher for the AI ​​group, at 6.70 per 1,000 women, compared to 5.70 per 1,000 women for the standard group. . In other words, one additional case of cancer was noted per 1,000 women screened when AI was used.

Crucially, the team said the number of women recalled for further investigation following a suspicious scan was about the same.

“In our study, we had a higher detection rate without a higher false positive rate,” says Katalinic. “This is a better result, with the same damage.”

The team said the tool’s “safety net” was activated 3,959 times in the AI ​​group, and led to 204 breast cancer diagnoses. In contrast, 20 breast cancer diagnoses in the AI ​​group would have been missed if doctors had not examined the scans that AI considered “normal.”

Stefan Bunk, another co-author and co-founder of Vara, the company that built the AI ​​tool, said the technology increased the speed at which radiologists examined scans marked as “normal.” Calculations even showed that these scans were not assessed. According to experts, the overall breast cancer detection rate would be higher and the recall rate lower than without the tool. According to him, this meant fewer false positives for women and a lower workload for radiologists.

Stephen Duffy, emeritus professor of cancer screening at Queen Mary University of London, who was not involved in the work, said the results are credible and impressive.

“Here in the UK there is particular interest in whether the use of AI plus one radiologist can safely replace reading by two radiologists. The sooner this is definitively investigated, the better,” he said.

Dr. Kristina Lång, from Lund University, said the study adds to the growing body of evidence supporting the potential benefits of incorporating AI into mammography screening. But she added that the large increase in the number of in situ cancers detected raises concerns because these cancers are more likely to be slow growing and could contribute to the overdiagnosis burden of screening.

“Long-term follow-up is essential to fully understand the clinical implications of integrating AI into mammography screening,” she said. “The results are encouraging, but it is essential to ensure we implement a method capable of detecting clinically relevant cancers at an early stage, where early detection can meaningfully improve patient outcomes.”

Dr. Katharine Halliday, president of the Royal College of Radiologists, said the organisation’s most recent census showed a 29% shortage of radiologists in the NHS.

“Any tools that can increase our accuracy and productivity are welcome. “But while the potential benefits are significant, so are the potential risks,” she said. “It is crucial that the use of AI in the NHS is done carefully, under the supervision of experts.”