The use of artificial intelligence for breast cancer screening is safe and can almost halve the workload of radiologists, according to the world’s largest trial of its kind.
Breast cancer is the most common cancer globally, according to the World Health Organization, with more than 2.3 million women developing the disease each year.
Screening can improve prognosis and reduce mortality by detecting breast cancer at an earlier, more treatable stage. Preliminary results from a large study suggest that AI screening is as good as two radiologists working together, does not increase false positives, and nearly cuts workload in half.
The preliminary safety analysis results of the first randomized controlled trial of its kind involving more than 80,000 women were published in Lancet Oncology journal.
Previous studies investigating whether AI can accurately diagnose breast cancer in mammography were conducted retrospectively, assessing scans that had been examined by clinicians.
But the latest study, which followed women from Sweden with an average age of 54, directly compared AI-assisted screening with standard care.
Half of the scans were assessed by two radiologists, while the other half were assessed by AI-assisted screening followed by interpretation by one or two radiologists.
In total, 244 women (28%) recalled from AI-assisted screening were found to have cancer, compared with 203 women (25%) recalled from standard screening. This resulted in 41 more cancers being detected with the support of AI, of which 19 were invasive and 22 were in situ cancers.
The use of artificial intelligence did not generate more false positives, where a scan is wrongly diagnosed as abnormal. The false-positive rate was 1.5% in both groups.
There were 36,886 fewer screen readings by radiologists in the AI group compared with the group receiving standard care, resulting in a 44% reduction in screen reading workload for radiologists, the authors said.
The final results, looking at whether AI can reduce the number of interval cancers – cases detected between screenings that generally have a worse prognosis – and whether the use of AI in screening is justified, are not expected until several years from now.
But the preliminary analysis concludes: “AI-assisted mammography screening resulted in a similar cancer detection rate compared to standard double-reading, with a significantly lower screen-reading workload, indicating that the use of AI in mammography screening is safe.”
The lead author, Dr. Kristina Lång, from Lund University in Sweden, said: “These promising preliminary safety results should be used to inform new trials and program-based evaluations to address the pronounced shortage of radiologists in many countries, but they are not enough on their own to confirm , that AI is ready to be implemented in mammography screening.
“We still need to understand the implications for patient outcomes, particularly whether the combination of radiologists’ expertise with AI can help detect interval cancers that are often missed by traditional screening, as well as the cost-effectiveness of the technology.
“The biggest potential of AI right now is that it can enable radiologists to be less burdened by the excessive amount of reading.
“While our AI-assisted screening system requires at least one radiologist responsible for detection, it could potentially eliminate the need for double-reading the majority of mammograms, easing workload pressures and allowing radiologists to focus on more advanced diagnostics . to shorten the waiting time for patients.”
Stephen Duffy, professor of cancer screening at Queen Mary University of London, who was not involved in the trial, hailed the “high quality” study but said there may be concerns that AI-driven increases in breast cancer detection could include overdetection of relatively harmless lesions.
“For example, the findings of this paper include an increase in the detection of ductal carcinoma in situ, which is believed to be potentially overdiagnosed,” he said.
Dr. Kotryna Temcinaite, head of research communications at the charity Breast Cancer Now, said the final trial results would ultimately determine whether AI could help improve breast cancer screening.
Meanwhile, she said “urgent issues” in breast screening programs need to be addressed, such as outdated IT systems that take up valuable staff time and delay improvements.
A spokesman for NHS England described the research as “very encouraging” and said it was already investigating how artificial intelligence could help speed up diagnosis for women, detect cancers at an earlier stage and save more lives.
Dr Katharine Halliday, the president of the Royal College of Radiologists, said: “AI holds huge promise and can save clinicians time by maximizing our efficiency, supporting our decision-making and helping to identify and prioritize the most urgent cases.”